技术领域technical field
本申请涉及商用车网联大数据应用技术领域，特别是涉及一种油品识别方法、装置、计算机设备、存储介质和计算机程序产品。The present application relates to the technical field of commercial vehicle networking big data application, and in particular to an oil product identification method, device, computer equipment, storage medium and computer program product.
背景技术Background technique
柴油是商用车产业的生命线。除了符合国家标准的车用柴油，我国柴油市场上还存在一种含硫量远超国家标准的劣质柴油，称为“高硫柴油”。车辆长时间使用高硫柴油后会引起车辆机械结构腐蚀、后处理系统颗粒捕集器堵塞(DPF)和选择性催化氧化器(SCR)硫中毒等故障，进而导致后处理系统崩溃和车辆寿命受损；另外，用户因违规加注高硫柴油引起车辆故障前往企业售后部门维修，二者往往会在责任界定等方面产生矛盾甚至陷入法律纠纷。高硫柴油在市场上的流行破坏了燃油市场，有悖排放法规，加重了企业运维部门的负担，远程监控目标车辆燃油品质、识别何时使用高硫柴油对企业具有重要意义。Diesel is the lifeline of the commercial vehicle industry. In addition to vehicle diesel that meets the national standards, there is also a low-quality diesel with a sulfur content far exceeding the national standards in my country's diesel market, called "high-sulfur diesel." Long-term use of high-sulfur diesel in vehicles will cause failures such as vehicle mechanical structure corrosion, after-treatment system particulate filter (DPF) clogging, and selective catalytic oxidizer (SCR) sulfur poisoning, which will lead to after-treatment system breakdown and shortened vehicle life. In addition, users go to the after-sales department of the company for repairs due to vehicle failures caused by illegal filling of high-sulfur diesel. The two often have conflicts in terms of responsibility definition and even fall into legal disputes. The popularity of high-sulfur diesel in the market has destroyed the fuel market, violated emission regulations, and increased the burden on the operation and maintenance department of enterprises. Remotely monitoring the fuel quality of target vehicles and identifying when to use high-sulfur diesel is of great significance to enterprises.
高硫柴油识别技术的相关方案主要是判断SCR性能、SCR转换效率是否下降，进而判断是否进行高硫柴油警报。但是，这种技术方案对仅通过SCR上下游数据识别油品，数据的利用程度不足，油品识别精度受限，并且规则参数的标定一般基于分辨率高、稳定的台架数据，并未考虑到网联场景下车网联数据存在不同程度的干扰因素，这些干扰因素严重影响最终的油品识别结果，油品识别的鲁棒性也有待提升。The relevant schemes of high-sulfur diesel identification technology are mainly to judge whether the SCR performance and SCR conversion efficiency have declined, and then judge whether to issue a high-sulfur diesel alarm. However, this technical solution only uses SCR upstream and downstream data to identify oil products, and the utilization of data is insufficient, and the accuracy of oil product identification is limited, and the calibration of rule parameters is generally based on high-resolution and stable bench data. In the Internet-connected scenario, there are different degrees of interference factors in the Internet-of-vehicle data. These interference factors seriously affect the final oil product identification results, and the robustness of oil product identification also needs to be improved.
目前的油品识别精度受限且鲁棒性不高。The current oil identification accuracy is limited and the robustness is not high.
发明内容Contents of the invention
基于此，有必要针对上述技术问题，提供一种能够提高油品识别精度和鲁棒性的油品识别方法、装置、计算机设备、计算机可读存储介质和计算机程序产品。Based on this, it is necessary to provide an oil product identification method, device, computer equipment, computer readable storage medium and computer program product that can improve the accuracy and robustness of oil product identification in order to address the above technical problems.
第一方面，本申请提供了一种油品识别方法。所述方法包括：In a first aspect, the present application provides a method for identifying oil products. The methods include:
根据目标车辆的网联数据获取目标车辆的数据序列，对数据序列进行分割，在数据序列中确定序列片段；数据序列包括目标车辆在多个单位时间的整车工况信息；Obtain the data sequence of the target vehicle according to the network data of the target vehicle, segment the data sequence, and determine the sequence fragments in the data sequence; the data sequence includes the vehicle working condition information of the target vehicle at multiple unit times;
基于氮氧化物浓度预测模型，获取序列片段对应的SCR系统下游的氮氧化物浓度预测序列；氮氧化物浓度预测模型基于正常油品训练集训练得到；Based on the nitrogen oxide concentration prediction model, obtain the nitrogen oxide concentration prediction sequence downstream of the SCR system corresponding to the sequence segment; the nitrogen oxide concentration prediction model is obtained based on the training of the normal oil product training set;
获取序列片段对应的SCR系统下游的氮氧化物浓度实际序列，根据氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的差异程度，确定序列片段的初始油品识别结果；Obtain the actual sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment, and determine the initial oil product identification result of the sequence fragment according to the degree of difference between the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration;
根据数据序列对应的加油动作，以及数据序列对应的SCR系统氨氮比的变化趋势，对序列片段的初始油品识别结果进行修正，得到序列片段的最终油品识别结果。According to the refueling action corresponding to the data sequence and the change trend of the ammonia-nitrogen ratio of the SCR system corresponding to the data sequence, the initial oil product identification result of the sequence segment is corrected to obtain the final oil product identification result of the sequence segment.
在其中一个实施例中，对数据序列进行分割，在数据序列中确定序列片段，包括：In one of the embodiments, the data sequence is segmented, and sequence fragments are determined in the data sequence, including:
根据网联数据，获取数据序列对应的油量信息，根据油量信息确定数据序列对应的加油动作；Obtain the fuel quantity information corresponding to the data sequence according to the network data, and determine the refueling action corresponding to the data sequence according to the fuel quantity information;
根据加油动作在数据序列中确定第一分割节点；determining the first split node in the data sequence according to the refueling action;
识别数据序列中的异常数据，根据异常数据在数据序列中确定第二分割节点；异常数据至少包括非油品异常数据和丢包掉帧数据；identifying abnormal data in the data sequence, and determining a second segmentation node in the data sequence according to the abnormal data; the abnormal data includes at least non-oil abnormal data and packet and frame loss data;
根据第一分割节点和第二分割节点对数据序列进行分割，获得数据序列中的序列片段；序列片段包括目标车辆在连续多个单位时间的整车工况信息。The data sequence is segmented according to the first segmentation node and the second segmentation node to obtain sequence fragments in the data sequence; the sequence fragments include the vehicle operating condition information of the target vehicle in a plurality of continuous unit times.
在其中一个实施例中，正常油品训练集的获取方式，包括：In one of the embodiments, the way to obtain the normal oil product training set includes:
获取目标车辆在正常油品状态下的样本数据序列和SCR系统下游的氮氧化物浓度样本序列；Obtain the sample data sequence of the target vehicle under normal oil condition and the sample sequence of nitrogen oxide concentration downstream of the SCR system;
基于样本数据序列和SCR系统下游的氮氧化物浓度样本序列获取训练实例，基于多个训练实例构建正常油品训练集。A training instance is obtained based on the sample data sequence and the nitrogen oxide concentration sample sequence downstream of the SCR system, and a normal oil product training set is constructed based on multiple training instances.
在其中一个实施例中，根据氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的差异程度，确定序列片段的初始油品识别结果，包括：In one of the embodiments, according to the degree of difference between the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration, the initial oil product identification result of the sequence fragment is determined, including:
获取氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的残差序列；残差序列包括多个残差值，各残差值用于表征氮氧化物浓度预测序列和氮氧化物浓度实际序列在单位时间的数据差值；Obtain the residual sequence between the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration; the residual sequence includes multiple residual values, and each residual value is used to represent the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration The data difference of the sequence in unit time;
在残差序列中，获取残差值大于残差阈值的异常点，并确定异常点数量；In the residual sequence, obtain the abnormal points whose residual value is greater than the residual threshold, and determine the number of abnormal points;
根据异常点数量和残差序列的长度，获取序列片段的异常分数；According to the number of abnormal points and the length of the residual sequence, the abnormal score of the sequence fragment is obtained;
在异常分数大于异常阈值的情况下，确定序列片段的初始油品识别结果为高硫油品；In the case where the abnormal score is greater than the abnormal threshold, the initial oil product identification result of the sequence fragment is determined to be a high-sulfur oil product;
在异常分数不大于异常阈值的情况下，确定序列片段的初始油品识别结果为正常油品。In the case that the abnormal score is not greater than the abnormal threshold, it is determined that the initial oil product recognition result of the sequence fragment is a normal oil product.
在其中一个实施例中，根据数据序列对应的加油动作，以及数据序列对应的SCR系统氨氮比的变化趋势，对序列片段的初始油品识别结果进行修正，得到序列片段的最终油品识别结果，包括：In one of the embodiments, according to the refueling action corresponding to the data sequence and the change trend of the ammonia-nitrogen ratio of the SCR system corresponding to the data sequence, the initial oil product identification result of the sequence segment is corrected to obtain the final oil product identification result of the sequence segment, include:
根据加油动作，确定序列片段的第一修正结果；Determine the first correction result of the sequence segment according to the refueling action;
根据SCR系统氨氮比的变化趋势，确定序列片段的第二修正结果；According to the change trend of the ammonia-nitrogen ratio of the SCR system, the second correction result of the sequence fragment is determined;
根据第一修正结果对序列片段的初始油品识别结果进行第一次修正，根据第二修正结果对第一次修正后的序列片段的初始油品识别结果进行第二次修正，得到序列片段的最终油品识别结果。According to the first correction result, the initial oil product identification result of the sequence fragment is corrected for the first time, and according to the second correction result, the initial oil product identification result of the sequence fragment corrected for the first time is corrected for the second time, and the sequence fragment's initial oil product identification result is obtained. The final oil identification result.
在其中一个实施例中，根据加油动作，确定序列片段的第一修正结果，包括：In one of the embodiments, according to the refueling action, determining the first correction result of the sequence segment includes:
根据网联数据，获取数据序列对应的油量信息，根据油量信息确定数据序列对应的加油动作；Obtain the fuel quantity information corresponding to the data sequence according to the network data, and determine the refueling action corresponding to the data sequence according to the fuel quantity information;
根据加油动作在数据序列中获取包含序列片段的片段集合；片段集合包括至少一个序列片段，各序列片段存在连续的先后顺序，且片段集合中不存在加油动作；Obtain a fragment set containing sequence fragments in the data sequence according to the refueling action; the fragment set includes at least one sequence fragment, each sequence fragment has a continuous sequence, and there is no refueling action in the fragment set;
根据片段集合中各序列片段的初始油品识别结果，以及各序列片段的片段时长，计算片段集合中的正常油品总时长和高硫油品总时长；According to the initial oil identification results of each sequence fragment in the fragment set, and the fragment duration of each sequence fragment, calculate the total duration of normal oil and the total duration of high-sulfur oil in the fragment collection;
根据片段集合的片段集合总时长、正常油品总时长和高硫油品总时长，计算高硫油品总时长在片段集合中的高硫油品时长占比；According to the total duration of the fragment collection, the total duration of normal oil products and the total duration of high-sulfur oil products in the fragment collection, calculate the proportion of the total duration of high-sulfur oil products in the duration of high-sulfur oil products in the fragment collection;
在高硫油品时长占比大于预设比例的片段集合总时长的情况下，确定片段集合的第一修正结果为高硫油品，将片段集合的第一修正结果作为片段集合中各序列片段的第一修正结果；In the case that the proportion of high-sulfur oil product duration is greater than the total duration of the fragment set with a preset ratio, determine that the first correction result of the fragment set is high-sulfur oil, and use the first correction result of the fragment set as each sequence fragment in the fragment set The result of the first revision of ;
在高硫油品时长占比不大于预设比例的片段集合总时长的情况下，确定片段集合的第一修正结果为正常油品，将片段集合的第一修正结果作为片段集合中各序列片段的第一修正结果。In the case that the proportion of high-sulfur oil product duration is not greater than the total duration of the fragment set in the preset proportion, the first correction result of the fragment set is determined to be normal oil, and the first correction result of the fragment set is used as each sequence fragment in the fragment set The first revised result of .
在其中一个实施例中，根据SCR系统氨氮比的变化趋势，确定序列片段的第二修正结果，包括：In one of the embodiments, according to the change trend of the ammonia-nitrogen ratio of the SCR system, the second correction result of the sequence fragment is determined, including:
根据网联数据，获取片段集合对应的SCR反应温度；Obtain the SCR reaction temperature corresponding to the fragment set according to the network data;
在片段集合中，获取SCR反应温度大于温度阈值的部分，作为片段集合的目标子集；In the fragment set, obtain the part whose SCR reaction temperature is greater than the temperature threshold, as a target subset of the fragment set;
计算目标子集的平均氨氮比，并对平均氨氮比进行向前差分，计算得到氨氮比向前差分；Calculate the average ammonia-nitrogen ratio of the target subset, and perform a forward difference on the average ammonia-nitrogen ratio, and calculate the forward difference of the ammonia-nitrogen ratio;
在氨氮比向前差分小于预设容忍度的情况下，确定片段集合的第二修正结果为高硫油品，将片段集合的第二修正结果作为片段集合中各序列片段的第二修正结果；When the forward difference of the ammonia nitrogen ratio is less than the preset tolerance, determine that the second correction result of the fragment set is a high-sulfur oil product, and use the second correction result of the fragment set as the second correction result of each sequence fragment in the fragment set;
在氨氮比向前差分不小于预设容忍度的情况下，确定片段集合的第二修正结果为正常油品，将片段集合的第二修正结果作为片段集合中各序列片段的第二修正结果。When the forward difference of the ammonia nitrogen ratio is not less than the preset tolerance, it is determined that the second correction result of the fragment set is normal oil, and the second correction result of the fragment set is used as the second correction result of each sequence fragment in the fragment set.
第二方面，本申请还提供了一种油品识别装置。所述装置包括：In the second aspect, the present application also provides an oil identification device. The devices include:
获取模块，用于根据目标车辆的网联数据获取目标车辆的数据序列，对数据序列进行分割，在数据序列中确定序列片段；数据序列包括目标车辆在多个单位时间的整车工况信息；The acquisition module is used to acquire the data sequence of the target vehicle according to the network connection data of the target vehicle, segment the data sequence, and determine the sequence fragments in the data sequence; the data sequence includes the vehicle working condition information of the target vehicle at multiple unit times;
预测模块，用于基于氮氧化物浓度预测模型，获取序列片段对应的SCR系统下游的氮氧化物浓度预测序列；氮氧化物浓度预测模型基于正常油品训练集训练得到；The prediction module is used to obtain the nitrogen oxide concentration prediction sequence downstream of the SCR system corresponding to the sequence segment based on the nitrogen oxide concentration prediction model; the nitrogen oxide concentration prediction model is obtained by training based on the normal oil product training set;
比对模块，用于获取序列片段对应的SCR系统下游的氮氧化物浓度实际序列，根据氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的差异程度，确定序列片段的初始油品识别结果；The comparison module is used to obtain the actual sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment, and determine the initial oil product identification of the sequence fragment according to the degree of difference between the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration result;
识别模块，用于根据数据序列对应的加油动作，以及数据序列对应的SCR系统氨氮比的变化趋势，对序列片段的初始油品识别结果进行修正，得到序列片段的最终油品识别结果。The identification module is used to correct the initial oil product identification result of the sequence segment according to the refueling action corresponding to the data sequence and the change trend of the ammonia nitrogen ratio of the SCR system corresponding to the data sequence, and obtain the final oil product identification result of the sequence segment.
第三方面，本申请还提供了一种计算机设备。所述计算机设备包括存储器和处理器，所述存储器存储有计算机程序，所述处理器执行所述计算机程序时实现以下步骤：In a third aspect, the present application also provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:
根据目标车辆的网联数据获取目标车辆的数据序列，对数据序列进行分割，在数据序列中确定序列片段；数据序列包括目标车辆在多个单位时间的整车工况信息；Obtain the data sequence of the target vehicle according to the network data of the target vehicle, segment the data sequence, and determine the sequence fragments in the data sequence; the data sequence includes the vehicle working condition information of the target vehicle at multiple unit times;
基于氮氧化物浓度预测模型，获取序列片段对应的SCR系统下游的氮氧化物浓度预测序列；氮氧化物浓度预测模型基于正常油品训练集训练得到；Based on the nitrogen oxide concentration prediction model, obtain the nitrogen oxide concentration prediction sequence downstream of the SCR system corresponding to the sequence segment; the nitrogen oxide concentration prediction model is obtained based on the training of the normal oil product training set;
获取序列片段对应的SCR系统下游的氮氧化物浓度实际序列，根据氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的差异程度，确定序列片段的初始油品识别结果；Obtain the actual sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment, and determine the initial oil product identification result of the sequence fragment according to the degree of difference between the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration;
根据数据序列对应的加油动作，以及数据序列对应的SCR系统氨氮比的变化趋势，对序列片段的初始油品识别结果进行修正，得到序列片段的最终油品识别结果。According to the refueling action corresponding to the data sequence and the change trend of the ammonia-nitrogen ratio of the SCR system corresponding to the data sequence, the initial oil product identification result of the sequence segment is corrected to obtain the final oil product identification result of the sequence segment.
第四方面，本申请还提供了一种计算机可读存储介质。所述计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时实现以下步骤：In a fourth aspect, the present application also provides a computer-readable storage medium. The computer-readable storage medium has a computer program stored thereon, and when the computer program is executed by a processor, the following steps are implemented:
根据目标车辆的网联数据获取目标车辆的数据序列，对数据序列进行分割，在数据序列中确定序列片段；数据序列包括目标车辆在多个单位时间的整车工况信息；Obtain the data sequence of the target vehicle according to the network data of the target vehicle, segment the data sequence, and determine the sequence fragments in the data sequence; the data sequence includes the vehicle working condition information of the target vehicle at multiple unit times;
基于氮氧化物浓度预测模型，获取序列片段对应的SCR系统下游的氮氧化物浓度预测序列；氮氧化物浓度预测模型基于正常油品训练集训练得到；Based on the nitrogen oxide concentration prediction model, obtain the nitrogen oxide concentration prediction sequence downstream of the SCR system corresponding to the sequence segment; the nitrogen oxide concentration prediction model is obtained based on the training of the normal oil product training set;
获取序列片段对应的SCR系统下游的氮氧化物浓度实际序列，根据氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的差异程度，确定序列片段的初始油品识别结果；Obtain the actual sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment, and determine the initial oil product identification result of the sequence fragment according to the degree of difference between the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration;
根据数据序列对应的加油动作，以及数据序列对应的SCR系统氨氮比的变化趋势，对序列片段的初始油品识别结果进行修正，得到序列片段的最终油品识别结果。According to the refueling action corresponding to the data sequence and the change trend of the ammonia-nitrogen ratio of the SCR system corresponding to the data sequence, the initial oil product identification result of the sequence segment is corrected to obtain the final oil product identification result of the sequence segment.
第五方面，本申请还提供了一种计算机程序产品。所述计算机程序产品，包括计算机程序，该计算机程序被处理器执行时实现以下步骤：In a fifth aspect, the present application also provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:
根据目标车辆的网联数据获取目标车辆的数据序列，对数据序列进行分割，在数据序列中确定序列片段；数据序列包括目标车辆在多个单位时间的整车工况信息；Obtain the data sequence of the target vehicle according to the network data of the target vehicle, segment the data sequence, and determine the sequence fragments in the data sequence; the data sequence includes the vehicle working condition information of the target vehicle at multiple unit times;
基于氮氧化物浓度预测模型，获取序列片段对应的SCR系统下游的氮氧化物浓度预测序列；氮氧化物浓度预测模型基于正常油品训练集训练得到；Based on the nitrogen oxide concentration prediction model, obtain the nitrogen oxide concentration prediction sequence downstream of the SCR system corresponding to the sequence segment; the nitrogen oxide concentration prediction model is obtained based on the training of the normal oil product training set;
获取序列片段对应的SCR系统下游的氮氧化物浓度实际序列，根据氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的差异程度，确定序列片段的初始油品识别结果；Obtain the actual sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment, and determine the initial oil product identification result of the sequence fragment according to the degree of difference between the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration;
根据数据序列对应的加油动作，以及数据序列对应的SCR系统氨氮比的变化趋势，对序列片段的初始油品识别结果进行修正，得到序列片段的最终油品识别结果。According to the refueling action corresponding to the data sequence and the change trend of the ammonia-nitrogen ratio of the SCR system corresponding to the data sequence, the initial oil product identification result of the sequence segment is corrected to obtain the final oil product identification result of the sequence segment.
上述油品识别方法、装置、计算机设备、存储介质和计算机程序产品，根据目标车辆的网联数据获取目标车辆的数据序列，对数据序列进行分割，在数据序列中确定序列片段；基于氮氧化物浓度预测模型，获取序列片段对应的SCR系统下游的氮氧化物浓度预测序列；获取序列片段对应的SCR系统下游的氮氧化物浓度实际序列，根据氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的差异程度，确定序列片段的初始油品识别结果；根据数据序列对应的加油动作，以及数据序列对应的SCR系统氨氮比的变化趋势，对序列片段的初始油品识别结果进行修正，得到序列片段的最终油品识别结果。在油品识别过程中，基于深度学习理论更为深入地挖掘SCR内部的反应特性，解决了正常柴油和高硫柴油数据样本不对等的问题，识别精度更高，并且考虑了网联数据的低分辨率特性，有效处理了网联数据的丢包、掉帧等问题，适用于各种复杂工况，鲁棒性更佳。The above oil identification method, device, computer equipment, storage medium and computer program product obtain the data sequence of the target vehicle according to the networked data of the target vehicle, segment the data sequence, and determine sequence fragments in the data sequence; Concentration prediction model, obtain the predicted sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment; obtain the actual sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment, based on the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration Determine the initial oil product identification result of the sequence segment; according to the refueling action corresponding to the data sequence and the change trend of the ammonia-nitrogen ratio of the SCR system corresponding to the data sequence, the initial oil product identification result of the sequence segment is corrected to obtain The final oil product identification results of the sequence fragments. In the process of oil product identification, based on the deep learning theory, the internal reaction characteristics of the SCR are further excavated, and the problem of unequal data samples between normal diesel and high-sulfur diesel is solved, the identification accuracy is higher, and the low network data is considered. The resolution feature effectively solves the problems of packet loss and frame drop of networked data, and is suitable for various complex working conditions with better robustness.
附图说明Description of drawings
图1为一个实施例中油品识别方法的应用环境图；Fig. 1 is an application environment diagram of the oil product identification method in an embodiment;
图2为一个实施例中油品识别方法的流程示意图；Fig. 2 is a schematic flow chart of an oil product identification method in an embodiment;
图3为一个实施例中对数据序列进行分割得到序列片段的流程示意图；FIG. 3 is a schematic flow diagram of segmenting a data sequence to obtain sequence fragments in one embodiment;
图4为一个实施例中SCR下游NOx预测模型示意图；Fig. 4 is a schematic diagram of the SCR downstream NOx prediction model in one embodiment;
图5为一个实施例中SCR下游NOx预测模型进行训练的流程示意图；FIG. 5 is a schematic flow diagram of training the SCR downstream NOx prediction model in an embodiment;
图6为一个实施例中确定序列片段的初始油品识别结果的流程示意图；Fig. 6 is a schematic flow chart of determining the initial oil product identification results of sequence fragments in one embodiment;
图7为一个实施例中对序列片段初始油品识别结果进行修正的流程示意图；Fig. 7 is a schematic flow chart of correcting the initial oil product identification result of the sequence segment in one embodiment;
图8为一个实施例中高硫柴油识别系统及云平台的结构示意图；Fig. 8 is a structural schematic diagram of a high-sulfur diesel identification system and a cloud platform in an embodiment;
图9为一个实施例中油品识别装置的结构框图；Fig. 9 is a structural block diagram of an oil identification device in an embodiment;
图10为一个实施例中计算机设备的内部结构图。Figure 10 is a diagram of the internal structure of a computer device in one embodiment.
具体实施方式Detailed ways
为了使本申请的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本申请进行进一步详细说明。应当理解，此处描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。In order to make the purpose, technical solution and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present application, and are not intended to limit the present application.
本申请实施例提供的油品识别方法，可以应用于如图1所示的应用环境中。其中，终端102通过网络与车联网云服务器104进行通信。数据存储系统可以存储车联网云服务器104上车辆的网联数据。数据存储系统可以集成在车联网云服务器104上，也可以放在云上或其他网络服务器上。其中，终端102可以但不限于是各种个人计算机、笔记本电脑、智能手机、平板电脑、物联网设备和便携式可穿戴设备，物联网设备可为智能车载设备等。便携式可穿戴设备可为智能手表、智能手环、头戴设备等。服务器104可以用独立的服务器或者是多个服务器组成的服务器集群来实现。The oil product identification method provided in the embodiment of the present application can be applied to the application environment shown in FIG. 1 . Wherein, the terminal 102 communicates with the Internet of Vehicles cloud server 104 through the network. The data storage system can store the Internet connection data of the vehicle on the Internet of Vehicles cloud server 104 . The data storage system can be integrated on the Internet of Vehicles cloud server 104, and can also be placed on the cloud or other network servers. Wherein, the terminal 102 may be, but not limited to, various personal computers, notebook computers, smart phones, tablet computers, Internet of Things devices and portable wearable devices, and the Internet of Things devices may be smart vehicle-mounted devices and the like. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, and the like. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers.
在一个实施例中，如图2所示，提供了一种油品识别方法，以该方法应用于图1中的终端102为例进行说明，包括以下步骤：In one embodiment, as shown in FIG. 2, a method for identifying oil products is provided. The method is applied to the terminal 102 in FIG. 1 as an example for illustration, including the following steps:
步骤202，根据目标车辆的网联数据获取目标车辆的数据序列，对数据序列进行分割，在数据序列中确定序列片段；数据序列包括目标车辆在多个单位时间的整车工况信息。Step 202: Obtain the data sequence of the target vehicle according to the networked data of the target vehicle, segment the data sequence, and determine the sequence fragments in the data sequence; the data sequence includes the vehicle operating condition information of the target vehicle at multiple unit times.
可选的，终端从服务器下载目标车辆的网联数据，基于需要进行油品识别的时间区间，根据网联数据中的发动机转速等整车工况信息提取车辆连续运行的数据序列，保证每个数据序列的运行连续性。然后根据燃油箱液位等油量信息识别车辆加油动作对数据序列进行第一次分割，保证每个片段的油品连续性。检查时间索引是否连续，进行第二次分割从而排除数据序列中丢包、掉帧等异常数据，保证每个片段的时间连续性。确定上下游NOx浓度等后处理系统工况信息的传感器异常工作区间，进行第二次分割从而排除数据序列中非油品异常的干扰。经过三次分割得到多个序列片段。Optionally, the terminal downloads the networked data of the target vehicle from the server, and based on the time interval for oil product identification, extracts the data sequence of the continuous operation of the vehicle according to the engine speed and other vehicle operating condition information in the networked data to ensure that each Running continuity of data series. Then, according to the oil quantity information such as the fuel tank level, the vehicle refueling action is recognized and the data sequence is segmented for the first time to ensure the continuity of the oil in each segment. Check whether the time index is continuous, and perform a second segmentation to eliminate abnormal data such as packet loss and frame drop in the data sequence, ensuring the time continuity of each segment. Determine the sensor abnormal working range of the post-processing system working condition information such as the upstream and downstream NOx concentration, and perform a second segmentation to eliminate the interference of non-oil abnormalities in the data sequence. Multiple sequence fragments were obtained through three divisions.
步骤204，基于氮氧化物浓度预测模型，获取序列片段对应的SCR系统下游的氮氧化物浓度预测序列；氮氧化物浓度预测模型基于正常油品训练集训练得到。Step 204, based on the NOx concentration prediction model, obtain the NOx concentration prediction sequence downstream of the SCR system corresponding to the sequence segment; the NOx concentration prediction model is trained based on the normal oil product training set.
其中，氮氧化物浓度通常表示为NOx浓度。正常油品训练集的获取方式，包括：获取目标车辆在正常油品状态下的样本数据序列和SCR系统下游的氮氧化物浓度样本序列；基于样本数据序列和SCR系统下游的氮氧化物浓度样本序列获取训练实例，基于多个训练实例构建正常油品训练集。Among them, the concentration of nitrogen oxides is generally expressed as the concentration of NOx. The method of obtaining the normal oil product training set includes: obtaining the sample data sequence of the target vehicle under normal oil state and the nitrogen oxide concentration sample sequence downstream of the SCR system; based on the sample data sequence and the nitrogen oxide concentration sample downstream of the SCR system Sequentially obtain training instances, and construct a normal oil product training set based on multiple training instances.
可选的，基于氮氧化物浓度预测模型对各序列片段的SCR系统下游的氮氧化物浓度进行预测，得到各序列片段对应的SCR系统下游的氮氧化物浓度预测序列。由于各序列片段中的数据是连续的，即连续单位时间的数据，因此各序列片段对应的SCR系统下游的氮氧化物浓度预测序列也是连续单位时间的数据。Optionally, the nitrogen oxide concentration downstream of the SCR system of each sequence fragment is predicted based on the nitrogen oxide concentration prediction model, and the nitrogen oxide concentration prediction sequence of the downstream of the SCR system corresponding to each sequence fragment is obtained. Since the data in each sequence fragment is continuous, that is, the data of continuous unit time, the nitrogen oxide concentration prediction sequence downstream of the SCR system corresponding to each sequence fragment is also continuous unit time data.
步骤206，获取序列片段对应的SCR系统下游的氮氧化物浓度实际序列，根据氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的差异程度，确定序列片段的初始油品识别结果。Step 206: Acquire the actual sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment, and determine the initial oil product identification result of the sequence fragment according to the degree of difference between the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration.
可选的，从网联数据中，确定各序列片段对应的SCR系统下游的氮氧化物浓度实际序列，氮氧化物浓度实际序列包括连续单位时间的数据。针对任意一个序列片段，将其对应的SCR系统下游的氮氧化物浓度实际序列和氮氧化物浓度预测序列作差，得到该序列片段对应的一个残差序列，残差序列能够表征氮氧化物浓度实际序列和氮氧化物浓度预测序列在各单位时间的差值。同理可以得到各序列片段对应的残差序列。根据各序列片段对应的残差序列中表现出的氮氧化物浓度实际序列和氮氧化物浓度预测序列的差异程度，确定各序列片段的初始油品识别结果。Optionally, from the networked data, the actual sequence of nitrogen oxide concentration downstream of the SCR system corresponding to each sequence fragment is determined, and the actual sequence of nitrogen oxide concentration includes data of continuous unit time. For any sequence segment, the actual sequence of nitrogen oxide concentration downstream of the corresponding SCR system is compared with the predicted sequence of nitrogen oxide concentration to obtain a residual sequence corresponding to the sequence segment, which can represent the concentration of nitrogen oxides The difference between the actual sequence and the predicted sequence of nitrogen oxide concentration at each unit time. Similarly, the residual sequence corresponding to each sequence segment can be obtained. According to the degree of difference between the actual sequence of nitrogen oxide concentration and the predicted sequence of nitrogen oxide concentration shown in the residual sequence corresponding to each sequence fragment, the initial oil product identification result of each sequence fragment is determined.
步骤208，根据数据序列对应的加油动作，以及数据序列对应的SCR系统氨氮比的变化趋势，对序列片段的初始油品识别结果进行修正，得到序列片段的最终油品识别结果。Step 208: Correct the initial oil product identification result of the sequence segment according to the refueling action corresponding to the data sequence and the change trend of the ammonia-nitrogen ratio of the SCR system corresponding to the data sequence, to obtain the final oil product identification result of the sequence segment.
可选的，先根据加油动作，确定各序列片段的第一修正结果，然后根据SCR系统氨氮比的变化趋势，确定各序列片段的第二修正结果。根据第一修正结果对各序列片段的初始油品识别结果进行第一次修正，根据第二修正结果对第一次修正后的各序列片段的初始油品识别结果进行第二次修正，得到各序列片段的最终油品识别结果。Optionally, the first correction result of each sequence segment is determined according to the refueling action, and then the second correction result of each sequence segment is determined according to the change trend of the ammonia-nitrogen ratio of the SCR system. According to the first correction result, the initial oil product identification results of each sequence segment were corrected for the first time, and the initial oil product identification results of each sequence segment after the first correction were corrected for the second time according to the second correction result, and each The final oil product identification results of the sequence fragments.
上述油品识别方法中，根据目标车辆的网联数据获取目标车辆的数据序列，对数据序列进行分割，在数据序列中确定序列片段；基于氮氧化物浓度预测模型，获取序列片段对应的SCR系统下游的氮氧化物浓度预测序列；获取序列片段对应的SCR系统下游的氮氧化物浓度实际序列，根据氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的差异程度，确定序列片段的初始油品识别结果；根据数据序列对应的加油动作，以及数据序列对应的SCR系统氨氮比的变化趋势，对序列片段的初始油品识别结果进行修正，得到序列片段的最终油品识别结果。在油品识别过程中，基于深度学习理论更为深入地挖掘SCR内部的反应特性，解决了正常柴油和高硫柴油数据样本不对等的问题，识别精度更高，并且考虑了网联数据的低分辨率特性，有效处理了网联数据的丢包、掉帧等问题，适用于各种复杂工况，鲁棒性更佳。In the above oil product identification method, the data sequence of the target vehicle is obtained according to the network data of the target vehicle, the data sequence is segmented, and the sequence fragment is determined in the data sequence; based on the nitrogen oxide concentration prediction model, the SCR system corresponding to the sequence fragment is obtained Downstream nitrogen oxide concentration prediction sequence; obtain the actual sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment, and determine the initial sequence fragment according to the degree of difference between the nitrogen oxide concentration prediction sequence and the actual sequence of nitrogen oxide concentration Oil product identification results: According to the refueling action corresponding to the data sequence and the change trend of the ammonia-nitrogen ratio of the SCR system corresponding to the data sequence, the initial oil product identification result of the sequence segment is corrected to obtain the final oil product identification result of the sequence segment. In the process of oil product identification, based on the deep learning theory, the internal reaction characteristics of the SCR are further excavated, and the problem of unequal data samples between normal diesel and high-sulfur diesel is solved, the identification accuracy is higher, and the low network data is considered. The resolution feature effectively solves the problems of packet loss and frame drop of networked data, and is suitable for various complex working conditions with better robustness.
在一个实施例中，对数据序列进行分割，在数据序列中确定序列片段，包括：根据网联数据，获取数据序列对应的油量信息，根据油量信息确定数据序列对应的加油动作；根据加油动作在数据序列中确定第一分割节点；识别数据序列中的异常数据，根据异常数据在数据序列中确定第二分割节点；异常数据至少包括非油品异常数据和丢包掉帧数据；根据第一分割节点和第二分割节点对数据序列进行分割，获得数据序列中的序列片段；序列片段包括目标车辆在连续多个单位时间的整车工况信息。In one embodiment, the data sequence is segmented, and the sequence fragments are determined in the data sequence, including: according to the network data, obtaining the fuel quantity information corresponding to the data sequence, and determining the refueling action corresponding to the data sequence according to the fuel quantity information; The action determines the first segmentation node in the data sequence; identifies the abnormal data in the data sequence, and determines the second segmentation node in the data sequence according to the abnormal data; the abnormal data includes at least non-oil abnormal data and packet loss and frame data; according to the first The first segmentation node and the second segmentation node segment the data sequence to obtain sequence fragments in the data sequence; the sequence fragments include the vehicle operating condition information of the target vehicle in multiple continuous unit times.
可选的，依据发动机转速等整车工况信息提取车辆连续运行的数据序列，保证每个数据序列的运行连续性。整车工况信息包括但不限于发动机转速、ECU车速、制动开关状态等能够反映车辆是否处在运行状态的数据通道。对于一个连续的待识别数据片段，去除了整车工况信息数据通道未工作时间索引的并集，保证获得的每个数据序列无工况间断或缺失。Optionally, the continuous operation data sequence of the vehicle is extracted according to the vehicle operating condition information such as the engine speed, so as to ensure the operation continuity of each data sequence. Vehicle working condition information includes but is not limited to engine speed, ECU speed, brake switch status and other data channels that can reflect whether the vehicle is in a running state. For a continuous data segment to be identified, the union of the non-working time index of the vehicle working condition information data channel is removed, so as to ensure that each data sequence obtained has no working condition discontinuity or absence.
依据燃油箱液位等油量信息识别车辆加油动作并作为分割节点，对数据序列进行分割，保证分割出的每个序列片段的油品连续性。油量信息包括但不限于燃油箱液位、油箱剩余油量等数据通道。确定油箱液位骤升的时刻作为节点对数据序列分割，保证分割出的每个序列片段都是同一种油品。According to the oil quantity information such as the fuel tank level, the vehicle refueling action is identified and used as a segmentation node to segment the data sequence to ensure the continuity of the oil in each segment of the segment. Fuel quantity information includes, but is not limited to, data channels such as the liquid level of the fuel tank and the remaining fuel quantity of the fuel tank. Determine the moment when the liquid level of the fuel tank suddenly rises as a node to segment the data sequence to ensure that each segment of the sequence is the same oil product.
检查时间索引是否连续以排除丢包、掉帧等异常，保证每个序列片段的时间连续性。顺序遍历时间索引，进一步检查片段的连续性，对于因硬件结构更换等因素导致的采样时间不一致的问题要通过降采样等方式统一采样时间。Check whether the time index is continuous to eliminate abnormalities such as packet loss and frame drop, and ensure the time continuity of each sequence segment. Sequentially traverse the time index to further check the continuity of the fragments. For the problem of inconsistent sampling time caused by factors such as hardware structure replacement, the sampling time must be unified through down-sampling and other methods.
确定上下游NOx浓度等后处理系统工况信息的传感器异常工作区间，排除非油品异常的干扰。后处理系统工况信息包括但不限于上下游NOx浓度、排气流量等数据通道。使用高硫柴油的最显著的表征是SCR是否发生硫中毒，SCR状态信息借由NOx传感器、温度传感器等采集，传感器未工作、初始化置位或故障时采集数据欠准，容易引起方法的误识别或系统的误动作，这些情况可称为“非油品异常”。锁定每个序列片段的非油品异常索引，并通过分割的方式去除该部分，减少对识别结果的影响。Determine the abnormal working range of sensors for after-treatment system working condition information such as upstream and downstream NOx concentrations, and eliminate the interference of non-oil abnormalities. The working condition information of the post-treatment system includes, but is not limited to, data channels such as upstream and downstream NOx concentrations and exhaust flow. The most obvious sign of using high-sulfur diesel is whether SCR has sulfur poisoning. SCR status information is collected by NOx sensors, temperature sensors, etc., and the collected data is inaccurate when the sensor is not working, initialized or faulty, which may easily lead to misidentification of the method Or malfunction of the system, these situations can be called "non-oil abnormality". Lock the non-oil abnormal index of each sequence fragment, and remove this part by segmentation to reduce the impact on the recognition result.
如图3所示，以上即为对数据序列进行分割得到序列片段的工作流程。在一种可行的实施方式中，也可根据数据的采集质量、应用需求去调整上述步骤或者上述步骤的执行顺序，得到最匹配、最高效的数据序列分割方案。As shown in FIG. 3 , the above is the workflow of segmenting the data sequence to obtain sequence fragments. In a feasible implementation manner, the above steps or the execution order of the above steps may also be adjusted according to the data collection quality and application requirements, so as to obtain the most matching and efficient data sequence segmentation scheme.
本实施例中，根据目标车辆的网联数据获取目标车辆的数据序列，对数据序列进行分割，在数据序列中确定序列片段。考虑了网联数据的低分辨率特性，有效处理了网联数据的丢包、掉帧等问题，适用于各种复杂工况，鲁棒性更佳。In this embodiment, the data sequence of the target vehicle is obtained according to the networked data of the target vehicle, the data sequence is divided, and the sequence fragments are determined in the data sequence. Considering the low-resolution characteristics of network data, it effectively handles the problems of packet loss and frame drop of network data, and is suitable for various complex working conditions with better robustness.
在一个实施例中，基于氮氧化物浓度预测模型，获取序列片段对应的SCR系统下游的氮氧化物浓度预测序列；氮氧化物浓度预测模型基于正常油品训练集训练得到；获取序列片段对应的SCR系统下游的氮氧化物浓度实际序列，根据氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的差异程度，确定序列片段的初始油品识别结果，包括：获取目标车辆在正常油品状态下的样本数据序列和SCR系统下游的氮氧化物浓度样本序列；基于样本数据序列和SCR系统下游的氮氧化物浓度样本序列获取训练实例，基于多个训练实例构建正常油品训练集。基于正常油品训练集训练得到氮氧化物浓度预测模型。基于氮氧化物浓度预测模型，获取序列片段对应的SCR系统下游的氮氧化物浓度预测序列；根据网联数据，获取序列片段对应的SCR系统下游的氮氧化物浓度实际序列。获取氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的残差序列；残差序列包括多个残差值，各残差值用于表征氮氧化物浓度预测序列和氮氧化物浓度实际序列在单位时间的数据差值；在残差序列中，获取残差值大于残差阈值的异常点，并确定异常点数量；根据异常点数量和残差序列的长度，获取序列片段的异常分数；在异常分数大于异常阈值的情况下，确定序列片段的初始油品识别结果为高硫油品；在异常分数不大于异常阈值的情况下，确定序列片段的初始油品识别结果为正常油品。In one embodiment, based on the nitrogen oxide concentration prediction model, the nitrogen oxide concentration prediction sequence downstream of the SCR system corresponding to the sequence fragment is obtained; the nitrogen oxide concentration prediction model is trained based on a normal oil product training set; the sequence fragment corresponding to The actual sequence of nitrogen oxide concentration downstream of the SCR system, according to the degree of difference between the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration, determines the initial oil product identification result of the sequence segment, including: obtaining the normal oil product of the target vehicle The sample data sequence under the state and the nitrogen oxide concentration sample sequence downstream of the SCR system; training examples are obtained based on the sample data sequence and the nitrogen oxide concentration sample sequence downstream of the SCR system, and a normal oil training set is constructed based on multiple training examples. The NOx concentration prediction model was trained based on the normal oil training set. Based on the nitrogen oxide concentration prediction model, obtain the predicted sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment; obtain the actual sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment according to the network data. Obtain the residual sequence between the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration; the residual sequence includes multiple residual values, and each residual value is used to represent the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration The data difference value of the sequence per unit time; in the residual sequence, obtain the abnormal points whose residual value is greater than the residual threshold, and determine the number of abnormal points; according to the number of abnormal points and the length of the residual sequence, obtain the abnormal score of the sequence fragment ; When the abnormal score is greater than the abnormal threshold, determine the initial oil identification result of the sequence segment as high-sulfur oil; when the abnormal score is not greater than the abnormal threshold, determine the initial oil identification result of the sequence segment as normal oil .
可选的，基于正常油品状态下的相关数据获得正常油品状态下的SCR下游NOx浓度预测模型。相较于车辆的正常行驶数据来说，车辆高硫柴油行驶数据属于小样本。样本不对等给油品识别造成了困扰，SCR下游NOx预测模型能够反映正常油品状态的车辆数据，对于未知数据而言，其下游NOx浓度的实际测量值和SCR下游NOx预测模型的预测值的残差反映了其异常程度，依据残差大小即可评估待识别数据所使用的燃油品质。Optionally, a NOx concentration prediction model downstream of the SCR under normal oil state is obtained based on relevant data under normal oil state. Compared with the normal driving data of the vehicle, the vehicle high-sulfur diesel driving data is a small sample. The inequalities of samples have caused troubles for oil product identification. The NOx prediction model downstream of SCR can reflect the vehicle data of normal oil product status. For unknown data, the difference between the actual measured value of downstream NOx concentration and the predicted value of the SCR downstream NOx prediction model The residual reflects the degree of abnormality, and the quality of the fuel used in the data to be identified can be evaluated according to the size of the residual.
SCR下游NOx预测模型的准确程度决定着残差网络判别模块的识别精度。示例性地，如图4所示，给出一个基于深度学习设计SCR下游NOx浓度预测模型的具体实施例。在该实施例中,预测模型的输入选择SCR上游NOx浓度、排气流量、发动机排气温度和SCR尿素喷射量，选取两层长短时记忆网络层(LSTM层)和一层全连接层去学习输入输出关系，最后经由RELU层限制输出为正。LSTM是一种专长于处理时间序列的网络结构，基于该网络设计的预测模型能够捕捉更深层次的SCR内部反应动态，预测准确度更高。The accuracy of the SCR downstream NOx prediction model determines the recognition accuracy of the residual network discriminant module. Exemplarily, as shown in FIG. 4 , a specific embodiment of designing a NOx concentration prediction model downstream of the SCR based on deep learning is given. In this embodiment, the input of the prediction model is the NOx concentration upstream of the SCR, the exhaust gas flow, the engine exhaust temperature and the SCR urea injection amount, and two layers of long-short-term memory network layers (LSTM layers) and one layer of fully connected layers are selected to learn The input-output relationship, and finally restrict the output to be positive through the RELU layer. LSTM is a network structure specialized in processing time series. The prediction model designed based on this network can capture the deeper internal reaction dynamics of SCR, and the prediction accuracy is higher.
如图5所示，基于深度学习的SCR下游NOx预测模型的离线训练流程。首先收集覆盖需求型号车辆具备健康的SCR的正常柴油行驶数据集，并划分为训练集与测试集；接着构建基于如图4所示网络结构，利用训练集不断调试训练参数和网络超参数，不断迭代训练网络；然后评估模型在已知数据的拟合程度，利用测试集评估模型在未知数据的泛化能力，若精度理想则获得最终的SCR下游NOx浓度预测模型，否则继续进行上两步骤，直至模型的预测精度符合需求。As shown in Figure 5, the offline training process of the SCR downstream NOx prediction model based on deep learning. Firstly, collect the normal diesel driving data set with healthy SCR of vehicles covering the required model, and divide it into training set and test set; then construct the network structure based on the network structure shown in Figure 4, use the training set to continuously adjust the training parameters and network hyperparameters, and continuously Iteratively train the network; then evaluate the fitting degree of the model on known data, use the test set to evaluate the generalization ability of the model on unknown data, if the accuracy is satisfactory, obtain the final NOx concentration prediction model downstream of the SCR, otherwise continue to the previous two steps, Until the prediction accuracy of the model meets the requirements.
上述SCR预测模型是本发明提供的示例性实施例而并非专一性限制，显而易见地，本领域的技术人员能够通过修改模型输入输出、网络架构去获得其它形式的模型。在一些可行的实施方式中，SCR模型也可通过标定、多项式拟合、机理推导等方式获得。The above SCR prediction model is an exemplary embodiment provided by the present invention rather than a specific limitation. Obviously, those skilled in the art can obtain other forms of models by modifying model input and output and network architecture. In some feasible implementation manners, the SCR model can also be obtained by means of calibration, polynomial fitting, mechanism derivation, and the like.
SCR下游NOx浓度预测模型计算各个序列片段的下游NOx浓度预测值，并同实测值作差获得残差。通过SCR下游NOx浓度预测模型计算得到若干和原始序列片段等长的一维残差序列，每个残差序列即体现原始序列片段的异常水平。将一维残差序列中残差值高于阈值且不位于非油品异常区间内的点标记为有效异常点。计算各个一维残差序列内有效异常点个数和一维残差序列片段长度的比值，获得各个一维残差序列的异常分数，也就是对应的各序列片段的异常分数。判定异常分数高于阈值的序列片段为车辆使用高硫柴油的序列片段，否则为使用正常柴油的序列片段。The NOx concentration prediction model downstream of the SCR calculates the predicted value of the downstream NOx concentration of each sequence segment, and makes a difference with the measured value to obtain the residual. Several one-dimensional residual sequences with the same length as the original sequence fragments were calculated through the SCR downstream NOx concentration prediction model, and each residual sequence reflected the abnormal level of the original sequence fragments. In the one-dimensional residual series, the points whose residual value is higher than the threshold and not located in the non-oil abnormal interval are marked as effective abnormal points. Calculate the ratio of the number of effective outliers in each one-dimensional residual sequence to the length of the one-dimensional residual sequence segment, and obtain the abnormal score of each one-dimensional residual sequence, that is, the corresponding abnormal score of each sequence segment. The sequence segment whose abnormal score is higher than the threshold is determined to be the sequence segment of the vehicle using high-sulfur diesel, otherwise it is the sequence segment of using normal diesel.
如图6所示，以上即为确定各序列片段的初始油品识别结果的工作流程。针对每一个序列片段，能够给出一个初步的油品识别结果，技术人员可按需设计SCR下游NOx浓度预测模型，自由调节阈值、以获得更高的准确度。As shown in Figure 6, the above is the workflow for determining the initial oil product identification results of each sequence fragment. For each sequence fragment, a preliminary oil product identification result can be given, and technicians can design the NOx concentration prediction model downstream of the SCR as needed, and freely adjust the threshold to obtain higher accuracy.
本实施例中，基于氮氧化物浓度预测模型，获取序列片段对应的SCR系统下游的氮氧化物浓度预测序列；获取序列片段对应的SCR系统下游的氮氧化物浓度实际序列，根据氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的差异程度，确定序列片段的初始油品识别结果。基于深度学习理论更为深入地挖掘SCR内部的反应特性，解决了正常柴油和高硫柴油数据样本不对等的问题，识别精度更高。In this embodiment, based on the nitrogen oxide concentration prediction model, the nitrogen oxide concentration prediction sequence downstream of the SCR system corresponding to the sequence fragment is obtained; the actual sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment is obtained, and according to the nitrogen oxide concentration The degree of difference between the predicted sequence and the actual sequence of nitrogen oxide concentration determines the initial oil identification result of the sequence fragment. Based on the deep learning theory, the internal reaction characteristics of the SCR are more deeply excavated, and the problem of unequal data samples between normal diesel and high-sulfur diesel is solved, and the recognition accuracy is higher.
在一个实施例中，根据数据序列对应的加油动作，以及数据序列对应的SCR系统氨氮比的变化趋势，对序列片段的初始油品识别结果进行修正，得到序列片段的最终油品识别结果，包括：根据网联数据，获取数据序列对应的油量信息，根据油量信息确定数据序列对应的加油动作；根据加油动作在数据序列中获取包含序列片段的片段集合；片段集合包括至少一个序列片段，各序列片段存在连续的先后顺序，且片段集合中不存在加油动作；根据片段集合中各序列片段的初始油品识别结果，以及各序列片段的片段时长，计算片段集合中的正常油品总时长和高硫油品总时长；根据片段集合的片段集合总时长、正常油品总时长和高硫油品总时长，计算高硫油品总时长在片段集合中的高硫油品时长占比；在高硫油品时长占比大于预设比例的片段集合总时长的情况下，确定片段集合的第一修正结果为高硫油品，将片段集合的第一修正结果作为片段集合中各序列片段的第一修正结果；在高硫油品时长占比不大于预设比例的片段集合总时长的情况下，确定片段集合的第一修正结果为正常油品，将片段集合的第一修正结果作为片段集合中各序列片段的第一修正结果。根据网联数据，获取片段集合对应的SCR反应温度；在片段集合中，获取SCR反应温度大于温度阈值的部分，作为片段集合的目标子集；计算目标子集的平均氨氮比，并对平均氨氮比进行向前差分，计算得到氨氮比向前差分；在氨氮比向前差分小于预设容忍度的情况下，确定片段集合的第二修正结果为高硫油品，将片段集合的第二修正结果作为片段集合中各序列片段的第二修正结果；在氨氮比向前差分不小于预设容忍度的情况下，确定片段集合的第二修正结果为正常油品，将片段集合的第二修正结果作为片段集合中各序列片段的第二修正结果。根据第一修正结果对序列片段的初始油品识别结果进行第一次修正，根据第二修正结果对第一次修正后的序列片段的初始油品识别结果进行第二次修正，得到序列片段的最终油品识别结果。In one embodiment, according to the refueling action corresponding to the data sequence and the change trend of the ammonia-nitrogen ratio of the SCR system corresponding to the data sequence, the initial oil product identification result of the sequence segment is corrected to obtain the final oil product identification result of the sequence segment, including : Obtain the fuel quantity information corresponding to the data sequence according to the network data, determine the refueling action corresponding to the data sequence according to the fuel quantity information; obtain a fragment set containing sequence fragments in the data sequence according to the refueling action; the fragment set includes at least one sequence fragment, There is a continuous order of each sequence segment, and there is no refueling action in the segment set; according to the initial oil product recognition results of each sequence segment in the segment set, and the segment duration of each sequence segment, calculate the total duration of normal oil products in the segment set and the total duration of high-sulfur oil products; according to the total duration of fragment collection, the total duration of normal oil products and the total duration of high-sulfur oil products of the fragment collection, calculate the proportion of the total duration of high-sulfur oil products in the duration of high-sulfur oil products in the fragment collection; In the case that the proportion of high-sulfur oil product duration is greater than the total duration of the fragment set with a preset ratio, determine that the first correction result of the fragment set is high-sulfur oil, and use the first correction result of the fragment set as each sequence fragment in the fragment set The first correction result of the high-sulfur oil product; in the case that the proportion of the high-sulfur oil product duration is not greater than the total duration of the fragment set of the preset proportion, the first correction result of the fragment set is determined to be a normal oil product, and the first correction result of the fragment set is used as The first correction result of each sequence segment in the segment set. According to the network connection data, obtain the SCR reaction temperature corresponding to the fragment set; in the fragment set, obtain the part whose SCR reaction temperature is greater than the temperature threshold, as the target subset of the fragment set; calculate the average ammonia nitrogen ratio of the target subset, and compare the average ammonia nitrogen The forward difference of the ammonia-nitrogen ratio is calculated to obtain the forward difference of the ammonia-nitrogen ratio; when the forward difference of the ammonia-nitrogen ratio is less than the preset tolerance, it is determined that the second correction result of the fragment set is high-sulfur oil, and the second correction of the fragment set The result is used as the second correction result of each sequence fragment in the fragment set; when the forward difference of the ammonia-nitrogen ratio is not less than the preset tolerance, the second correction result of the fragment set is determined to be normal oil, and the second correction result of the fragment set is The result is used as the second correction result of each sequence fragment in the fragment set. According to the first correction result, the initial oil product identification result of the sequence fragment is corrected for the first time, and according to the second correction result, the initial oil product identification result of the sequence fragment corrected for the first time is corrected for the second time, and the sequence fragment's initial oil product identification result is obtained. The final oil identification result.
可选的，车辆网联大数据应用的一个重大挑战就是数据质量不稳定。在真实驾驶场景的各种未知的复杂工况下，初始油品识别结果可能会出现一定的误判断。因此借助一系列的规则去修正初始油品识别结果，使获得的最终的油品识别结果准确度更高，提升高硫柴油识别方法的鲁棒性和可靠性。Optionally, a major challenge in the application of big data in vehicle networking is unstable data quality. Under various unknown and complex working conditions in real driving scenarios, the initial oil product identification results may be misjudged to some extent. Therefore, a series of rules are used to correct the initial oil product identification results, so that the final oil product identification results obtained are more accurate, and the robustness and reliability of the high-sulfur diesel identification method are improved.
对未发生加油行为的范围内的各序列片段依据初步识别结果投票，统一识别结果。对于各序列片段，每个序列片段之间存在时间上的先后顺序。多个序列片段之间若车辆未发生加油行为，则其油品也不会发生改变。以车辆加油时间为节点将各序列片段划分若干个片段集合，对每个片段集合中所有的序列片段的初步识别结果按片段时长赋权投票，统一未发生加油行为的范围内片段的识别结果，提升精度并使之符合客观事实。Vote for each sequence segment within the range where no refueling behavior occurs based on the preliminary recognition results, and unify the recognition results. For each sequence fragment, there is a time sequence among each sequence fragment. If the vehicle does not refuel between multiple sequence segments, its oil quality will not change. Taking the refueling time of the vehicle as the node, divide each sequence fragment into several fragment sets, and vote for the initial recognition results of all the sequence fragments in each fragment set according to the length of the fragment, and unify the recognition results of the fragments in the range where no refueling behavior occurs, Improve accuracy and align it with objective facts.
例如，假设数据序列是{x1,x2,x3,x4,...,xn-1,xn}，其中每一个序列片段有效时长是{t1,t2,t3,t4,...,tn-1,tn}，各序列片段的初步识别结果是{1,1,0,0,...,1}，其中1代表正常柴油，0代表高硫柴油。假设在x3和x4之间存在加油行为，则划分为{x1,x2,x3}和{x4,...,xn-1,xn}两个片段集合，对于{x1,x2,x3}的初步识别结果，分别计算其正常柴油总时长T1＝t1+t2和高硫柴油总时长T2＝t3，投票原则如下式：For example, suppose the data sequence is {x1 ,x2 ,x3 ,x4 ,...,xn-1 ,xn }, where the effective duration of each sequence segment is {t1 ,t2 ,t3 , t4 ,...,tn-1 ,tn }, the preliminary identification result of each sequence fragment is {1,1,0,0,...,1}, where 1 represents normal diesel oil and 0 represents high sulfur diesel fuel. Assuming that there is refueling behavior between x3 and x4 , it is divided into two fragment sets {x1 ,x2 ,x3 } and {x4 ,...,xn-1 ,xn }, for { For the preliminary identification results of x1 , x2 , x3 }, calculate the total duration T1 = t1 + t2 of normal diesel and T2 = t3 of high-sulfur diesel respectively. The voting principle is as follows:
其中R代表每个片段集合的识别结果，α是可调缩放因子，可自行标定或调节。对于其它片段集合重复上述投票流程，修正后的识别结果在未发生加油行为的同一片段集合内的所有序列片段都一致。Among them, R represents the recognition result of each fragment set, and α is an adjustable scaling factor, which can be calibrated or adjusted by itself. Repeat the above voting process for other fragment sets, and the corrected recognition results are consistent with all the sequence fragments in the same fragment set without refueling behavior.
计算每个片段集合的SCR反应温度高于T部分的平均氨氮比c，即反应NOx总量与尿素喷射总量之比。每个片段集合的平均氨氮比是SCR健康状态的定量体现。为了控制变量，设置温度窗口去取SCR的高效工作区间。考虑到反应器内化学反应和物料传输的时滞性，窗口可适当放宽一些。窗口内的反应的NOx总量是SCR上下游NOx浓度之差的积分，尿素喷射总量是窗口内尿素喷射量的积分。Calculate the average ammonia-nitrogen ratio c of the SCR reaction temperature higher than T for each segment set, that is, the ratio of the total amount of reaction NOx to the total amount of urea injection. The average ammonia-nitrogen ratio of each fragment collection is a quantitative reflection of the SCR health status. In order to control the variables, set the temperature window to take the efficient working range of the SCR. Considering the time lag of chemical reaction and material transfer in the reactor, the window can be appropriately relaxed. The total amount of NOx reacted within the window is the integral of the difference between the upstream and downstream NOx concentrations of the SCR, and the total amount of urea injection is the integral of the urea injection amount within the window.
计算每个片段集合间的氨氮比向前差分Δc，若Δc大于容忍度Q1，则将片段集合中各序列片段的油品识别结果修正为正常柴油，若Δc小于容忍度Q2，则将片段集合中各序列片段的油品识别结果修正为高硫柴油，若Δc处于容忍度Q1到Q2范围内，则不做修正。从SCR的健康状态的变化趋势去进一步修正，获得最终的识别结果。车辆在使用高硫柴油一段时间后SCR状态趋于恶化，平均氨氮比减小；在使用正常柴油一段时间后SCR状态趋于平稳或缓解，平均氨氮比基本不变或增加。超出容忍度Q1到Q2范围的片段集合认为是显著背离机理逻辑的识别结果，这些结果需要被修正，其余则维持原判。Calculate the forward difference Δc of the ammonia-nitrogen ratio between each fragment set, if Δc is greater than the tolerance Q1 , then correct the oil product recognition result of each sequence fragment in the fragment set to normal diesel, if Δc is less than the tolerance Q2 , then set The oil identification results of each sequence fragment in the fragment set are corrected to high-sulfur diesel oil, and if Δc is within the range of tolerance Q1 to Q2 , no correction will be made. Further corrections are made from the changing trend of the health state of the SCR to obtain the final recognition result. After using high-sulfur diesel for a period of time, the SCR state of the vehicle tends to deteriorate, and the average ammonia-to-nitrogen ratio decreases; after using normal diesel for a period of time, the SCR state tends to stabilize or ease, and the average ammonia-to-nitrogen ratio basically remains unchanged or increases. Fragment collections that exceed the range of tolerance Q1 to Q2 are considered to be identification results that significantly deviate from the logic of the mechanism, and these results need to be corrected, while the rest remain the original judgment.
如图7所示，以上即为对序列片段的初始油品识别结果进行修正的工作流程。使用机理或经验知识去提升油品识别结果的可靠性和鲁棒性。在一些可行的实施方式中，基于规则的油品判别也可从SCR转化效率趋势，DPF的工作状态等角度出发，也可结合数据质量和需求去调整规则顺序、增减修正步骤从而获得新的识别规则。As shown in Figure 7, the above is the workflow for correcting the initial oil product identification results of the sequence fragments. Use mechanistic or empirical knowledge to improve the reliability and robustness of oil identification results. In some feasible implementations, rule-based oil product discrimination can also be based on the SCR conversion efficiency trend, DPF working status, etc., and can also be combined with data quality and demand to adjust the order of rules, increase or decrease correction steps to obtain new Identification rules.
本实施例中，根据数据序列对应的加油动作，以及数据序列对应的SCR系统氨氮比的变化趋势，对序列片段的初始油品识别结果进行两次修正，得到序列片段的最终油品识别结果，提高了油品识别的精度。In this embodiment, according to the refueling action corresponding to the data sequence and the change trend of the ammonia-nitrogen ratio of the SCR system corresponding to the data sequence, the initial oil product identification result of the sequence segment is corrected twice to obtain the final oil product identification result of the sequence segment, Improve the accuracy of oil identification.
在一个实施例中，一种油品识别方法。所述方法包括：In one embodiment, a method for identifying oil products. The methods include:
获取目标车辆在正常油品状态下的样本数据序列和SCR系统下游的氮氧化物浓度样本序列；基于样本数据序列和SCR系统下游的氮氧化物浓度样本序列获取训练实例，基于多个训练实例构建正常油品训练集。基于正常油品训练集训练得到氮氧化物浓度预测模型。Obtain the sample data sequence of the target vehicle in the normal oil state and the nitrogen oxide concentration sample sequence downstream of the SCR system; obtain training examples based on the sample data sequence and the nitrogen oxide concentration sample sequence downstream of the SCR system, and construct based on multiple training examples Normal oil training set. The NOx concentration prediction model was trained based on the normal oil training set.
根据目标车辆的网联数据获取目标车辆的数据序列，数据序列包括目标车辆在多个单位时间的整车工况信息。根据网联数据，获取数据序列对应的油量信息，根据油量信息确定数据序列对应的加油动作；根据加油动作在数据序列中确定第一分割节点；识别数据序列中的异常数据，根据异常数据在数据序列中确定第二分割节点；异常数据至少包括非油品异常数据和丢包掉帧数据；根据第一分割节点和第二分割节点对数据序列进行分割，获得数据序列中的序列片段；序列片段包括目标车辆在连续多个单位时间的整车工况信息。The data sequence of the target vehicle is obtained according to the networked data of the target vehicle, and the data sequence includes the vehicle working condition information of the target vehicle at multiple unit times. According to the network data, obtain the oil quantity information corresponding to the data sequence, determine the refueling action corresponding to the data sequence according to the oil quantity information; determine the first segmentation node in the data sequence according to the refueling action; identify the abnormal data in the data sequence, according to the abnormal data Determining the second segmentation node in the data sequence; the abnormal data at least includes non-oil abnormal data and packet loss and frame data; segmenting the data sequence according to the first segmentation node and the second segmentation node to obtain sequence fragments in the data sequence; The sequence fragments include the vehicle operating condition information of the target vehicle in multiple consecutive unit times.
基于氮氧化物浓度预测模型，获取序列片段对应的SCR系统下游的氮氧化物浓度预测序列。根据网联数据，获取序列片段对应的SCR系统下游的氮氧化物浓度实际序列。获取氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的残差序列；残差序列包括多个残差值，各残差值用于表征氮氧化物浓度预测序列和氮氧化物浓度实际序列在单位时间的数据差值；在残差序列中，获取残差值大于残差阈值的异常点，并确定异常点数量；根据异常点数量和残差序列的长度，获取序列片段的异常分数；在异常分数大于异常阈值的情况下，确定序列片段的初始油品识别结果为高硫油品；在异常分数不大于异常阈值的情况下，确定序列片段的初始油品识别结果为正常油品。Based on the NOx concentration prediction model, the NOx concentration prediction sequence downstream of the SCR system corresponding to the sequence fragment is obtained. According to the network data, the actual sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment is obtained. Obtain the residual sequence between the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration; the residual sequence includes multiple residual values, and each residual value is used to represent the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration The data difference value of the sequence per unit time; in the residual sequence, obtain the abnormal points whose residual value is greater than the residual threshold, and determine the number of abnormal points; according to the number of abnormal points and the length of the residual sequence, obtain the abnormal score of the sequence fragment ; When the abnormal score is greater than the abnormal threshold, determine the initial oil identification result of the sequence segment as high-sulfur oil; when the abnormal score is not greater than the abnormal threshold, determine the initial oil identification result of the sequence segment as normal oil .
根据网联数据，获取数据序列对应的油量信息，根据油量信息确定数据序列对应的加油动作；根据加油动作在数据序列中获取包含序列片段的片段集合；片段集合包括至少一个序列片段，各序列片段存在连续的先后顺序，且片段集合中不存在加油动作；根据片段集合中各序列片段的初始油品识别结果，以及各序列片段的片段时长，计算片段集合中的正常油品总时长和高硫油品总时长；根据片段集合的片段集合总时长、正常油品总时长和高硫油品总时长，计算高硫油品总时长在片段集合中的高硫油品时长占比；在高硫油品时长占比大于预设比例的片段集合总时长的情况下，确定片段集合的第一修正结果为高硫油品，将片段集合的第一修正结果作为片段集合中各序列片段的第一修正结果；在高硫油品时长占比不大于预设比例的片段集合总时长的情况下，确定片段集合的第一修正结果为正常油品，将片段集合的第一修正结果作为片段集合中各序列片段的第一修正结果。According to the network connection data, the oil quantity information corresponding to the data sequence is obtained, and the refueling action corresponding to the data sequence is determined according to the oil quantity information; according to the refueling action, a fragment set containing sequence fragments is obtained in the data sequence; the fragment set includes at least one sequence fragment, each There is a continuous sequence of sequence fragments, and there is no refueling action in the fragment set; according to the initial oil product recognition results of each sequence fragment in the fragment set, and the fragment duration of each sequence fragment, calculate the total duration and sum of normal oil products in the fragment set The total duration of high-sulfur oil products; according to the total duration of the fragment collection, the total duration of normal oil products and the total duration of high-sulfur oil products, calculate the proportion of the total duration of high-sulfur oil products to the duration of high-sulfur oil products in the fragment collection; When the proportion of high-sulfur oil product duration is greater than the total duration of the fragment set with a preset ratio, the first correction result of the fragment set is determined to be high-sulfur oil, and the first correction result of the fragment set is used as the sequence fragment of each sequence fragment in the fragment set. The first correction result; when the proportion of high-sulfur oil product duration is not greater than the total duration of the fragment set of the preset ratio, determine that the first correction result of the fragment set is normal oil, and use the first correction result of the fragment set as a fragment The first correction result of each sequence segment in the set.
根据网联数据，获取片段集合对应的SCR反应温度；在片段集合中，获取SCR反应温度大于温度阈值的部分，作为片段集合的目标子集；计算目标子集的平均氨氮比，并对平均氨氮比进行向前差分，计算得到氨氮比向前差分；在氨氮比向前差分小于预设容忍度的情况下，确定片段集合的第二修正结果为高硫油品，将片段集合的第二修正结果作为片段集合中各序列片段的第二修正结果；在氨氮比向前差分不小于预设容忍度的情况下，确定片段集合的第二修正结果为正常油品，将片段集合的第二修正结果作为片段集合中各序列片段的第二修正结果。According to the network connection data, obtain the SCR reaction temperature corresponding to the fragment set; in the fragment set, obtain the part whose SCR reaction temperature is greater than the temperature threshold, as the target subset of the fragment set; calculate the average ammonia nitrogen ratio of the target subset, and compare the average ammonia nitrogen The forward difference of the ammonia-nitrogen ratio is calculated to obtain the forward difference of the ammonia-nitrogen ratio; when the forward difference of the ammonia-nitrogen ratio is less than the preset tolerance, it is determined that the second correction result of the fragment set is high-sulfur oil, and the second correction of the fragment set The result is used as the second correction result of each sequence fragment in the fragment set; when the forward difference of the ammonia-nitrogen ratio is not less than the preset tolerance, the second correction result of the fragment set is determined to be normal oil, and the second correction result of the fragment set is The result is used as the second correction result of each sequence fragment in the fragment set.
根据第一修正结果对序列片段的初始油品识别结果进行第一次修正，根据第二修正结果对第一次修正后的序列片段的初始油品识别结果进行第二次修正，得到序列片段的最终油品识别结果。According to the first correction result, the initial oil product identification result of the sequence fragment is corrected for the first time, and according to the second correction result, the initial oil product identification result of the sequence fragment corrected for the first time is corrected for the second time, and the sequence fragment's initial oil product identification result is obtained. The final oil identification result.
在一个实施例中，一种高硫柴油识别系统及云平台，是一个结合硬件结构、数据存储、数据交互、云计算以及软件服务的数据服务系统，是高硫柴油识别问题和车辆远程油品监控的系统级解决方案。该系统的具体实施例的结构如图8所示，包括车辆远程终端、车辆远程大数据服务云平台和企业端用户接口三部分。In one embodiment, a high-sulfur diesel identification system and cloud platform is a data service system that combines hardware structure, data storage, data interaction, cloud computing, and software services. System-level solutions for monitoring. The structure of a specific embodiment of the system is shown in Figure 8, including three parts: a vehicle remote terminal, a vehicle remote big data service cloud platform, and an enterprise user interface.
车载远程终端是一种用于完成车辆流数据采集、数据存储和数据通信功能的电子设备。在一些具体实施例中，车载远程终端由OBD模块、存储模块、通信模块和控制模块组成。OBD模块采集车辆运行时产生的流数据并暂存于存储模块，通信模块同云平台建立通信协议完成数据发送和指令接受的功能，控制模块统筹进行上述流程。The vehicle-mounted remote terminal is an electronic device used to complete vehicle flow data collection, data storage and data communication functions. In some specific embodiments, the vehicle-mounted remote terminal is composed of an OBD module, a storage module, a communication module and a control module. The OBD module collects stream data generated during vehicle operation and temporarily stores it in the storage module. The communication module establishes a communication protocol with the cloud platform to complete the functions of data transmission and instruction acceptance. The control module coordinates the above processes.
车辆远程大数据服务云平台是一种用于完成数据云计算、云存储和交互功能的数据服务平台。在一些具体实施例中，所述云平台由大型高性能计算机集群、高容量存储介质和先进数据管理计算机软件组成。所述云平台中的数据存储功能单元是一种用于存储车载远程终端发送的数据和平台内部运算处理次级数据的计算机存储介质。结合第一方面，所述云平台中的高硫柴油识别功能单元是本发明提供的高硫柴油识别方法的一种可部署于云平台的软件实现，依托平台计算能力完成车辆油品识别任务。所述云平台中的人机交互界面是一种高硫柴油识别系统及云平台与用户之间交流的媒介，在一些具体实施例中，人机交互界面一般设置为完成各种复杂视觉显示功能的前端UI。The vehicle remote big data service cloud platform is a data service platform used to complete data cloud computing, cloud storage and interactive functions. In some specific embodiments, the cloud platform is composed of large-scale high-performance computer clusters, high-capacity storage media and advanced data management computer software. The data storage functional unit in the cloud platform is a computer storage medium for storing data sent by the vehicle-mounted remote terminal and secondary data for internal calculation and processing of the platform. In combination with the first aspect, the high-sulfur diesel identification functional unit in the cloud platform is a software implementation of the high-sulfur diesel identification method provided by the present invention that can be deployed on the cloud platform, and relies on the computing power of the platform to complete the vehicle oil product identification task. The human-computer interaction interface in the cloud platform is a communication medium between the high-sulfur diesel identification system and the cloud platform and the user. In some specific embodiments, the human-computer interaction interface is generally set to complete various complex visual display functions front-end UI.
企业访问终端是一种用户同车辆远程大数据服务云平台进行数据交互的软硬件设备。在一些具体实施例中，企业端访问终端可选为用户手机及APP、企业计算机及浏览器或操作台等。优选地，系统可设置加密模块来仅对受限的IP地址访问终端开放数据访问服务，以保障企业和用户机密。The enterprise access terminal is a hardware and software device for data interaction between the user and the vehicle remote big data service cloud platform. In some specific embodiments, the enterprise-side access terminal can be selected as the user's mobile phone and APP, enterprise computer and browser or operating console. Preferably, the system can set an encryption module to open data access services only to restricted IP address access terminals, so as to protect the confidentiality of enterprises and users.
高硫柴油识别系统及云平台的工作流程是：车载远程终端采集车辆运行流数据，实时传输至车辆远程大数据服务云平台的数据存储功能单元并完成存储；数据存储功能单元按一定窗口长度周期性向高硫柴油识别功能单元提供原始网联数据；高硫柴油识别功能单元经由数据预处理模块、残差网络判别模块和规则修正模块等步骤进行计算，将识别结果传递至数据存储功能单元完成存储；当企业端用户围绕车辆油品问题有信息检索、质询回溯等需求时，通过访问终端登录云平台人机交互界面并键入需求，数据存储功能单元就会向用户上传目标车辆目标时刻油品识别结果并显示。The workflow of the high-sulfur diesel identification system and cloud platform is as follows: the vehicle-mounted remote terminal collects vehicle operation flow data, transmits it to the data storage function unit of the vehicle remote big data service cloud platform in real time and completes the storage; Provide the original network connection data to the high-sulfur diesel identification functional unit; the high-sulfur diesel identification functional unit performs calculations through the steps of data preprocessing module, residual network identification module and rule correction module, and transfers the identification results to the data storage functional unit for storage ; When the enterprise user has information retrieval, query backtracking and other needs around the vehicle oil problem, he can log in to the human-computer interaction interface of the cloud platform through the access terminal and enter the demand, and the data storage function unit will upload the oil product identification of the target vehicle at the target time to the user. result and display.
本实施例中，高硫柴油识别系统及云平台实现了“企业—车辆—云平台”三端数据协同交互，满足企业在智能网联场景下对车辆远程油品监控的需求，为企业在解决高硫柴油相关法律纠纷、故障维修等方面提供数据支撑。In this embodiment, the high-sulfur diesel identification system and the cloud platform realize the collaborative interaction of three-terminal data of "enterprise-vehicle-cloud platform", which meets the needs of enterprises for remote oil product monitoring of vehicles in the intelligent network connection scenario, and provides solutions for enterprises Provide data support for legal disputes related to high-sulfur diesel, breakdown maintenance, etc.
需要说明的是，本发明中提到的“第一”,“第二”并不指代严格的逻辑顺序或其它逻辑关系，仅是为了将每个发明个体区分开来。本发明中提到的相似的简称文字和全称文字均指代同一个体，如“高硫柴油识别系统”和“系统”，“车辆远程大数据服务云平台”和“平台”等。本发明中提到的“包含”、“包括”、“由……组成”和“实现……功能”等概括性词语是示例性的而非排他性的，一切其它的并列要素都应视为本发明实施例所涵盖的。It should be noted that the "first" and "second" mentioned in the present invention do not refer to a strict logical order or other logical relationship, but are only for the purpose of distinguishing each individual invention. Similar abbreviations and full names mentioned in the present invention refer to the same entity, such as "high-sulfur diesel identification system" and "system", "vehicle remote big data service cloud platform" and "platform", etc. General words such as "comprising", "comprising", "consisting of" and "realizing the function" mentioned in the present invention are exemplary and not exclusive, and all other parallel elements should be regarded as the present invention. covered by the embodiments of the invention.
应该理解的是，虽然如上所述的各实施例所涉及的流程图中的各个步骤按照箭头的指示依次显示，但是这些步骤并不是必然按照箭头指示的顺序依次执行。除非本文中有明确的说明，这些步骤的执行并没有严格的顺序限制，这些步骤可以以其它的顺序执行。而且，如上所述的各实施例所涉及的流程图中的至少一部分步骤可以包括多个步骤或者多个阶段，这些步骤或者阶段并不必然是在同一时刻执行完成，而是可以在不同的时刻执行，这些步骤或者阶段的执行顺序也不必然是依次进行，而是可以与其它步骤或者其它步骤中的步骤或者阶段的至少一部分轮流或者交替地执行。It should be understood that although the steps in the flow charts involved in the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed sequentially in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least some of the steps in the flow charts involved in the above-mentioned embodiments may include multiple steps or stages, and these steps or stages are not necessarily executed at the same time, but may be performed at different times For execution, the execution order of these steps or stages is not necessarily performed sequentially, but may be executed in turn or alternately with other steps or at least a part of steps or stages in other steps.
基于同样的发明构思，本申请实施例还提供了一种用于实现上述所涉及的油品识别方法的油品识别装置。该装置所提供的解决问题的实现方案与上述方法中所记载的实现方案相似，故下面所提供的一个或多个油品识别装置实施例中的具体限定可以参见上文中对于油品识别方法的限定，在此不再赘述。Based on the same inventive concept, an embodiment of the present application also provides an oil product identification device for implementing the above-mentioned oil product identification method. The solution to the problem provided by the device is similar to the implementation described in the above method, so the specific limitations in one or more embodiments of the oil product identification device provided below can be referred to above for the oil product identification method limited and will not be repeated here.
在一个实施例中，如图9所示，提供了一种油品识别装置900，包括：获取模块901、预测模块902、比对模块903和识别模块904，其中：In one embodiment, as shown in FIG. 9 , an oil identification device 900 is provided, including: an acquisition module 901, a prediction module 902, a comparison module 903 and an identification module 904, wherein:
获取模块901，用于根据目标车辆的网联数据获取目标车辆的数据序列，对数据序列进行分割，在数据序列中确定序列片段；数据序列包括目标车辆在多个单位时间的整车工况信息。The acquisition module 901 is used to acquire the data sequence of the target vehicle according to the network connection data of the target vehicle, segment the data sequence, and determine the sequence fragments in the data sequence; the data sequence includes the vehicle operating condition information of the target vehicle at multiple unit times .
预测模块902，用于基于氮氧化物浓度预测模型，获取序列片段对应的SCR系统下游的氮氧化物浓度预测序列；氮氧化物浓度预测模型基于正常油品训练集训练得到。The prediction module 902 is configured to obtain the nitrogen oxide concentration prediction sequence downstream of the SCR system corresponding to the sequence segment based on the nitrogen oxide concentration prediction model; the nitrogen oxide concentration prediction model is trained based on the normal oil product training set.
比对模块903，用于获取序列片段对应的SCR系统下游的氮氧化物浓度实际序列，根据氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的差异程度，确定序列片段的初始油品识别结果。The comparison module 903 is used to obtain the actual sequence of nitrogen oxide concentration downstream of the SCR system corresponding to the sequence fragment, and determine the initial oil product of the sequence fragment according to the degree of difference between the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration recognition result.
识别模块904，用于根据数据序列对应的加油动作，以及数据序列对应的SCR系统氨氮比的变化趋势，对序列片段的初始油品识别结果进行修正，得到序列片段的最终油品识别结果。The identification module 904 is used to correct the initial oil product identification results of the sequence segments according to the refueling actions corresponding to the data sequences and the change trend of the ammonia-nitrogen ratio of the SCR system corresponding to the data sequences, and obtain the final oil product identification results of the sequence segments.
在一个实施例中，获取模块901还用于根据网联数据，获取数据序列对应的油量信息，根据油量信息确定数据序列对应的加油动作；根据加油动作在数据序列中确定第一分割节点；识别数据序列中的异常数据，根据异常数据在数据序列中确定第二分割节点；异常数据至少包括非油品异常数据和丢包掉帧数据；根据第一分割节点和第二分割节点对数据序列进行分割，获得数据序列中的序列片段；序列片段包括目标车辆在连续多个单位时间的整车工况信息。In one embodiment, the obtaining module 901 is also used to obtain the fuel quantity information corresponding to the data sequence according to the network data, determine the refueling action corresponding to the data sequence according to the fuel quantity information; determine the first segmentation node in the data sequence according to the refueling action ; Identify the abnormal data in the data sequence, and determine the second segmentation node in the data sequence according to the abnormal data; the abnormal data includes at least non-oil abnormal data and packet loss and frame data; according to the first segmentation node and the second segmentation node for data The sequence is segmented to obtain the sequence fragments in the data sequence; the sequence fragments include the vehicle working condition information of the target vehicle in multiple continuous unit times.
在一个实施例中，预测模块902还用于获取目标车辆在正常油品状态下的样本数据序列和SCR系统下游的氮氧化物浓度样本序列；基于样本数据序列和SCR系统下游的氮氧化物浓度样本序列获取训练实例，基于多个训练实例构建正常油品训练集。In one embodiment, the prediction module 902 is also used to obtain the sample data sequence of the target vehicle in normal oil state and the sample sequence of nitrogen oxide concentration downstream of the SCR system; based on the sample data sequence and the concentration of nitrogen oxide downstream of the SCR system The sample sequence obtains training instances, and a normal oil product training set is constructed based on multiple training instances.
在一个实施例中，比对模块903还用于获取氮氧化物浓度预测序列和氮氧化物浓度实际序列之间的残差序列；残差序列包括多个残差值，各残差值用于表征氮氧化物浓度预测序列和氮氧化物浓度实际序列在单位时间的数据差值；在残差序列中，获取残差值大于残差阈值的异常点，并确定异常点数量；根据异常点数量和残差序列的长度，获取序列片段的异常分数；在异常分数大于异常阈值的情况下，确定序列片段的初始油品识别结果为高硫油品；在异常分数不大于异常阈值的情况下，确定序列片段的初始油品识别结果为正常油品。In one embodiment, the comparison module 903 is also used to obtain the residual sequence between the predicted sequence of nitrogen oxide concentration and the actual sequence of nitrogen oxide concentration; the residual sequence includes a plurality of residual values, and each residual value is used for Characterize the data difference between the nitrogen oxide concentration prediction sequence and the actual nitrogen oxide concentration sequence per unit time; in the residual sequence, obtain the abnormal points whose residual value is greater than the residual threshold, and determine the number of abnormal points; according to the number of abnormal points and the length of the residual sequence to obtain the abnormal score of the sequence fragment; in the case that the abnormal score is greater than the abnormal threshold, determine that the initial oil product identification result of the sequence fragment is high-sulfur oil; in the case of the abnormal score not greater than the abnormal threshold, The initial oil product identification result of the sequence fragment was determined to be normal oil product.
在一个实施例中，识别模块904还用于根据加油动作，确定序列片段的第一修正结果；根据SCR系统氨氮比的变化趋势，确定序列片段的第二修正结果；根据第一修正结果对序列片段的初始油品识别结果进行第一次修正，根据第二修正结果对第一次修正后的序列片段的初始油品识别结果进行第二次修正，得到序列片段的最终油品识别结果。In one embodiment, the identification module 904 is also used to determine the first correction result of the sequence segment according to the refueling action; determine the second correction result of the sequence segment according to the change trend of the ammonia-nitrogen ratio of the SCR system; The initial oil product identification result of the segment is corrected for the first time, and the initial oil product identification result of the first corrected sequence segment is corrected for the second time according to the second correction result to obtain the final oil product identification result of the sequence segment.
在一个实施例中，识别模块904还用于根据网联数据，获取数据序列对应的油量信息，根据油量信息确定数据序列对应的加油动作；根据加油动作在数据序列中获取包含序列片段的片段集合；片段集合包括至少一个序列片段，各序列片段存在连续的先后顺序，且片段集合中不存在加油动作；根据片段集合中各序列片段的初始油品识别结果，以及各序列片段的片段时长，计算片段集合中的正常油品总时长和高硫油品总时长；根据片段集合的片段集合总时长、正常油品总时长和高硫油品总时长，计算高硫油品总时长在片段集合中的高硫油品时长占比；在高硫油品时长占比大于预设比例的片段集合总时长的情况下，确定片段集合的第一修正结果为高硫油品，将片段集合的第一修正结果作为片段集合中各序列片段的第一修正结果；在高硫油品时长占比不大于预设比例的片段集合总时长的情况下，确定片段集合的第一修正结果为正常油品，将片段集合的第一修正结果作为片段集合中各序列片段的第一修正结果。In one embodiment, the identification module 904 is also used to obtain the fuel quantity information corresponding to the data sequence according to the network data, determine the refueling action corresponding to the data sequence according to the fuel quantity information; Fragment collection; the fragment collection includes at least one sequence fragment, each sequence fragment has a continuous sequence, and there is no refueling action in the fragment collection; according to the initial oil product recognition result of each sequence fragment in the fragment collection, and the fragment duration of each sequence fragment , to calculate the total duration of normal oil products and the total duration of high-sulfur oil products in the fragment set; according to the total duration of fragment collection, the total duration of normal oil products and the total duration of high-sulfur oil products in the The duration ratio of high-sulfur oil products in the collection; in the case that the duration ratio of high-sulfur oil products is greater than the total duration of the fragment collection of the preset ratio, the first correction result of the fragment collection is determined to be high-sulfur oil products, and the fragment collection The first correction result is taken as the first correction result of each sequence fragment in the fragment set; when the proportion of high-sulfur oil product duration is not greater than the preset proportion of the total duration of the fragment set, it is determined that the first correction result of the fragment set is normal oil product, using the first correction result of the fragment set as the first correction result of each sequence fragment in the fragment set.
在一个实施例中，识别模块904还用于根据网联数据，获取片段集合对应的SCR反应温度；在片段集合中，获取SCR反应温度大于温度阈值的部分，作为片段集合的目标子集；计算目标子集的平均氨氮比，并对平均氨氮比进行向前差分，计算得到氨氮比向前差分；在氨氮比向前差分小于预设容忍度的情况下，确定片段集合的第二修正结果为高硫油品，将片段集合的第二修正结果作为片段集合中各序列片段的第二修正结果；在氨氮比向前差分不小于预设容忍度的情况下，确定片段集合的第二修正结果为正常油品，将片段集合的第二修正结果作为片段集合中各序列片段的第二修正结果。In one embodiment, the identification module 904 is also used to obtain the SCR reaction temperature corresponding to the fragment set according to the network connection data; in the fragment set, obtain the part whose SCR reaction temperature is greater than the temperature threshold, as the target subset of the fragment set; calculate The average ammonia-nitrogen ratio of the target subset, and the average ammonia-nitrogen ratio is forwarded to calculate the forward difference of the ammonia-nitrogen ratio; when the forward difference of the ammonia-nitrogen ratio is less than the preset tolerance, the second correction result of the determined fragment set is For high-sulfur oil products, the second correction result of the fragment set is used as the second correction result of each sequence fragment in the fragment set; when the forward difference of the ammonia nitrogen ratio is not less than the preset tolerance, the second correction result of the fragment set is determined For normal oil products, the second correction result of the fragment set is used as the second correction result of each sequence fragment in the fragment set.
上述油品识别装置中的各个模块可全部或部分通过软件、硬件及其组合来实现。上述各模块可以硬件形式内嵌于或独立于计算机设备中的处理器中，也可以以软件形式存储于计算机设备中的存储器中，以便于处理器调用执行以上各个模块对应的操作。Each module in the above-mentioned oil identification device can be fully or partially realized by software, hardware and combinations thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, and can also be stored in the memory of the computer device in the form of software, so that the processor can invoke and execute the corresponding operations of the above-mentioned modules.
在一个实施例中，提供了一种计算机设备，该计算机设备可以是服务器，其内部结构图可以如图10所示。该计算机设备包括处理器、存储器、输入/输出接口(Input/Output，简称I/O)和通信接口。其中，处理器、存储器和输入/输出接口通过系统总线连接，通信接口通过输入/输出接口连接到系统总线。其中，该计算机设备的处理器用于提供计算和控制能力。该计算机设备的存储器包括非易失性存储介质和内存储器。该非易失性存储介质存储有操作系统、计算机程序和数据库。该内存储器为非易失性存储介质中的操作系统和计算机程序的运行提供环境。该计算机设备的数据库用于存储车辆网联数据。该计算机设备的输入/输出接口用于处理器与外部设备之间交换信息。该计算机设备的通信接口用于与外部的终端通过网络连接通信。该计算机程序被处理器执行时以实现一种油品识别方法。In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure may be as shown in FIG. 10 . The computer device includes a processor, a memory, an input/output interface (Input/Output, I/O for short), and a communication interface. Wherein, the processor, the memory and the input/output interface are connected through the system bus, and the communication interface is connected to the system bus through the input/output interface. Wherein, the processor of the computer device is used to provide calculation and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs and databases. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer equipment is used to store vehicle networking data. The input/output interface of the computer device is used for exchanging information between the processor and external devices. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, an oil identification method is realized.
本领域技术人员可以理解，图10中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的计算机设备的限定，具体的计算机设备可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。Those skilled in the art can understand that the structure shown in Figure 10 is only a block diagram of a part of the structure related to the solution of this application, and does not constitute a limitation to the computer equipment on which the solution of this application is applied. The specific computer equipment can be More or fewer components than shown in the figures may be included, or some components may be combined, or have a different arrangement of components.
在一个实施例中，还提供了一种计算机设备，包括存储器和处理器，存储器中存储有计算机程序，该处理器执行计算机程序时实现上述各方法实施例中的步骤。In one embodiment, there is also provided a computer device, including a memory and a processor, where a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.
在一个实施例中，提供了一种计算机可读存储介质，其上存储有计算机程序，该计算机程序被处理器执行时实现上述各方法实施例中的步骤。In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the steps in the foregoing method embodiments are implemented.
在一个实施例中，提供了一种计算机程序产品，包括计算机程序，该计算机程序被处理器执行时实现上述各方法实施例中的步骤。In one embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the foregoing method embodiments are implemented.
需要说明的是，本申请所涉及的用户信息(包括但不限于用户设备信息、用户个人信息等)和数据(包括但不限于用于分析的数据、存储的数据、展示的数据等)，均为经用户授权或者经过各方充分授权的信息和数据，且相关数据的收集、使用和处理需要遵守相关国家和地区的相关法律法规和标准。It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all It is information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data need to comply with relevant laws, regulations and standards of relevant countries and regions.
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机程序来指令相关的硬件来完成，所述的计算机程序可存储于一非易失性计算机可读取存储介质中，该计算机程序在执行时，可包括如上述各方法的实施例的流程。其中，本申请所提供的各实施例中所使用的对存储器、数据库或其它介质的任何引用，均可包括非易失性和易失性存储器中的至少一种。非易失性存储器可包括只读存储器(Read-OnlyMemory，ROM)、磁带、软盘、闪存、光存储器、高密度嵌入式非易失性存储器、阻变存储器(ReRAM)、磁变存储器(Magnetoresistive Random Access Memory，MRAM)、铁电存储器(Ferroelectric Random Access Memory，FRAM)、相变存储器(Phase Change Memory，PCM)、石墨烯存储器等。易失性存储器可包括随机存取存储器(Random Access Memory，RAM)或外部高速缓冲存储器等。作为说明而非局限，RAM可以是多种形式，比如静态随机存取存储器(Static Random Access Memory，SRAM)或动态随机存取存储器(Dynamic RandomAccess Memory，DRAM)等。本申请所提供的各实施例中所涉及的数据库可包括关系型数据库和非关系型数据库中至少一种。非关系型数据库可包括基于区块链的分布式数据库等，不限于此。本申请所提供的各实施例中所涉及的处理器可为通用处理器、中央处理器、图形处理器、数字信号处理器、可编程逻辑器、基于量子计算的数据处理逻辑器等，不限于此。Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above-mentioned embodiments can be completed by instructing related hardware through computer programs, and the computer programs can be stored in a non-volatile computer-readable memory In the medium, when the computer program is executed, it may include the processes of the embodiments of the above-mentioned methods. Wherein, any reference to storage, database or other media used in the various embodiments provided in the present application may include at least one of non-volatile and volatile storage. Non-volatile memory can include read-only memory (Read-Only Memory, ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive variable memory (ReRAM), magnetic variable memory (Magnetoresistive Random Access Memory, MRAM), Ferroelectric Random Access Memory (FRAM), Phase Change Memory (Phase Change Memory, PCM), graphene memory, etc. The volatile memory may include random access memory (Random Access Memory, RAM) or external cache memory. As an illustration and not a limitation, RAM can be in various forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (Dynamic Random Access Memory, DRAM). The databases involved in the various embodiments provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a blockchain-based distributed database, etc., but is not limited thereto. The processors involved in the various embodiments provided by this application can be general-purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, data processing logic devices based on quantum computing, etc., and are not limited to this.
以上实施例的各技术特征可以进行任意的组合，为使描述简洁，未对上述实施例中的各个技术特征所有可能的组合都进行描述，然而，只要这些技术特征的组合不存在矛盾，都应当认为是本说明书记载的范围。The technical features of the above embodiments can be combined arbitrarily. To make the description concise, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be It is considered to be within the range described in this specification.
以上所述实施例仅表达了本申请的几种实施方式，其描述较为具体和详细，但并不能因此而理解为对本申请专利范围的限制。应当指出的是，对于本领域的普通技术人员来说，在不脱离本申请构思的前提下，还可以做出若干变形和改进，这些都属于本申请的保护范围。因此，本申请的保护范围应以所附权利要求为准。The above-mentioned embodiments only express several implementation modes of the present application, and the description thereof is relatively specific and detailed, but should not be construed as limiting the patent scope of the present application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be determined by the appended claims.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310494582.2ACN116541676B (en) | 2023-05-04 | 2023-05-04 | Oil product identification method, device, computer equipment and storage medium |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310494582.2ACN116541676B (en) | 2023-05-04 | 2023-05-04 | Oil product identification method, device, computer equipment and storage medium |
| Publication Number | Publication Date |
|---|---|
| CN116541676Atrue CN116541676A (en) | 2023-08-04 |
| CN116541676B CN116541676B (en) | 2025-09-23 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310494582.2AActiveCN116541676B (en) | 2023-05-04 | 2023-05-04 | Oil product identification method, device, computer equipment and storage medium |
| Country | Link |
|---|---|
| CN (1) | CN116541676B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119601128A (en)* | 2024-11-20 | 2025-03-11 | 北京铝能清新环境技术有限公司 | A flue gas outlet NOx concentration prediction method based on real-time calibration of LSTM model |
| CN120375984A (en)* | 2025-06-27 | 2025-07-25 | 中国环境科学研究院 | Method, device and equipment for on-line recognition of gas station and oil anomaly |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102012211709A1 (en)* | 2012-07-05 | 2014-01-09 | Robert Bosch Gmbh | Method for checking cross sensitivity of ammonia in SCR catalyst system of diesel engine of motor car, involves recognizing fault of nitrogen oxide sensor in predetermined period according to increase in temperature of catalyst |
| KR20180023233A (en)* | 2016-08-25 | 2018-03-07 | 현대자동차주식회사 | Method for checking of using high-sulfur fuel |
| CN110761883A (en)* | 2019-12-27 | 2020-02-07 | 潍柴动力股份有限公司 | Method and apparatus for predicting sulfur poisoning |
| CN114233471A (en)* | 2021-12-24 | 2022-03-25 | 湖南道依茨动力有限公司 | High-sulfur oil identification and post-processing method and device and working vehicle |
| CN115270631A (en)* | 2022-08-01 | 2022-11-01 | 一汽解放汽车有限公司 | Material circulation degree prediction method and device, computer equipment and storage medium |
| CN115840902A (en)* | 2022-06-29 | 2023-03-24 | 东风商用车有限公司 | Method for diagnosing poor-quality fuel adding place |
| WO2023045413A1 (en)* | 2021-09-23 | 2023-03-30 | 湖南道依茨动力有限公司 | Engine control method, engine control apparatus, and vehicle |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102012211709A1 (en)* | 2012-07-05 | 2014-01-09 | Robert Bosch Gmbh | Method for checking cross sensitivity of ammonia in SCR catalyst system of diesel engine of motor car, involves recognizing fault of nitrogen oxide sensor in predetermined period according to increase in temperature of catalyst |
| KR20180023233A (en)* | 2016-08-25 | 2018-03-07 | 현대자동차주식회사 | Method for checking of using high-sulfur fuel |
| CN110761883A (en)* | 2019-12-27 | 2020-02-07 | 潍柴动力股份有限公司 | Method and apparatus for predicting sulfur poisoning |
| WO2023045413A1 (en)* | 2021-09-23 | 2023-03-30 | 湖南道依茨动力有限公司 | Engine control method, engine control apparatus, and vehicle |
| CN114233471A (en)* | 2021-12-24 | 2022-03-25 | 湖南道依茨动力有限公司 | High-sulfur oil identification and post-processing method and device and working vehicle |
| CN115840902A (en)* | 2022-06-29 | 2023-03-24 | 东风商用车有限公司 | Method for diagnosing poor-quality fuel adding place |
| CN115270631A (en)* | 2022-08-01 | 2022-11-01 | 一汽解放汽车有限公司 | Material circulation degree prediction method and device, computer equipment and storage medium |
| Title |
|---|
| 冯曦;李军;陶汉国;李志军;申博玺;: "柴油车非循环工况NO_x排放评价方法研究", 内燃机工程, no. 05, 15 October 2018 (2018-10-15)* |
| 吴建华;袁小芳;刘凌佳;: "基于WebGIS的船舶尾气遥测数据分析系统设计", 现代电子技术, no. 12, 15 June 2020 (2020-06-15)* |
| 杨蓉 , 杨林 , 谭盛兰 , 张松 , 黄伟 , 黄俊明: "基于遗传算法-优化长短期记忆神经网络的柴油机瞬态NOx排放预测模型研究", 《内燃机工程》, 28 February 2022 (2022-02-28)* |
| 苏盈盈 , 张气皓 , 罗妤 , 周昊 , 何亚平 , 阎垒: "基于AE-LSTM混合神经网络模型的NOx排放预测", 《西南师范大学学报（自然科学版）》, 30 April 2023 (2023-04-30)* |
| 郭振华: "大功率船用柴油机NO_, x_排放测试计算及预测方法研究", 硕士学位论文, 10 March 2002 (2002-03-10)* |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119601128A (en)* | 2024-11-20 | 2025-03-11 | 北京铝能清新环境技术有限公司 | A flue gas outlet NOx concentration prediction method based on real-time calibration of LSTM model |
| CN120375984A (en)* | 2025-06-27 | 2025-07-25 | 中国环境科学研究院 | Method, device and equipment for on-line recognition of gas station and oil anomaly |
| Publication number | Publication date |
|---|---|
| CN116541676B (en) | 2025-09-23 |
| Publication | Publication Date | Title |
|---|---|---|
| CN116541676A (en) | Oil product identification method, device, computer equipment and storage medium | |
| CN113919235B (en) | Mobile source pollution abnormal emission detection method and medium based on LSTM evolution clustering | |
| CN113295826B (en) | Motor vehicle exhaust emission test management system and intelligent diagnosis method based on Internet of vehicles | |
| CN118292970B (en) | Heavy vehicle DPF monitoring method, device, medium and equipment based on remote data | |
| CN113870046A (en) | Power equipment fault diagnosis method and equipment | |
| CN113657651A (en) | Diesel vehicle emission prediction method, medium and equipment based on deep migration learning | |
| CN118688641A (en) | Battery health status prediction method, device and computer readable storage medium | |
| CN114970926A (en) | Model training method, enterprise operation risk prediction method and device | |
| CN116741304A (en) | Diesel engine original exhaust nitrogen oxide prediction method based on Internet of vehicles big data | |
| CN117762770A (en) | Buried point data real-time verification method, system, electronic equipment and storage medium | |
| CN115221218A (en) | Quality evaluation method and device for vehicle data, computer equipment and storage medium | |
| Yang et al. | Enhancing vehicular emissions monitoring: A GA-GRU-based soft sensors approach for HDDVs | |
| CN118861927A (en) | A method and device for detecting power lithium battery data fault | |
| CN114740815B (en) | Passenger car fault diagnosis method and device based on neural network and electronic equipment | |
| CN110470481A (en) | Fault Diagnosis of Engine based on BP neural network | |
| CN114611615A (en) | Object classification processing method and device, computer equipment and storage medium | |
| CN117171921B (en) | Method and device for evaluating and processing health state of diesel particulate tail gas purification device | |
| CN117633680B (en) | Metering standard analysis method, device and equipment for direct-current electric energy meter and storage medium | |
| CN117035563B (en) | Product quality safety risk monitoring method, device, monitoring system and medium | |
| CN118014006A (en) | Method and system for predicting multitasking business process of sewage treatment plant | |
| CN116859890A (en) | Regeneration failure prediction method, device, computer equipment and storage medium | |
| WO2023246391A1 (en) | Extraction of risk feature description | |
| Lins et al. | Methodology for Commercial Vehicle Mechanical Systems Maintenance: Data-Driven and Deep-Learning-Based Prediction | |
| CN115994073A (en) | Log analysis method and device and electronic equipment | |
| CN115564410A (en) | State monitoring method and device for relay protection equipment |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |