技术领域Technical Field
本发明涉及管道技术领域，具体地涉及一种管道上方威胁事件的识别模型的训练方法及识别方法。The present invention relates to the technical field of pipelines, and in particular to a training method and an identification method for an identification model of threat events above pipelines.
背景技术Background technique
管道是用管子、管子联接件和阀门等联接成的用于输送气体、液体或带固体颗粒的流体的装置。通常，流体经鼓风机、压缩机、泵和锅炉等增压后，从管道的高压处流向低压处，也可利用流体自身的压力或重力输送。管道的用途很广泛，主要用在给水、排水、供热、供煤气、长距离输送石油和天然气、农业灌溉、水力工程和各种工业装置中。以天然气管道和输油管道进行示例，天然气管道是指将天然气(包括油田生产的伴生气)从开采地或处理厂输送到城市配气中心或工业企业用户的管道，又称输气管道。利用天然气管道输送天然气，是陆地上大量输送天然气的方式。输油管道(也称管线、管路)是由油管及其附件所组成，并按照工艺流程的需要，配备相应的油泵机组，设计安装成一个完整的管道系统，用于完成油料接卸及输转任务。A pipeline is a device connected by pipes, pipe connectors and valves for transporting gas, liquid or fluid with solid particles. Usually, the fluid flows from the high pressure part of the pipeline to the low pressure part after being pressurized by blowers, compressors, pumps and boilers. It can also be transported by the pressure of the fluid itself or gravity. Pipelines are widely used, mainly in water supply, drainage, heating, gas supply, long-distance transportation of oil and natural gas, agricultural irrigation, hydraulic engineering and various industrial devices. Taking natural gas pipelines and oil pipelines as examples, natural gas pipelines refer to pipelines that transport natural gas (including associated gas produced by oil fields) from mining sites or processing plants to urban gas distribution centers or industrial enterprise users, also known as gas pipelines. Using natural gas pipelines to transport natural gas is a way to transport natural gas in large quantities on land. Oil pipelines (also called pipelines and pipelines) are composed of oil pipes and their accessories, and are equipped with corresponding oil pump units according to the needs of the process flow. They are designed and installed into a complete pipeline system to complete the tasks of oil loading and unloading and transportation.
管道的安全运行有着重大的社会和经济意义。传统的管道安全防护主要通过安装摄像头并利用监控人工识别是否存在管道上方作业的威胁事件，人工监控识别并不能及时发现破坏管道的行为，导致管道的安全性较低。The safe operation of pipelines has great social and economic significance. Traditional pipeline safety protection mainly relies on installing cameras and using monitoring to manually identify whether there are threatening events above the pipeline. Manual monitoring and identification cannot detect behaviors that damage the pipeline in time, resulting in low pipeline safety.
发明内容Summary of the invention
为了克服现有技术存在的不足，本发明实施例提供了管道上方威胁事件的识别模型的训练方法及识别方法。In order to overcome the deficiencies of the prior art, an embodiment of the present invention provides a training method and an identification method for an identification model of threat events above a pipeline.
为了实现上述目的，本发明第一方面提供一种管道上方威胁事件的识别模型的训练方法，包括：In order to achieve the above object, the first aspect of the present invention provides a method for training a recognition model of threat events above a pipeline, comprising:
利用设置于管道上方的光纤获取第一振动数据；Acquiring first vibration data using an optical fiber disposed above the pipeline;
根据第一振动数据生成对应的第一相位数据和第一强度数据；generating corresponding first phase data and first intensity data according to the first vibration data;
根据第一相位数据和第一强度数据得到第一特征数据，其中第一特征数据包括第一瀑布图图像的颜色变化趋势的数据，第一瀑布图图像是基于第一振动数据生成的；Obtaining first characteristic data according to the first phase data and the first intensity data, wherein the first characteristic data includes data on a color change trend of a first waterfall chart image, the first waterfall chart image being generated based on the first vibration data;
利用第一特征数据构建样本集；constructing a sample set using the first feature data;
根据样本集对深度神经网络分类模型进行训练，得到管道上方威胁事件的识别模型。The deep neural network classification model is trained according to the sample set to obtain the recognition model of threat events above the pipeline.
在本发明实施例中，样本集包括训练集和验证集，利用第一特征数据构建样本集包括：In the embodiment of the present invention, the sample set includes a training set and a validation set, and constructing the sample set using the first feature data includes:
将第一特征数据分成训练集和验证集；Divide the first feature data into a training set and a validation set;
根据样本集对深度神经网络分类模型进行训练，得到管道上方威胁事件的识别模型包括：The deep neural network classification model is trained based on the sample set, and the recognition model of threat events above the pipeline is obtained, including:
初始化深度神经网络分类模型的权重梯度；Initialize the weight gradients of the deep neural network classification model;
将训练集分批次输入至深度神经网络分类模型；Input the training set into the deep neural network classification model in batches;
根据深度神经网络分类模型的设计层数，逐层进行前向传播到输出层；According to the designed number of layers of the deep neural network classification model, forward propagation is performed layer by layer to the output layer;
根据损失函数确定输出层的输出结果与真实标签的损失值；Determine the loss value between the output result of the output layer and the true label according to the loss function;
基于损失值进行反向传播，并反向链式地逐层更新权重梯度；Back propagates based on the loss value and updates the weight gradient layer by layer in a reverse chain manner;
在验证集验证深度神经网络分类模型的准确率大于预设阈值的情况下，得到管道上方威胁事件的识别模型。When the accuracy of the deep neural network classification model verified by the validation set is greater than the preset threshold, the recognition model of the threat events above the pipeline is obtained.
本发明第二方面提供一种管道上方威胁事件的识别方法，包括：A second aspect of the present invention provides a method for identifying a threat event above a pipeline, comprising:
利用设置于管道上方的光纤获取第二振动数据；Acquiring second vibration data using an optical fiber disposed above the pipeline;
根据第二振动数据生成对应的第二相位数据和第二强度数据；generating corresponding second phase data and second intensity data according to the second vibration data;
根据第二相位数据和第二强度数据得到第二特征数据，其中第二特征数据包括第二瀑布图图像的颜色变化趋势的数据，第二瀑布图图像是基于第二振动数据生成的；Obtaining second characteristic data according to the second phase data and the second intensity data, wherein the second characteristic data includes data on a color change trend of a second waterfall chart image, the second waterfall chart image being generated based on the second vibration data;
将第二特征数据输入至管道上方威胁事件的识别模型，得到管道上方威胁事件的识别结果，其中识别模型通过如上述的管道上方威胁事件的识别模型的训练方法得到。The second characteristic data is input into a recognition model of threat events above the pipeline to obtain a recognition result of the threat events above the pipeline, wherein the recognition model is obtained by the training method of the recognition model of threat events above the pipeline as described above.
在本发明实施例中，利用设置于管道上方的光纤获取第二振动数据包括：In an embodiment of the present invention, obtaining the second vibration data by using an optical fiber disposed above the pipeline includes:
利用设置于管道上方的光纤获取光纤传感瑞利散射回的第二振动数据。The optical fiber disposed above the pipeline is used to obtain second vibration data sent back by the optical fiber sensing Rayleigh scattering.
本发明第三方面提供一种管道上方威胁事件的识别模型的训练装置，包括：A third aspect of the present invention provides a training device for a recognition model of threat events above a pipeline, comprising:
第一获取模块，用于利用设置于管道上方的光纤获取第一振动数据；A first acquisition module, used for acquiring first vibration data by using an optical fiber arranged above the pipeline;
第一生成模块，用于根据第一振动数据生成对应的第一相位数据和第一强度数据；A first generating module, used for generating corresponding first phase data and first intensity data according to the first vibration data;
第一得到模块，用于根据第一相位数据和第一强度数据得到第一特征数据，其中第一特征数据包括第一瀑布图图像的颜色变化趋势的数据，第一瀑布图图像是基于第一振动数据生成的；A first obtaining module, configured to obtain first characteristic data according to the first phase data and the first intensity data, wherein the first characteristic data includes data of a color change trend of a first waterfall chart image, and the first waterfall chart image is generated based on the first vibration data;
构建模块，用于利用第一特征数据构建样本集；A construction module, used to construct a sample set using the first feature data;
训练模块，用于根据样本集对深度神经网络分类模型进行训练，得到管道上方威胁事件的识别模型。The training module is used to train the deep neural network classification model according to the sample set to obtain an identification model for threat events above the pipeline.
在本发明实施例中，样本集包括训练集和验证集，构建模块包括：In the embodiment of the present invention, the sample set includes a training set and a validation set, and the construction module includes:
分成单元，用于将第一特征数据分成训练集和验证集；A division unit is used to divide the first feature data into a training set and a validation set;
训练模块包括：The training modules include:
初始化单元，用于初始化深度神经网络分类模型的权重梯度；Initialization unit, used to initialize the weight gradient of the deep neural network classification model;
输入单元，用于将训练集分批次输入至所述深度神经网络分类模型；An input unit, used to input the training set into the deep neural network classification model in batches;
传播单元，用于根据深度神经网络分类模型的设计层数，逐层进行前向传播到输出层；The propagation unit is used to perform forward propagation to the output layer layer by layer according to the designed number of layers of the deep neural network classification model;
确定单元，用于根据损失函数确定所述输出层的输出结果与真实标签的损失值；A determination unit, used to determine a loss value between an output result of the output layer and a true label according to a loss function;
更新单元，用于基于损失值进行反向传播，并反向链式地逐层更新权重梯度；The update unit is used to perform back propagation based on the loss value and update the weight gradient layer by layer in a reverse chain manner;
验证单元，用于在验证集验证深度神经网络分类模型的准确率大于预设阈值的情况下，得到管道上方威胁事件的识别模型。The verification unit is used to obtain an identification model of threat events above the pipeline when the accuracy of the deep neural network classification model verified by the verification set is greater than a preset threshold.
本发明第四方面提供一种管道上方威胁事件的识别装置，包括：A fourth aspect of the present invention provides a device for identifying a threat event above a pipeline, comprising:
第二获取模块，用于利用设置于管道上方的光纤获取第二振动数据；A second acquisition module, used for acquiring second vibration data by using an optical fiber arranged above the pipeline;
第二生成模块，用于根据第二振动数据生成对应的第二相位数据和第二强度数据；A second generating module, used for generating corresponding second phase data and second intensity data according to the second vibration data;
第二得到模块，用于根据第二相位数据和第二强度数据得到第二特征数据，其中第二特征数据包括第二瀑布图图像的颜色变化趋势的数据，第二瀑布图图像是基于第二振动数据生成的；A second obtaining module is used to obtain second characteristic data according to the second phase data and the second intensity data, wherein the second characteristic data includes data of a color change trend of a second waterfall chart image, and the second waterfall chart image is generated based on the second vibration data;
输入模块，用于将第二特征数据输入至管道上方威胁事件的识别模型，得到管道上方威胁事件的识别结果，其中识别模型通过上述的管道上方威胁事件的识别模型的训练方法得到。The input module is used to input the second feature data into the recognition model of the threat event above the pipeline to obtain the recognition result of the threat event above the pipeline, wherein the recognition model is obtained by the training method of the recognition model of the threat event above the pipeline.
在本发明实施例中，第二获取模块包括：In an embodiment of the present invention, the second acquisition module includes:
获取单元，用于利用设置于管道上方的光纤获取光纤传感瑞利散射回的第二振动数据。The acquisition unit is used to acquire the second vibration data sent back by the optical fiber sensing Rayleigh scattering by using the optical fiber arranged above the pipeline.
本发明第五方面提供一种计算机设备，包括存储器以及处理器，存储器存储有计算机程序，计算机程序在处理器上运行时执行上述的管道上方威胁事件的识别装置的训练方法，或上述的管道上方威胁事件的识别方法。A fifth aspect of the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program runs on the processor, it executes the training method of the above-mentioned device for identifying threat events above the pipeline, or the above-mentioned method for identifying threat events above the pipeline.
本发明第六方面提供一种机器可读存储介质，该机器可读存储介质上存储有指令，该指令用于使得机器执行根据上述的管道上方威胁事件的识别装置的训练方法，或上述的管道上方威胁事件的识别方法。A sixth aspect of the present invention provides a machine-readable storage medium having instructions stored thereon, the instructions being used to enable a machine to execute a training method for an apparatus for identifying threat events above a pipeline, or a method for identifying threat events above a pipeline.
在本发明实施例中，管道上方威胁事件的识别模型的训练方法包括：利用设置于管道上方的光纤获取第一振动数据；根据第一振动数据生成对应的第一相位数据和第一强度数据；根据第一相位数据和第一强度数据得到第一特征数据，其中第一特征数据包括第一瀑布图图像的颜色变化趋势的数据，第一瀑布图图像是基于第一振动数据生成的；利用第一特征数据构建样本集；根据样本集对深度神经网络分类模型进行训练，得到管道上方威胁事件的识别模型。In an embodiment of the present invention, a training method for an identification model of threat events above a pipeline includes: obtaining first vibration data using an optical fiber disposed above the pipeline; generating corresponding first phase data and first intensity data based on the first vibration data; obtaining first feature data based on the first phase data and the first intensity data, wherein the first feature data includes data on a color change trend of a first waterfall chart image, and the first waterfall chart image is generated based on the first vibration data; constructing a sample set using the first feature data; and training a deep neural network classification model based on the sample set to obtain an identification model of threat events above the pipeline.
管道上方发生威胁事件(例如在管道上方施工)与管道上方未发生威胁事件(例如非管道上方施工)对应的瀑布图图像的颜色变化趋势是不一样的。由于第一特征数据包括第一瀑布图图像的颜色变化趋势的数据，第一瀑布图图像是基于第一振动数据生成的，这样深度神经网络分类模型可以学习管道上方发生威胁事件时(例如施工)瀑布图图像的特征，然后基于管道上方威胁事件的识别模型，可以实现利用瀑布图图像的颜色变化趋势判定是否在管道上方作业(施工)，实时检测管道上方是否存在施工作业，及时识别出管道上方的威胁事件，保护管道(例如油气管道)免受破坏，保证管道的安全运行，提高管道的安全性。另外，利用设置于管道上方的光纤获取振动数据，抗干扰性强，不受天气亮度等因素影响，管道上方威胁事件的识别模型的识别准确度较高。The color change trend of the waterfall image corresponding to a threat event (e.g., construction above the pipeline) and a threat event (e.g., construction not above the pipeline) above the pipeline is different. Since the first feature data includes data on the color change trend of the first waterfall image, and the first waterfall image is generated based on the first vibration data, the deep neural network classification model can learn the features of the waterfall image when a threat event (e.g., construction) occurs above the pipeline. Then, based on the recognition model of the threat event above the pipeline, it is possible to use the color change trend of the waterfall image to determine whether there is work (construction) above the pipeline, detect in real time whether there is construction work above the pipeline, and promptly identify the threat event above the pipeline, protect the pipeline (e.g., oil and gas pipeline) from damage, ensure the safe operation of the pipeline, and improve the safety of the pipeline. In addition, the vibration data is obtained by using the optical fiber set above the pipeline, which has strong anti-interference and is not affected by factors such as weather brightness. The recognition accuracy of the recognition model of the threat event above the pipeline is high.
附图说明BRIEF DESCRIPTION OF THE DRAWINGS
附图是用来提供对本发明实施例的进一步理解，并且构成说明书的一部分，与下面的具体实施方式一起用于解释本发明实施例，但并不构成对本发明实施例的限制。在附图中：The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:
图1示意性示出了根据本发明实施例的管道上方威胁事件的识别模型的训练方法的流程图；FIG1 schematically shows a flow chart of a method for training a recognition model of threat events above a pipeline according to an embodiment of the present invention;
图2示意性示出了根据本发明实施例的深度神经网络分类模型的示意图；FIG2 schematically shows a schematic diagram of a deep neural network classification model according to an embodiment of the present invention;
图3示意性示出了根据本发明实施例的非管道上方作业的瀑布图图像的示意图；FIG3 schematically shows a schematic diagram of a waterfall chart image of a non-pipeline above operation according to an embodiment of the present invention;
图4示意性示出了根据本发明实施例的在管道上方作业的瀑布图图像的示意图之一；FIG4 schematically shows one of the schematic diagrams of a waterfall diagram image operating above a pipeline according to an embodiment of the present invention;
图5示意性示出了根据本发明实施例的在管道上方作业的瀑布图图像的示意图之二；FIG5 schematically shows a second schematic diagram of a waterfall diagram image operating above a pipeline according to an embodiment of the present invention;
图6示意性示出了根据本发明实施例的管道上方威胁事件的识别方法的流程图。FIG6 schematically shows a flow chart of a method for identifying threat events above a pipeline according to an embodiment of the present invention.
具体实施方式Detailed ways
以下结合附图对本发明实施例的具体实施方式进行详细说明。应当理解的是，此处所描述的具体实施方式仅用于说明和解释本发明实施例，并不用于限制本发明实施例。The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.
需要说明，若本申请实施方式中有涉及方向性指示(诸如上、下、左、右、前、后……)，则该方向性指示仅用于解释在某一特定姿态(如附图所示)下各部件之间的相对位置关系、运动情况等，如果该特定姿态发生改变时，则该方向性指示也相应地随之改变。It should be noted that if the implementation methods of the present application involve directional indications (such as up, down, left, right, front, back...), such directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.
另外，若本申请实施方式中有涉及“第一”、“第二”等的描述，则该“第一”、“第二”等的描述仅用于描述目的，而不能理解为指示或暗示其相对重要性或者隐含指明所指示的技术特征的数量。由此，限定有“第一”、“第二”的特征可以明示或者隐含地包括至少一个该特征。另外，各个实施方式之间的技术方案可以相互结合，但是必须是以本领域普通技术人员能够实现为基础，当技术方案的结合出现相互矛盾或无法实现时应当认为这种技术方案的结合不存在，也不在本申请要求的保护范围之内。In addition, if there are descriptions involving "first", "second", etc. in the implementation methods of this application, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various implementation methods can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.
在保证管道的安全运行中，传统的解决思路需要通过人工巡检，或者安装摄像头并利用监控人工识别是否存在管道上方作业(施工)，人工巡检无法做到实时性，无法全周期监控，安装摄像头的成本高，而且偏远地区存在取电困难的问题。In ensuring the safe operation of pipelines, the traditional solution requires manual inspections, or installing cameras and using monitoring to manually identify whether there is any work (construction) above the pipeline. Manual inspections cannot be done in real time or throughout the entire cycle. The cost of installing cameras is high, and there is difficulty in obtaining electricity in remote areas.
图1示意性示出了根据本发明实施例的管道上方威胁事件的识别模型的训练方法的流程图。如图1所示，为了解决上述存在的问题，在本发明一实施例中，提供了一种管道上方威胁事件的识别模型的训练方法，包括以下步骤：Fig. 1 schematically shows a flow chart of a method for training a recognition model of a threat event above a pipeline according to an embodiment of the present invention. As shown in Fig. 1, in order to solve the above-mentioned problems, in one embodiment of the present invention, a method for training a recognition model of a threat event above a pipeline is provided, comprising the following steps:
步骤101，利用设置于管道上方的光纤获取第一振动数据；Step 101, obtaining first vibration data using an optical fiber disposed above a pipeline;
步骤102，根据第一振动数据生成对应的第一相位数据和第一强度数据；Step 102, generating corresponding first phase data and first intensity data according to the first vibration data;
步骤103，根据第一相位数据和第一强度数据得到第一特征数据，其中第一特征数据包括第一瀑布图图像的颜色变化趋势的数据，第一瀑布图图像是基于第一振动数据生成的；Step 103, obtaining first characteristic data according to the first phase data and the first intensity data, wherein the first characteristic data includes data of a color change trend of a first waterfall image, and the first waterfall image is generated based on the first vibration data;
步骤104，利用第一特征数据构建样本集；Step 104, constructing a sample set using the first feature data;
步骤105，根据样本集对深度神经网络分类模型进行训练，得到管道上方威胁事件的识别模型。Step 105: Train the deep neural network classification model according to the sample set to obtain a recognition model for threat events above the pipeline.
在步骤101中，利用设置于管道上方的光纤为介质，采集第一振动数据(或者称为光纤传感数据)，抗干扰性强，不受天气亮度等因素影响，使得管道上方威胁事件的识别模型的识别准确度较高。在步骤102和步骤103中，可以将光纤传感数据理解为原始未处理数据，利用收集到的原始未处理数据，生成第一相位数据，再转换为第一强度数据，再定位信号的相位，截取得到第一特征数据。具体地，通过定位第一强度数据中定位数据的位置，通过滑动窗口的方法定位截取第一相位数据，最后得到第一特征数据。In step 101, the optical fiber disposed above the pipeline is used as a medium to collect the first vibration data (or optical fiber sensing data), which has strong anti-interference performance and is not affected by factors such as weather brightness, so that the recognition accuracy of the recognition model of the threat event above the pipeline is relatively high. In steps 102 and 103, the optical fiber sensing data can be understood as raw unprocessed data, and the collected raw unprocessed data is used to generate the first phase data, which is then converted into the first intensity data, and the phase of the signal is then located to intercept the first characteristic data. Specifically, by locating the position of the positioning data in the first intensity data, the first phase data is located and intercepted by the sliding window method, and finally the first characteristic data is obtained.
管道上方发生威胁事件(例如在管道上方施工)与管道上方未发生威胁事件(例如非管道上方施工)对应的瀑布图图像的颜色变化趋势是不一样的。由于第一特征数据包括第一瀑布图图像的颜色变化趋势的数据，第一瀑布图图像是基于第一振动数据生成的，这样深度神经网络分类模型可以学习管道上方发生威胁事件时(例如施工)瀑布图图像的特征，然后基于管道上方威胁事件的识别模型，可以实现利用瀑布图图像的颜色变化趋势判定是否在管道上方作业(施工)，实时检测管道上方是否存在施工作业，及时识别出管道上方的威胁事件，实现全周期监控并保证检测的实时性，保护管道(例如油气管道)免受破坏，保证管道的安全运行，提高管道的安全性。The color change trend of the waterfall image corresponding to a threat event (e.g., construction above the pipeline) and a threat event (e.g., construction not above the pipeline) above the pipeline is different. Since the first feature data includes data on the color change trend of the first waterfall image, and the first waterfall image is generated based on the first vibration data, the deep neural network classification model can learn the features of the waterfall image when a threat event (e.g., construction) occurs above the pipeline, and then based on the recognition model of the threat event above the pipeline, it is possible to use the color change trend of the waterfall image to determine whether there is work (construction) above the pipeline, detect in real time whether there is construction work above the pipeline, and promptly identify the threat event above the pipeline, realize full-cycle monitoring and ensure the real-time detection, protect the pipeline (e.g., oil and gas pipeline) from damage, ensure the safe operation of the pipeline, and improve the safety of the pipeline.
在一实施例中，样本集包括训练集和验证集，利用第一特征数据构建样本集包括：将第一特征数据分成训练集和验证集；根据样本集对深度神经网络分类模型进行训练，得到管道上方威胁事件的识别模型包括：初始化深度神经网络分类模型的权重梯度；将训练集分批次输入至深度神经网络分类模型；根据深度神经网络分类模型的设计层数，逐层进行前向传播到输出层；根据损失函数确定输出层的输出结果与真实标签的损失值；基于损失值进行反向传播，并反向链式地逐层更新权重梯度；在验证集验证深度神经网络分类模型的准确率大于预设阈值的情况下，得到管道上方威胁事件的识别模型。In one embodiment, the sample set includes a training set and a validation set, and constructing the sample set using the first feature data includes: dividing the first feature data into a training set and a validation set; training a deep neural network classification model according to the sample set, and obtaining an identification model for threat events above the pipeline includes: initializing the weight gradient of the deep neural network classification model; inputting the training set into the deep neural network classification model in batches; forward propagating to the output layer layer by layer according to the designed number of layers of the deep neural network classification model; determining the loss value between the output result of the output layer and the true label according to the loss function; performing back propagation based on the loss value, and updating the weight gradient layer by layer in a reverse chain; when the accuracy of the deep neural network classification model verified by the validation set is greater than a preset threshold, the identification model for threat events above the pipeline is obtained.
在训练深度神经网络分类模型时，保存验证集上准确率提升的模型，最终得到准确率达标的管道上方威胁事件的识别模型。图2示意性示出了根据本发明实施例的深度神经网络分类模型的示意图，可参见图2，每轮训练开始后，将训练集分批次送入深度神经网络分类模型中，验证集用于训练时每轮去验证模型的准确率。When training the deep neural network classification model, the model with improved accuracy on the validation set is saved, and finally a recognition model of threat events above the pipeline with an accuracy rate that meets the standard is obtained. FIG2 schematically shows a schematic diagram of a deep neural network classification model according to an embodiment of the present invention. Referring to FIG2 , after each round of training starts, the training set is sent to the deep neural network classification model in batches, and the validation set is used to verify the accuracy of the model in each round of training.
在本发明实施例中，可以利用光纤传感数据(即第一振动数据)生成瀑布图图像。管道上方发生威胁事件(例如在管道上方施工)与管道上方未发生威胁事件(例如非管道上方施工)对应的瀑布图图像的颜色变化趋势是不一样的，瀑布图图像的作用范围和振动强弱信息也是不一样的，深度神经网络分类模型学习管道上方发生威胁事件时瀑布图图像的特征，然后推理判断瀑布图图像的差异(异常度)来识别是否在管道上方施工作业。In an embodiment of the present invention, a waterfall image can be generated using optical fiber sensing data (i.e., first vibration data). The color change trend of the waterfall image corresponding to a threatening event occurring above the pipeline (e.g., construction above the pipeline) and a threatening event not occurring above the pipeline (e.g., construction not occurring above the pipeline) is different, and the scope of action and vibration intensity information of the waterfall image are also different. The deep neural network classification model learns the features of the waterfall image when a threatening event occurs above the pipeline, and then infers and judges the difference (abnormality) of the waterfall image to identify whether construction is being performed above the pipeline.
在监测的整个管道，根据采样周期连续在光纤的模拟信号上取值，根据每次取值信号强度差异化的统计结果，得到一个以纵轴为时间轴，横轴为距离，距离的起始点为管道开始点，至整个管道结束，也就是得到了以振动光纤长度为距离的二维图形。此时实时监测的信号会变成一个二维数据，形成了以Y轴为时间轴(采样周期)，在时间上是连续的，X轴为距离，起始点为管道开始点，至整个管道结束，也就是得到了振动光纤的距离的二维图形，随着连续不断的采样，事件的行为最终通过图像形成了一张张瀑布图图像。In the entire pipeline being monitored, the analog signal of the optical fiber is continuously sampled according to the sampling period. Based on the statistical results of the signal strength differentiation of each sample, a two-dimensional graph with the vertical axis as the time axis and the horizontal axis as the distance is obtained. The starting point of the distance is the starting point of the pipeline and the end of the entire pipeline is obtained, that is, the distance is obtained. At this time, the real-time monitoring signal will become a two-dimensional data, forming a two-dimensional graph with the Y axis as the time axis (sampling period), which is continuous in time, and the X axis as the distance. The starting point is the starting point of the pipeline and the end of the entire pipeline. That is, the distance of the vibrating optical fiber is obtained. With continuous sampling, the behavior of the event is finally formed into waterfall images through images.
根据光纤传感数据绘制瀑布图图像，可以感知管道周围环境信息。具体地，收集光纤传感瑞利散射回来的信号数据，去除干扰信号，进行图像增强等处理，然后利用滤波后的光纤传感数据绘制瀑布图图像，从瀑布图图像中感知管道周围环境信息。需要说明的是，在图像增强和图像滤波等处理过程中，图像的形状没有改变，依然是二维矩阵。By drawing a waterfall image based on the fiber optic sensor data, the environment information around the pipeline can be sensed. Specifically, the signal data sent back by the fiber optic sensor Rayleigh scattering is collected, the interference signal is removed, and image enhancement and other processing are performed. Then, the waterfall image is drawn using the filtered fiber optic sensor data, and the environment information around the pipeline is sensed from the waterfall image. It should be noted that during the image enhancement and image filtering processes, the shape of the image does not change and is still a two-dimensional matrix.
图3示意性示出了根据本发明实施例的非管道上方作业的瀑布图图像的示意图；图4示意性示出了根据本发明实施例的在管道上方作业的瀑布图图像的示意图之一；图5示意性示出了根据本发明实施例的在管道上方作业的瀑布图图像的示意图之二。基于土壤介质传导的原理，使瀑布图图像颜色有变化的趋势，从而判定出现的施工位置是否在管道(正)上方。在一实施例中，在管道上方施工时(即在管道上方作业的威胁事件)，作用范围大，振动强度非常强，瀑布图图像上的背景部分散杂点较多。在一实施例中，非管道上方施工时(即非威胁事件)，振动通过土壤传播，土壤传播会导致信号减弱，作用范围变小，瀑布图图像上显示的背景部分干净，施工信号相比也会减弱。其中作用范围对应的就是距离维度，通过距离维度能够判定其对应的监控区间范围。FIG3 schematically shows a schematic diagram of a waterfall image of a non-pipeline operation according to an embodiment of the present invention; FIG4 schematically shows one of the schematic diagrams of a waterfall image of an operation above a pipeline according to an embodiment of the present invention; and FIG5 schematically shows one of the schematic diagrams of a waterfall image of an operation above a pipeline according to an embodiment of the present invention. Based on the principle of soil medium conduction, the color of the waterfall image has a tendency to change, so as to determine whether the construction position is (directly) above the pipeline. In one embodiment, when construction is performed above the pipeline (i.e., a threatening event of operation above the pipeline), the scope of action is large, the vibration intensity is very strong, and there are many scattered points in the background part of the waterfall image. In one embodiment, when construction is performed above the pipeline (i.e., a non-threatening event), the vibration is transmitted through the soil, and the soil transmission will cause the signal to be weakened, the scope of action becomes smaller, the background part displayed on the waterfall image is clean, and the construction signal will be weakened by comparison. The scope of action corresponds to the distance dimension, and the corresponding monitoring interval range can be determined by the distance dimension.
需要说明的是，原始的瀑布图图像为彩色图像，主要有深蓝色、浅蓝色、红色和白色，瀑布图图像上点坐标的值代表振动强度，深蓝色代表振动强度较小或者没有，浅蓝色或者偏白色代表振动强度较大。即瀑布图图像上颜色的深浅可以代表振动强度，当颜色从蓝色到红色时，越红代表振动强度越高，越蓝代表振动强度越低。It should be noted that the original waterfall image is a color image, mainly dark blue, light blue, red and white. The coordinate values of the points on the waterfall image represent the vibration intensity. Dark blue represents a small or no vibration intensity, and light blue or off-white represents a large vibration intensity. That is, the depth of the color on the waterfall image can represent the vibration intensity. When the color ranges from blue to red, the redder the color, the higher the vibration intensity, and the bluer the color, the lower the vibration intensity.
图6示意性示出了根据本发明实施例的管道上方威胁事件的识别方法的流程图。如图6所示，在本发明一实施例中，提供了一种管道上方威胁事件的识别方法，包括以下步骤：FIG6 schematically shows a flow chart of a method for identifying a threat event above a pipeline according to an embodiment of the present invention. As shown in FIG6 , in one embodiment of the present invention, a method for identifying a threat event above a pipeline is provided, comprising the following steps:
步骤601，利用设置于管道上方的光纤获取第二振动数据；Step 601, obtaining second vibration data using an optical fiber disposed above the pipeline;
步骤602，根据第二振动数据生成对应的第二相位数据和第二强度数据；Step 602, generating corresponding second phase data and second intensity data according to the second vibration data;
步骤603，根据第二相位数据和第二强度数据得到第二特征数据，其中第二特征数据包括第二瀑布图图像的颜色变化趋势的数据，第二瀑布图图像是基于第二振动数据生成的；Step 603, obtaining second characteristic data according to the second phase data and the second intensity data, wherein the second characteristic data includes data on a color change trend of a second waterfall image, and the second waterfall image is generated based on the second vibration data;
步骤604，将第二特征数据输入至管道上方威胁事件的识别模型，得到管道上方威胁事件的识别结果，其中识别模型通过上述任意实施例的管道上方威胁事件的识别模型的训练方法得到。Step 604: input the second feature data into a recognition model of threat events above the pipeline to obtain a recognition result of the threat events above the pipeline, wherein the recognition model is obtained by the training method of the recognition model of threat events above the pipeline in any of the above embodiments.
第一振动数据和第二振动数据的区别是，第一振动数据可以理解是用于训练深度神经网络分类模型的数据，第二振动数据是在深度神经网络分类模型训练好之后，即得到管道上方威胁事件的识别模型之后，实时收集到的光纤传感数据。管道上方发生威胁事件(例如在管道上方施工)与管道上方未发生威胁事件(例如非管道上方施工)对应的瀑布图图像的颜色变化趋势是不一样的。由于第二特征数据包括第二瀑布图图像的颜色变化趋势的数据，第二瀑布图图像是基于第二振动数据生成的，这样将第二特征数据输入至管道上方威胁事件的识别模型，得到管道上方威胁事件的识别结果，就可以实现利用瀑布图图像的颜色变化趋势判定是否在管道(正)上方作业，实时检测管道上方是否存在施工作业，及时识别出管道上方的威胁事件，实现全周期监控并保证检测的实时性，保护管道(例如油气管道)免受破坏，保证管道的安全运行，提高管道的安全性。The difference between the first vibration data and the second vibration data is that the first vibration data can be understood as data used to train the deep neural network classification model, and the second vibration data is the optical fiber sensing data collected in real time after the deep neural network classification model is trained, that is, after the recognition model of the threat event above the pipeline is obtained. The color change trend of the waterfall image corresponding to the threat event above the pipeline (for example, construction above the pipeline) and the threat event above the pipeline (for example, construction not above the pipeline) is different. Since the second feature data includes the data of the color change trend of the second waterfall image, the second waterfall image is generated based on the second vibration data, so the second feature data is input into the recognition model of the threat event above the pipeline to obtain the recognition result of the threat event above the pipeline, and it is possible to use the color change trend of the waterfall image to determine whether the operation is (positively) above the pipeline, detect whether there is construction work above the pipeline in real time, and timely identify the threat event above the pipeline, so as to achieve full-cycle monitoring and ensure the real-time detection, protect the pipeline (for example, oil and gas pipeline) from damage, ensure the safe operation of the pipeline, and improve the safety of the pipeline.
本发明实施例提供一种管道上方威胁事件的识别模型的训练装置，包括：An embodiment of the present invention provides a training device for a recognition model of threat events above a pipeline, comprising:
第一获取模块，用于利用设置于管道上方的光纤获取第一振动数据；A first acquisition module, used for acquiring first vibration data by using an optical fiber arranged above the pipeline;
第一生成模块，用于根据第一振动数据生成对应的第一相位数据和第一强度数据；A first generating module, used for generating corresponding first phase data and first intensity data according to the first vibration data;
第一得到模块，用于根据第一相位数据和第一强度数据得到第一特征数据，其中第一特征数据包括第一瀑布图图像的颜色变化趋势的数据，第一瀑布图图像是基于第一振动数据生成的；A first obtaining module, configured to obtain first characteristic data according to the first phase data and the first intensity data, wherein the first characteristic data includes data of a color change trend of a first waterfall chart image, and the first waterfall chart image is generated based on the first vibration data;
构建模块，用于利用第一特征数据构建样本集；A construction module, used to construct a sample set using the first feature data;
训练模块，用于根据样本集对深度神经网络分类模型进行训练，得到管道上方威胁事件的识别模型。The training module is used to train the deep neural network classification model according to the sample set to obtain an identification model for threat events above the pipeline.
在本发明实施例中，样本集包括训练集和验证集，构建模块包括：In the embodiment of the present invention, the sample set includes a training set and a validation set, and the construction module includes:
分成单元，用于将第一特征数据分成训练集和验证集；A division unit is used to divide the first feature data into a training set and a validation set;
训练模块包括：The training modules include:
初始化单元，用于初始化深度神经网络分类模型的权重梯度；Initialization unit, used to initialize the weight gradient of the deep neural network classification model;
输入单元，用于将训练集分批次输入至深度神经网络分类模型；An input unit, used to input the training set into the deep neural network classification model in batches;
传播单元，用于根据深度神经网络分类模型的设计层数，逐层进行前向传播到输出层；The propagation unit is used to perform forward propagation to the output layer layer by layer according to the designed number of layers of the deep neural network classification model;
确定单元，用于根据损失函数确定所述输出层的输出结果与真实标签的损失值；A determination unit, used to determine a loss value between an output result of the output layer and a true label according to a loss function;
更新单元，用于基于损失值进行反向传播，并反向链式地逐层更新权重梯度；The update unit is used to perform back propagation based on the loss value and update the weight gradient layer by layer in a reverse chain manner;
验证单元，用于在验证集验证深度神经网络分类模型的准确率大于预设阈值的情况下，得到管道上方威胁事件的识别模型。The verification unit is used to obtain an identification model of threat events above the pipeline when the accuracy of the deep neural network classification model verified by the verification set is greater than a preset threshold.
本发明实施例提供一种管道上方威胁事件的识别装置，包括：An embodiment of the present invention provides a device for identifying threat events above a pipeline, comprising:
第二获取模块，用于利用设置于管道上方的光纤获取第二振动数据；A second acquisition module, used for acquiring second vibration data by using an optical fiber arranged above the pipeline;
第二生成模块，用于根据第二振动数据生成对应的第二相位数据和第二强度数据；A second generating module, used for generating corresponding second phase data and second intensity data according to the second vibration data;
第二得到模块，用于根据第二相位数据和第二强度数据得到第二特征数据，其中第二特征数据包括第二瀑布图图像的颜色变化趋势的数据，第二瀑布图图像是基于第二振动数据生成的；A second obtaining module is used to obtain second characteristic data according to the second phase data and the second intensity data, wherein the second characteristic data includes data of a color change trend of a second waterfall chart image, and the second waterfall chart image is generated based on the second vibration data;
输入模块，用于将第二特征数据输入至管道上方威胁事件的识别模型，得到管道上方威胁事件的识别结果，其中识别模型通过如上述的管道上方威胁事件的识别模型的训练方法得到。The input module is used to input the second feature data into the recognition model of the threat event above the pipeline to obtain the recognition result of the threat event above the pipeline, wherein the recognition model is obtained by the training method of the recognition model of the threat event above the pipeline as described above.
在本发明实施例中，第二获取模块包括：In an embodiment of the present invention, the second acquisition module includes:
获取单元，用于利用设置于管道上方的光纤获取光纤传感瑞利散射回的第二振动数据。The acquisition unit is used to acquire the second vibration data sent back by the optical fiber sensing Rayleigh scattering by using the optical fiber arranged above the pipeline.
本发明实施例提供一种计算机设备，包括存储器以及处理器，存储器存储有计算机程序，计算机程序在处理器上运行时执行上述的管道上方威胁事件的识别装置的训练方法，或上述的管道上方威胁事件的识别方法。An embodiment of the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program runs on the processor, it executes the training method of the above-mentioned device for identifying threat events above a pipeline, or the above-mentioned method for identifying threat events above a pipeline.
本发明实施例提供一种机器可读存储介质，该机器可读存储介质上存储有指令，该指令用于使得机器执行根据上述的管道上方威胁事件的识别装置的训练方法，或上述的管道上方威胁事件的识别方法。An embodiment of the present invention provides a machine-readable storage medium having instructions stored thereon, the instructions being used to enable a machine to execute the training method of the apparatus for identifying threat events above a pipeline, or the method for identifying threat events above a pipeline.
本领域内的技术人员应明白，本申请的实施例可提供为方法、系统、或计算机程序产品。因此，本申请可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本申请可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.
本申请是参照根据本申请实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。The present application is described with reference to the flowchart and/or block diagram of the method, device (system) and computer program product according to the embodiment of the present application. It should be understood that each process and/or box in the flowchart and/or block diagram, and the combination of the process and/or box in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for realizing the function specified in one process or multiple processes in the flowchart and/or one box or multiple boxes in the block diagram.
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device that implements the functions specified in one or more processes in the flowchart and/or one or more boxes in the block diagram.
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。These computer program instructions may also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and/or one or more boxes in the block diagram.
在一个典型的配置中，计算设备包括一个或多个处理器(CPU)、输入/输出接口、网络接口和内存。In a typical configuration, a computing device includes one or more processors (CPU), input/output interfaces, network interfaces, and memory.
存储器可能包括计算机可读介质中的非永久性存储器，随机存取存储器(RAM)和/或非易失性内存等形式，如只读存储器(ROM)或闪存(flash RAM)。存储器是计算机可读介质的示例。The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and/or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.
计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带磁磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。按照本文中的界定，计算机可读介质不包括暂存电脑可读媒体(transitory media)，如调制的数据信号和载波。Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
还需要说明的是，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、商品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、商品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括要素的过程、方法、商品或者设备中还存在另外的相同要素。It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
以上仅为本申请的实施例而已，并不用于限制本申请。对于本领域技术人员来说，本申请可以有各种更改和变化。凡在本申请的精神和原理之内所作的任何修改、等同替换、改进等，均应包含在本申请的权利要求范围之内。The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310083227.6ACN118196465A (en) | 2023-01-17 | 2023-01-17 | Training method and identification method of recognition model for threat events above pipeline |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310083227.6ACN118196465A (en) | 2023-01-17 | 2023-01-17 | Training method and identification method of recognition model for threat events above pipeline |
| Publication Number | Publication Date |
|---|---|
| CN118196465Atrue CN118196465A (en) | 2024-06-14 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310083227.6APendingCN118196465A (en) | 2023-01-17 | 2023-01-17 | Training method and identification method of recognition model for threat events above pipeline |
| Country | Link |
|---|---|
| CN (1) | CN118196465A (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105931402A (en)* | 2016-06-27 | 2016-09-07 | 上海波汇科技股份有限公司 | Optical fiber perimeter intrusion monitoring method based on image recognition |
| CN111147133A (en)* | 2019-12-24 | 2020-05-12 | 武汉理工光科股份有限公司 | Real-time monitoring system and method for traffic flow based on phi-OTDR |
| CN112703457A (en)* | 2018-05-07 | 2021-04-23 | 强力物联网投资组合2016有限公司 | Method and system for data collection, learning and machine signal streaming for analysis and maintenance using industrial internet of things |
| CN112883802A (en)* | 2021-01-20 | 2021-06-01 | 刘信 | Method for identifying destructive event of pipeline optical fiber vibration safety early warning system |
| CN113531399A (en)* | 2020-04-16 | 2021-10-22 | 中国石油天然气股份有限公司 | Pipeline monitoring method, pipeline monitoring device, computer equipment and storage medium |
| CN113743528A (en)* | 2021-09-15 | 2021-12-03 | 集美大学 | Voltage source inverter fault diagnosis method and system based on convolutional neural network |
| CN114742096A (en)* | 2022-03-22 | 2022-07-12 | 上海微波技术研究所（中国电子科技集团公司第五十研究所） | Intrusion alarm method and system based on vibration optical fiber detection and complete action extraction |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105931402A (en)* | 2016-06-27 | 2016-09-07 | 上海波汇科技股份有限公司 | Optical fiber perimeter intrusion monitoring method based on image recognition |
| CN112703457A (en)* | 2018-05-07 | 2021-04-23 | 强力物联网投资组合2016有限公司 | Method and system for data collection, learning and machine signal streaming for analysis and maintenance using industrial internet of things |
| CN111147133A (en)* | 2019-12-24 | 2020-05-12 | 武汉理工光科股份有限公司 | Real-time monitoring system and method for traffic flow based on phi-OTDR |
| CN113531399A (en)* | 2020-04-16 | 2021-10-22 | 中国石油天然气股份有限公司 | Pipeline monitoring method, pipeline monitoring device, computer equipment and storage medium |
| CN112883802A (en)* | 2021-01-20 | 2021-06-01 | 刘信 | Method for identifying destructive event of pipeline optical fiber vibration safety early warning system |
| CN113743528A (en)* | 2021-09-15 | 2021-12-03 | 集美大学 | Voltage source inverter fault diagnosis method and system based on convolutional neural network |
| CN114742096A (en)* | 2022-03-22 | 2022-07-12 | 上海微波技术研究所（中国电子科技集团公司第五十研究所） | Intrusion alarm method and system based on vibration optical fiber detection and complete action extraction |
| Title |
|---|
| 张利波等: "物联网技术在输气站场建设中的应用展望", 油气田地面工程, vol. 36, no. 6, 30 June 2017 (2017-06-30)* |
| Publication | Publication Date | Title |
|---|---|---|
| Gupta et al. | A selective literature review on leak management techniques for water distribution system | |
| Zhou et al. | A pipeline leak detection and localization approach based on ensemble TL1DCNN | |
| Wu et al. | Efficient delineation of nested depression hierarchy in digital elevation models for hydrological analysis using level‐set method | |
| Yussof et al. | Review of water leak detection methods in smart building applications | |
| KR102413399B1 (en) | Leak diagnosis system for offshore plant pipelines based on machine learning | |
| NL2025790B1 (en) | Intelligent system and method for inspecting offshore oil and gas pipelines | |
| Wiseman | Automatic alert system for worn out pipes in autonomous vehicles | |
| CN113781513A (en) | Method and system for detecting leakage of water supply pipeline of power plant | |
| CN114898277A (en) | Method, apparatus, storage medium, and processor for identifying pipeline safety risks | |
| CN114693114A (en) | Monitoring method and device for underground space structure, computer equipment and storage medium | |
| CN118675321A (en) | DAS traffic parameter real-time estimation method, system, equipment and medium | |
| CN116644963A (en) | Pipeline safety supervision and assessment method, system and readable storage medium | |
| Guan et al. | A deep learning approach for construction vehicles fill factor estimation and bucket detection in extreme environments | |
| JP2022552498A (en) | Leak detection method | |
| CN118196465A (en) | Training method and identification method of recognition model for threat events above pipeline | |
| von Benzon et al. | Mapping damages from inspection images to 3D digital twins of large‐scale structures | |
| Kim et al. | A leak detection and 3D source localization method on a plant piping system by using multiple cameras | |
| CN111079766A (en) | Intelligent method for P & ID (peer-to-peer) graph | |
| Ferreira et al. | Corrosion-like defect severity estimation in pipelines using convolutional neural networks | |
| Fathi et al. | High-Precision Single-Leak Detection and Localization in Single-Phase Liquid Pipelines Using the Negative Pressure Wave Technique: An Application in a Real-Field Case Study | |
| US11726071B2 (en) | Finding a contaminant source in a volume of flowing fluid | |
| CN118211025A (en) | Method, device and processor for determining environmental information near a pipeline | |
| Rahmes et al. | Continuous environmentally efficient pipeline leak detection | |
| Chen et al. | Pipe Flaws Detection by Using the Mindstorm Robot | |
| CN118193938A (en) | Manual excavation construction identification method, device, processor and readable storage medium |
| 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 |