技术领域Technical Field
本发明涉及车辆维护管理领域，特别是涉及一种ATV全地形车整车维护管理系统及方法。The present invention relates to the field of vehicle maintenance management, and in particular to an ATV all-terrain vehicle whole vehicle maintenance management system and method.
背景技术Background Art
全地形车(All TerrainVehicle，ATV)在越野和各种地形中的使用已经变得越来越普遍。然而，由于ATV在极端条件下的使用，其各个部件，特别是底盘轴、轮轴、驱动轴和传动轴等关键部位，容易受到高频振动的影响。这些振动可能源自不平整的地形、高速行驶时的震动等因素，对ATV的整体性能和寿命产生负面影响。在目前的技术状况下，虽然有一些振动监测系统，但大多数侧重于车辆的引擎或其他传统传感器监测的数据，而缺乏对底盘关键部位振动的全面监测。这使得对底盘结构健康状况的实时监测和准确分析变得困难。此外，传统的振动监测系统可能无法提供足够的时频分析和振动特征频率检测，因此难以精确诊断底盘部件的异常情况。The use of All Terrain Vehicles (ATVs) in off-road and various terrains has become increasingly common. However, due to the use of ATVs in extreme conditions, its various components, especially key parts such as chassis shafts, wheel axles, drive shafts and transmission shafts, are susceptible to high-frequency vibrations. These vibrations may originate from factors such as uneven terrain and vibrations during high-speed driving, which have a negative impact on the overall performance and life of the ATV. Under the current state of technology, although there are some vibration monitoring systems, most of them focus on the data monitored by the vehicle's engine or other traditional sensors, and lack comprehensive monitoring of the vibration of key parts of the chassis. This makes it difficult to monitor and accurately analyze the health of the chassis structure in real time. In addition, traditional vibration monitoring systems may not provide sufficient time-frequency analysis and vibration characteristic frequency detection, making it difficult to accurately diagnose abnormal conditions of chassis components.
发明内容Summary of the invention
为解决现有技术存在的上述问题，本发明提供了一种ATV全地形车整车维护管理系统及方法。In order to solve the above problems existing in the prior art, the present invention provides an ATV all-terrain vehicle maintenance management system and method.
为实现上述目的，本发明提供了如下方案：To achieve the above object, the present invention provides the following solutions:
一种ATV全地形车整车维护管理系统，包括：An ATV all-terrain vehicle maintenance management system, comprising:
振动信号采集模块，用于以设定采样频率获取ATV全地形车的原始振动信号；A vibration signal acquisition module is used to acquire the original vibration signal of the ATV all-terrain vehicle at a set sampling frequency;
A/D转换模块，与所述振动信号采集模块连接，用于将所述原始振动信号转化为数字振动数据；An A/D conversion module, connected to the vibration signal acquisition module, for converting the original vibration signal into digital vibration data;
车辆控制器，与所述A/D转换模块连接，用于将所述数字振动数据进行无线传输；a vehicle controller, connected to the A/D conversion module, for wirelessly transmitting the digital vibration data;
服务器，与所述车辆控制器无线连接，用于对所述数字振动数据进行处理，并基于处理后的数字振动数据确定振动特征频率，用于基于所述振动特征频率生成底盘部件检测结果；a server, wirelessly connected to the vehicle controller, for processing the digital vibration data, determining a vibration characteristic frequency based on the processed digital vibration data, and generating a chassis component detection result based on the vibration characteristic frequency;
当所述底盘部件检测结果存在异常时，所述服务器还用于生成维护提醒和通知。When the chassis component detection result is abnormal, the server is also used to generate a maintenance reminder and notification.
可选地，所述振动信号采集模块包括：Optionally, the vibration signal acquisition module includes:
第一传动轴振动传感器，用于采集ATV全地形车传动轴的连接点或距ATV全地形车传动轴的连接点设定范围内的振动信号；A first transmission shaft vibration sensor is used to collect vibration signals at a connection point of the ATV all-terrain vehicle transmission shaft or within a set range from the connection point of the ATV all-terrain vehicle transmission shaft;
第二传动轴振动传感器，用于采集ATV全地形车底盘轮轴连接点位置处的振动信号；The second transmission shaft vibration sensor is used to collect vibration signals at the connection point of the ATV chassis wheel axle;
第一驱动轴振动传感器和第二驱动轴振动传感器，均用于采集ATV全地形车中传动轴、差速器或驱动轴结合位置处的振动信号。The first drive shaft vibration sensor and the second drive shaft vibration sensor are both used to collect vibration signals at the connection position of the transmission shaft, differential or drive shaft in the ATV all-terrain vehicle.
可选地，所述A/D转换模块采用CAN协议传输与所述车辆控制器进行数据通讯。Optionally, the A/D conversion module uses CAN protocol transmission to communicate data with the vehicle controller.
可选地，所述服务器包括：Optionally, the server includes:
数据处理和降噪单元，与所述车辆控制器无线连接，用于对所述数字振动数据进行滤波处理，用于采用滑动平均法对滤波处理后的数字振动数据进行平滑处理，得到平滑信号；A data processing and noise reduction unit, wirelessly connected to the vehicle controller, for filtering the digital vibration data, and for smoothing the filtered digital vibration data using a sliding average method to obtain a smoothed signal;
时频变换单元，与所述数据处理和降噪单元连接，用于将对所述平滑信号进行快速傅里叶变换，得到频域信号；A time-frequency transformation unit, connected to the data processing and noise reduction unit, for performing a fast Fourier transform on the smoothed signal to obtain a frequency domain signal;
振动特征频率计算单元，与所述时频变换单元连接，用于基于所述频域信号确定振动特征频率；a vibration characteristic frequency calculation unit, connected to the time-frequency conversion unit, and configured to determine the vibration characteristic frequency based on the frequency domain signal;
检测单元，与所述振动特征频率计算单元连接，用于基于所述振动特征频率生成底盘部件检测结果。The detection unit is connected to the vibration characteristic frequency calculation unit and is used to generate a chassis component detection result based on the vibration characteristic frequency.
可选地，所述数据处理和降噪单元包括：Optionally, the data processing and noise reduction unit includes:
低通滤波器，使用fir1函数建立得到，用于对所述数字振动数据进行滤波处理。The low-pass filter is established using the fir1 function and is used to perform filtering processing on the digital vibration data.
一种ATV全地形车整车维护管理方法，所述方法采用上述提供的ATV全地形车整车维护管理系统实现；所述方法包括：A maintenance management method for an ATV all-terrain vehicle is provided, wherein the method is implemented by using the above-mentioned ATV all-terrain vehicle maintenance management system; the method comprises:
获取原始振动信号；Obtaining the original vibration signal;
将所述原始振动信号转化为数字振动数据；Converting the raw vibration signal into digital vibration data;
对所述数字振动数据进行处理，并基于处理后的数字振动数据确定振动特征频率；processing the digital vibration data and determining a vibration characteristic frequency based on the processed digital vibration data;
基于所述振动特征频率生成底盘部件检测结果；generating chassis component detection results based on the vibration characteristic frequency;
当所述底盘部件检测结果存在异常时，所述服务器还用于生成维护提醒和通知；When the chassis component detection result is abnormal, the server is also used to generate maintenance reminders and notifications;
当所述底盘部件检测结果无异常时，不做处理。When the chassis component detection result is normal, no processing is performed.
可选地，对所述数字振动数据进行处理，并基于处理后的数字振动数据确定振动特征频率，具体包括：Optionally, processing the digital vibration data and determining the vibration characteristic frequency based on the processed digital vibration data specifically includes:
使用fir1函数建立滤波器，并使用汉明窗口作为所述滤波器的窗口函数，对所述数字振动数据进行滤波处理，得到滤波信号；Use the fir1 function to establish a filter, and use a Hamming window as the window function of the filter to filter the digital vibration data to obtain a filtered signal;
采用滑动平均法对所述滤波信号进行平滑处理，得到平滑信号；Smoothing the filtered signal using a sliding average method to obtain a smoothed signal;
对所述平滑信号进行快速傅里叶变换，得到频域信号；Performing fast Fourier transform on the smoothed signal to obtain a frequency domain signal;
基于所述频域信号得到所述振动特征频率。The vibration characteristic frequency is obtained based on the frequency domain signal.
可选地，基于所述振动特征频率生成底盘部件检测结果，具体包括：Optionally, generating chassis component detection results based on the vibration characteristic frequency specifically includes:
基于振动特征频率生成频谱数据；generating spectrum data based on the vibration characteristic frequencies;
采用Butterworth高通滤波器对所述频谱数据进行滤波处理，得到正常状态频谱和异常状态频谱；Using a Butterworth high-pass filter to filter the spectrum data to obtain a normal state spectrum and an abnormal state spectrum;
对正常状态频谱和异常状态频谱进行差异分析，得到差异结果；Perform difference analysis on the normal state spectrum and the abnormal state spectrum to obtain difference results;
当所述差异结果超过设定阈值时，确定所述ATV全地形车存在初步局部异常情况；当所述差异结果不超过设定阈值时，确定所述ATV全地形车无异常；When the difference result exceeds the set threshold, it is determined that the ATV all-terrain vehicle has a preliminary local abnormality; when the difference result does not exceed the set threshold, it is determined that the ATV all-terrain vehicle has no abnormality;
根据正常工作状态下的频谱数据构建的统计模型；A statistical model constructed based on spectrum data under normal working conditions;
检测实际频谱数据与所述统计模型得到的频谱数据间的偏差；Detecting a deviation between actual spectrum data and spectrum data obtained by the statistical model;
判断所述偏差是否超出一定阈值，得到第一判断结果；Determine whether the deviation exceeds a certain threshold value to obtain a first determination result;
当所述第一判断结果为是时，确定ATV全地形车存在存在同一部件的整体异常情况；When the first judgment result is yes, it is determined that the ATV all-terrain vehicle has an overall abnormality of the same component;
当所述第一判断结果为否时，不做处理。When the first judgment result is no, no processing is performed.
可选地，基于所述振动特征频率生成底盘部件检测结果，具体包括：Optionally, generating chassis component detection results based on the vibration characteristic frequency specifically includes:
基于振动特征频率生成频谱数据；generating spectrum data based on the vibration characteristic frequencies;
采用Butterworth高通滤波器对所述频谱数据进行滤波处理，得到正常状态频谱和异常状态频谱；Using a Butterworth high-pass filter to filter the spectrum data to obtain a normal state spectrum and an abnormal state spectrum;
基于所述异常状态频谱得到异常振动频率；Obtaining an abnormal vibration frequency based on the abnormal state spectrum;
判断单位时间段内异常振动频率出现的次数是否在第一设定阈值范围内，得到第二判断结果；Determine whether the number of occurrences of the abnormal vibration frequency within a unit time period is within a first set threshold range, and obtain a second determination result;
当所述第二判断结果为是，则确定ATV全地形车底盘轮轴连接位置处或传动轴、差速器或驱动轴结合位置处的健康状态为正常状态或轻微告警状态；When the second judgment result is yes, the health status of the ATV chassis wheel axle connection position or the transmission shaft, differential or drive shaft combination position is determined to be a normal state or a slight warning state;
当所述第二判断结果为否，则判断单位时间段内异常振动频率出现的次数是否在第二设定阈值范围内，得到第三判断结果；When the second judgment result is no, it is determined whether the number of occurrences of the abnormal vibration frequency in the unit time period is within the second set threshold range to obtain a third judgment result;
当所述第三判断结果为是时，则确定ATV全地形车底盘轮轴连接位置处或传动轴、差速器或驱动轴结合位置处的健康状态的为中度告警状态；When the third judgment result is yes, it is determined that the health status of the ATV chassis wheel axle connection position or the transmission shaft, differential or drive shaft combination position is a moderate warning state;
当所述第三判断结果为否时，则判断单位时间段内异常振动频率出现的次数是否在第三设定阈值范围内，得到第四判断结果；When the third judgment result is no, it is judged whether the number of occurrences of the abnormal vibration frequency in the unit time period is within the third set threshold range to obtain a fourth judgment result;
当所述第四判断结果为是时，则确定ATV全地形车底盘轮轴连接位置处或传动轴、差速器或驱动轴结合位置处的健康状态为严重告警状态。When the fourth judgment result is yes, it is determined that the health status of the ATV all-terrain vehicle chassis wheel axle connection position or the transmission shaft, differential or drive shaft combination position is a serious warning state.
可选地，基于所述振动特征频率生成底盘部件检测结果，具体包括：Optionally, generating chassis component detection results based on the vibration characteristic frequency specifically includes:
使用预训练分类模型进行特征提取，使用SUOD集成异常检测加速器集成多种异常检测加权方法；Use pre-trained classification models for feature extraction and use the SUOD integrated anomaly detection accelerator to integrate multiple anomaly detection weighting methods;
采用每种异常检测方法基于历史振动数据得到异常分数，并对得到的异常分数进行加权计算，获得最终的综合异常分数；Each anomaly detection method is used to obtain an anomaly score based on historical vibration data, and the obtained anomaly scores are weighted to obtain the final comprehensive anomaly score;
对所述综合异常分数的历史振动数据进行人工标定；Manually calibrating the historical vibration data of the comprehensive anomaly score;
将标定过的数据作为训练数据，并对所述训练数据进行归一化处理；Using the calibrated data as training data, and performing normalization processing on the training data;
使用归一化处理后的训练数据对Anomaly Transformer模型进行训练，得到训练好的Anomaly Transformer模型；Use the normalized training data to train the Anomaly Transformer model to obtain a trained Anomaly Transformer model;
通过convert_tflite_model函数将训练好的Anomaly Transformer模型转换为TensorFlow Lite模型；Convert the trained Anomaly Transformer model to a TensorFlow Lite model using the convert_tflite_model function;
将所述振动特征频率转换成张量的形式输入所述TensorFlow Lite模型得到所述底盘部件检测结果。The vibration characteristic frequency is converted into a tensor form and input into the TensorFlow Lite model to obtain the chassis component detection result.
根据本发明提供的具体实施例，本发明公开了以下技术效果：According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:
本发明通过设置振动信号采集模块，可以实现全面监测和实时检测底盘关键连接处的振动情况，提高了维护的及时性。服务器中通过异常频率和特征频率计算，生成详细的故障信息，并通过多渠道通知维护人员。同时，服务器和车辆控制器采用无线连接，可以提供远程故障诊断和数据记录的功能，使得维护人员能够远程获取车辆振动数据，降低了维护成本，提高了维护效率。通过服务器中采用降噪和准确性提高措施，进一步保障了振动信号的稳定性。综合而言，本发明在全面性、实时性、智能性等方面为整车维护带来了显著的优势，能够精确诊断底盘部件的异常情况。The present invention can realize comprehensive monitoring and real-time detection of vibration conditions at key chassis connections by setting a vibration signal acquisition module, thereby improving the timeliness of maintenance. The server generates detailed fault information by calculating abnormal frequencies and characteristic frequencies, and notifies maintenance personnel through multiple channels. At the same time, the server and vehicle controller are wirelessly connected, which can provide remote fault diagnosis and data recording functions, allowing maintenance personnel to remotely obtain vehicle vibration data, reducing maintenance costs and improving maintenance efficiency. By adopting noise reduction and accuracy improvement measures in the server, the stability of the vibration signal is further guaranteed. In summary, the present invention brings significant advantages to vehicle maintenance in terms of comprehensiveness, real-time performance, intelligence, etc., and can accurately diagnose abnormal conditions of chassis components.
附图说明BRIEF DESCRIPTION OF THE DRAWINGS
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
图1为本发明实施例提供的ATV全地形车整车维护管理系统的安装示意图；FIG1 is a schematic diagram of the installation of an ATV all-terrain vehicle maintenance management system provided by an embodiment of the present invention;
图2为本发明实施例提供的ATV全地形车整车维护管理系统实施流程图；FIG2 is a flowchart of an implementation of an ATV all-terrain vehicle maintenance management system provided by an embodiment of the present invention;
附图标号说明：Description of Figure Numbers:
1-第一传动轴振动传感器，2-第二传动轴振动传感器，3-第一驱动轴振动传感器，4-第二驱动轴振动传感器，5-车辆控制器，6-电机，7-差速器，8-驱动轴，9-传动轴，10-轮轴。1-first transmission shaft vibration sensor, 2-second transmission shaft vibration sensor, 3-first drive shaft vibration sensor, 4-second drive shaft vibration sensor, 5-vehicle controller, 6-motor, 7-differential, 8-drive shaft, 9-transmission shaft, 10-wheel axle.
具体实施方式DETAILED DESCRIPTION
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
本发明的目的是提供一种ATV全地形车整车维护管理系统及方法，旨在能够进行高效的时频分析和振动特征频率计算，从而精确诊断底盘部件的异常情况。The purpose of the present invention is to provide an ATV all-terrain vehicle maintenance management system and method, aiming to be able to perform efficient time-frequency analysis and vibration characteristic frequency calculation, so as to accurately diagnose abnormal conditions of chassis components.
为使本发明的上述目的、特征和优点能够更加明显易懂，下面结合附图和具体实施方式对本发明作进一步详细的说明。In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.
本发明提供了一种ATV全地形车整车维护管理系统，该系统包括：振动信号采集模块、A/D转换模块、车辆控制器5和服务器。The present invention provides an ATV all-terrain vehicle maintenance management system, which includes: a vibration signal acquisition module, an A/D conversion module, a vehicle controller 5 and a server.
振动信号采集模块用于以设定采样频率获取ATV全地形车的原始振动信号。在实际应用过程中，振动信号采集模块可以采用第一传动轴振动传感器1、第二传动轴振动传感器2、第一驱动轴振动传感器3和第二驱动轴振动传感器4。其中，将第一传动轴振动传感器1、第二传动轴振动传感器2、第一驱动轴振动传感器3、第二驱动轴振动传感器4在底盘轴与轴之间的连接处，此处的轴主要包括轮轴10、驱动轴8、传动轴9等关键部位。例如，在车辆底盘轮轴连接点处，安装第二传动轴振动传感器2。在驱动轴8附近的合适位置，在传动轴9、差速器7或驱动轴8结合处，安装第一驱动轴振动传感器3、第二驱动轴振动传感器4。在传动轴9的连接点或附近，安装第一传动轴振动传感器1。在座舱内部位置，安装车辆控制器5。在轮轴10处，安装电机6。在电机6处，安装差速锁(即差速器7)。The vibration signal acquisition module is used to obtain the original vibration signal of the ATV all-terrain vehicle at a set sampling frequency. In the actual application process, the vibration signal acquisition module can use the first transmission shaft vibration sensor 1, the second transmission shaft vibration sensor 2, the first drive shaft vibration sensor 3 and the second drive shaft vibration sensor 4. Among them, the first transmission shaft vibration sensor 1, the second transmission shaft vibration sensor 2, the first drive shaft vibration sensor 3, and the second drive shaft vibration sensor 4 are connected between the chassis shaft and the shaft, and the shaft here mainly includes the key parts such as the wheel shaft 10, the drive shaft 8, and the drive shaft 9. For example, at the connection point of the vehicle chassis wheel shaft, the second drive shaft vibration sensor 2 is installed. At a suitable position near the drive shaft 8, at the junction of the drive shaft 9, the differential 7 or the drive shaft 8, the first drive shaft vibration sensor 3 and the second drive shaft vibration sensor 4 are installed. At or near the connection point of the drive shaft 9, the first drive shaft vibration sensor 1 is installed. The vehicle controller 5 is installed at the internal position of the cabin. At the wheel shaft 10, the motor 6 is installed. At the motor 6, the differential lock (i.e., the differential 7) is installed.
(1)轮轴10：在车辆底盘轮轴10连接点处，安装第二传动轴振动传感器2，使其贴合轮轴10表面或固定在轮轴10附近，确保能捕捉到与轮轴10相关的振动。(1) Axle 10: A second transmission shaft vibration sensor 2 is installed at the connection point of the axle 10 to the vehicle chassis so that it fits the surface of the axle 10 or is fixed near the axle 10 to ensure that vibrations related to the axle 10 can be captured.
(2)驱动轴8：将第一驱动轴振动传感器3、第二驱动轴振动传感器4安装在驱动轴8附近的合适位置，在传动轴9、差速器7或驱动轴8结合处，以监测与驱动系统相关的振动。(2) Drive shaft 8: Install the first drive shaft vibration sensor 3 and the second drive shaft vibration sensor 4 at a suitable position near the drive shaft 8, at the junction of the propeller shaft 9, the differential 7 or the drive shaft 8, to monitor vibrations associated with the drive system.
(3)传动轴9：在传动轴9的连接点或附近，确保第一传动轴振动传感器1直接接触传动轴9或其连接部件，以全面监测与传动系统相关的振动。(3) Transmission shaft 9: At or near the connection point of the transmission shaft 9, ensure that the first transmission shaft vibration sensor 1 is in direct contact with the transmission shaft 9 or its connection parts to comprehensively monitor the vibrations related to the transmission system.
进一步，选择1000Hz采样频率进行振动信号的采集，以确保准确捕捉振动。以实时捕捉和记录振动信号，特别是对于突发的振动变化。Furthermore, a sampling frequency of 1000 Hz is selected for collecting vibration signals to ensure accurate capture of vibrations, so as to capture and record vibration signals in real time, especially for sudden vibration changes.
A/D转换模块与振动信号采集模块连接。A/D转换模块可以位于每个振动传感器的电子控制单元(ECU)。通过对振动传感器采集的模拟振动信号进行高精度的量化处理，该处理过程包括将连续的模拟信号划分成离散的数字值。利用模拟到数字转换技术确保振动信号的精准数字化。之后通过内部线缆，经A/D转换模块获得的数字振动数据被传输到车辆控制器5。在传输过程中，CAN协议被采用，通过该协议，数字振动数据以高效且可靠的方式传递至车辆控制器5，确保车辆控制器5能够及时获取并准确解读振动数据。The A/D conversion module is connected to the vibration signal acquisition module. The A/D conversion module can be located in the electronic control unit (ECU) of each vibration sensor. The analog vibration signal collected by the vibration sensor is subjected to high-precision quantization processing, and the processing process includes dividing the continuous analog signal into discrete digital values. The analog-to-digital conversion technology is used to ensure the accurate digitization of the vibration signal. Afterwards, the digital vibration data obtained by the A/D conversion module is transmitted to the vehicle controller 5 through the internal cable. During the transmission process, the CAN protocol is adopted, through which the digital vibration data is transmitted to the vehicle controller 5 in an efficient and reliable manner, ensuring that the vehicle controller 5 can obtain and accurately interpret the vibration data in a timely manner.
服务器与车辆控制器5无线连接，用于对数字振动数据进行处理，并基于处理后的数字振动数据确定振动特征频率，用于基于振动特征频率生成底盘部件检测结果。当底盘部件检测结果存在异常时，服务器还用于生成维护提醒和通知。The server is wirelessly connected to the vehicle controller 5, and is used to process the digital vibration data, and determine the vibration characteristic frequency based on the processed digital vibration data, and is used to generate a chassis component detection result based on the vibration characteristic frequency. When the chassis component detection result is abnormal, the server is also used to generate a maintenance reminder and notification.
在具体实施过程中，车辆控制器5与服务器间可以采用高效的无线通信协议，如Wi-Fi、4G/5G等，确保数据能够迅速、稳定地传输到云端分析服务器或高性能设备，例如：本地服务器或高性能电脑。三轴振动数据(即采用的原始振动数据)被封装成数据包格式，包括数据头、数据体和校验位等信息。车辆控制器5与服务器之间建立稳定的连接，通过使用心跳包等机制监测连接状态确保数据传输通道的可用性。服务器接收到振动数据后，向控制器发送确认信号确保数据已成功传输，若发现传输错误则请求重新传输。服务器接收到振动数据后将其存储在数据库中，通过定期备份以防止数据丢失并支持后续的历史数据分析。In the specific implementation process, an efficient wireless communication protocol, such as Wi-Fi, 4G/5G, etc., can be used between the vehicle controller 5 and the server to ensure that the data can be quickly and stably transmitted to the cloud analysis server or high-performance equipment, such as a local server or a high-performance computer. The three-axis vibration data (i.e., the original vibration data used) is encapsulated into a data packet format, including information such as a data header, a data body, and a check bit. A stable connection is established between the vehicle controller 5 and the server, and the connection status is monitored by using mechanisms such as heartbeat packets to ensure the availability of the data transmission channel. After receiving the vibration data, the server sends a confirmation signal to the controller to ensure that the data has been successfully transmitted. If a transmission error is found, retransmission is requested. After receiving the vibration data, the server stores it in a database, and backs it up regularly to prevent data loss and support subsequent historical data analysis.
进一步，在具体实施过程中，服务器的功能主要包括以下方面：Furthermore, in the specific implementation process, the functions of the server mainly include the following aspects:
(1)数据处理和降噪：(1) Data processing and noise reduction:
在服务器中用数字滤波器对原始振动信号进行降噪处理以去除高频噪声。选择数字滤波器对原始振动信号进行降噪处理。通过在频域分析原始振动信号，选取低通滤波器以去除高频噪声。频域分析确定需要保留或去除的频率范围，确保只有噪声被过滤而不影响关键的振动信息。基于此，可以设计一个能够在降噪的同时不引入额外相位延迟或失真的滤波器。使用fir1函数建立滤波器，使用汉明窗口作为滤波器的窗口函数，滤波器长度为1024个系数，截止频率为48Hz，通过将设计好的数字滤波器应用于采集到的原始振动信号，实现实时的降噪效果。In the server, a digital filter is used to perform noise reduction on the original vibration signal to remove high-frequency noise. A digital filter is selected to perform noise reduction on the original vibration signal. By analyzing the original vibration signal in the frequency domain, a low-pass filter is selected to remove high-frequency noise. Frequency domain analysis determines the frequency range that needs to be retained or removed, ensuring that only noise is filtered without affecting critical vibration information. Based on this, a filter can be designed that can reduce noise without introducing additional phase delay or distortion. The filter is established using the fir1 function, and the Hamming window is used as the window function of the filter. The filter length is 1024 coefficients and the cutoff frequency is 48Hz. By applying the designed digital filter to the collected original vibration signal, a real-time noise reduction effect is achieved.
使用滑动平均法平滑信号，减小瞬时波动的影响。定义用于计算平均值的滑动窗口大小，确保平滑效果符合振动信号性质和分析需求。对降噪后的数据逐点计算平均值，对窗口内的振动信号进行平滑处理。通过滑动平均，有效减小瞬时波动对后续分析的干扰，使振动信号更加稳定。Use the sliding average method to smooth the signal and reduce the impact of instantaneous fluctuations. Define the size of the sliding window used to calculate the average value to ensure that the smoothing effect meets the nature of the vibration signal and the analysis requirements. Calculate the average value of the noise-reduced data point by point and smooth the vibration signal within the window. Through sliding average, the interference of instantaneous fluctuations on subsequent analysis is effectively reduced, making the vibration signal more stable.
(2)时频变换：(2) Time-frequency transformation:
输入信号是在时域上采集的振动数据。将时域信号进行离散化处理，得到一系列离散的数据点。在时域信号上应用汉宁窗函数。对信号进行快速傅里叶变换FFT，将信号转换到频域。The input signal is the vibration data collected in the time domain. The time domain signal is discretized to obtain a series of discrete data points. The Hanning window function is applied to the time domain signal. The signal is transformed into the frequency domain by performing a fast Fourier transform (FFT).
(3)计算振动特征频率：(3) Calculate the vibration characteristic frequency:
1)频域分析的异常检测：通过FFT结果的频谱检测到的频率峰值或频率带，表示为：frequency_peaks＝find_peaks(spectrum)。其中spectrum为FFT结果的频谱，find_peaks为检测频谱中的峰值函数，frequency_peaks为检测到的频率峰值或频率带。通过设定阈值检查是否存在异常的频率峰值或频率带，同时通过分析频谱中能量的分布情况，察觉异常情况可能导致频谱中某些频率的能量显著增加。1) Anomaly detection by frequency domain analysis: The frequency peaks or frequency bands detected by the spectrum of the FFT result are expressed as: frequency_peaks = find_peaks(spectrum). Where spectrum is the spectrum of the FFT result, find_peaks is the peak function for detecting the spectrum, and frequency_peaks is the detected frequency peaks or frequency bands. By setting a threshold to check whether there are abnormal frequency peaks or frequency bands, and by analyzing the distribution of energy in the spectrum, it is detected that abnormal conditions may cause a significant increase in the energy of certain frequencies in the spectrum.
2)异常频率成分的识别：通过一个Butterworth高通滤波器，根据正常工作状态下的振动频谱实施频率滤波，滤除正常频率成分，使异常频率成分得以凸显。对正常状态和异常状态的频谱进行差异分析，找出差异较大的频率成分，确定初步局部异常情况。这里的差异较大的频率成分是指某些频率成分的幅度明显高于正常状态。此次统计为同一部件异常频率数量，为后续车辆部件健康状态做判断标准。2) Identification of abnormal frequency components: Through a Butterworth high-pass filter, frequency filtering is performed according to the vibration spectrum under normal working conditions to filter out normal frequency components and highlight abnormal frequency components. The frequency spectrum of normal and abnormal conditions is analyzed to find out the frequency components with large differences and determine the preliminary local abnormal conditions. The frequency components with large differences here refer to the amplitudes of some frequency components that are significantly higher than the normal state. This statistics is the number of abnormal frequencies of the same component, which will serve as a criterion for judging the health status of subsequent vehicle components.
同时建立根据正常工作状态下的频谱数据构建的统计模型，包括均值和标准差。At the same time, a statistical model is established based on the spectrum data under normal working conditions, including the mean and standard deviation.
其中，mean_spectrum表示频谱数据在所有频率点上的平均值。对于N个正常工作状态下的频谱数据normal_spectrum_i，计算它们在每个频率点上的平均值，对所有频谱数据进行求和并取平均。in, mean_spectrum represents the average value of the spectrum data at all frequency points. For N normal working spectrum data normal_spectrum_i, calculate their average values at each frequency point, sum and average all spectrum data.
std_spectrum表示频谱数据在所有频率点上的变异程度，即分布的离散程度。对于N个正常工作状态下的频谱数据normal_spectrum_i，计算了它们在每个频率点上与均值mean_spectrum的偏差的平方和的平均值的平方根，对所有频谱数据进行偏差平方和的平均值，然后取平方根。检测实际频谱与统计模型的偏差，判断是否超出一定阈值，从而确定是否存在同一部件整体异常情况。 std_spectrum indicates the degree of variation of the spectrum data at all frequency points, that is, the degree of dispersion of the distribution. For N normal working spectrum data normal_spectrum_i, the square root of the average of the sum of squares of their deviations from the mean mean_spectrum at each frequency point is calculated, and the average of the sum of squares of deviations of all spectrum data is taken, and then the square root is taken. The deviation between the actual spectrum and the statistical model is detected to determine whether it exceeds a certain threshold, so as to determine whether there is an overall abnormality in the same component.
(4)生成维护提醒和通知：(4) Generate maintenance reminders and notifications:
判断单位时间段内异常振动频率出现的次数，以确定连接处的健康状态。当单位时间段内异常振动频率少量出现在一定阈值范围内，则确定连接处的健康状态为正常状态或轻微告警状态。当单位时间段内异常振动频率中度出现超过一定阈值，则确定连接处的健康状态为中度告警状态。当单位时间段内异常振动频率大量出现超过一定阈值，则确定连接处的健康状态为严重告警状态。生成维护提醒，包括故障类型和建议的维修步骤(其中建议的维修是通过预先存储在服务器中的历史建议和维修方案提取得到)。生成的维护提醒通过无线通讯渠道传输至车主或维修人员。这可以通过手机应用等形式实现。通知中包含了故障的详细信息，以及建议的行动步骤，使车主或维修人员能够迅速采取适当的措施，提高车辆维护的效率和及时性。判断实际频谱与统计模型的偏差，判断是否超出一定阈值，从而确定这一部件是否存在频谱异常情况。Determine the number of times the abnormal vibration frequency occurs within a unit time period to determine the health status of the connection. When the abnormal vibration frequency occurs in a small amount within a certain threshold range within a unit time period, the health status of the connection is determined to be normal or a slight alarm state. When the abnormal vibration frequency occurs moderately within a unit time period and exceeds a certain threshold, the health status of the connection is determined to be a moderate alarm state. When the abnormal vibration frequency occurs in large quantities within a unit time period and exceeds a certain threshold, the health status of the connection is determined to be a serious alarm state. Generate a maintenance reminder, including the fault type and the recommended maintenance steps (where the recommended maintenance is obtained by extracting historical suggestions and maintenance plans pre-stored in the server). The generated maintenance reminder is transmitted to the owner or maintenance personnel through a wireless communication channel. This can be achieved in the form of a mobile phone application, etc. The notification contains detailed information about the fault and the recommended action steps, so that the owner or maintenance personnel can quickly take appropriate measures to improve the efficiency and timeliness of vehicle maintenance. Determine the deviation between the actual spectrum and the statistical model, and determine whether it exceeds a certain threshold, so as to determine whether there is a spectrum abnormality in this component.
(5)振动传感器矩阵实时损坏检测：(5) Real-time damage detection of vibration sensor matrix:
从第一传动轴9、第二传动轴9、第一驱动轴8和第二驱动轴振动传感器4获取真实行驶测试采集的三轴振动数据。The three-axis vibration data collected by the actual driving test is acquired from the first transmission shaft 9 , the second transmission shaft 9 , the first drive shaft 8 and the second drive shaft vibration sensor 4 .
对每个传感器的三轴振动数据进行实时采集和存储。The three-axis vibration data of each sensor is collected and stored in real time.
使用预训练分类模型进行特征提取，使用SUOD集成异常检测加速器集成孤立森林、lof(局部离群因子)、KNN(K近邻)等多种异常检测加权方法。Use pre-trained classification models for feature extraction, and use SUOD integrated anomaly detection accelerator to integrate multiple anomaly detection weighted methods such as isolation forest, lof (local outlier factor), KNN (K nearest neighbor), etc.
模型集成多种异常检测加权方法，对每种异常检测方法得到的异常分数进行加权计算，获得最终的综合异常分数。The model integrates multiple anomaly detection weighted methods, performs weighted calculation on the anomaly scores obtained by each anomaly detection method, and obtains the final comprehensive anomaly score.
对异常得分高的数据进行最后的人工标定，标定足够量的异常数据。对这些异常数据做人工标注，判断各个部件是否损坏。The data with high abnormal scores are finally manually calibrated to calibrate a sufficient amount of abnormal data. These abnormal data are manually labeled to determine whether each component is damaged.
将标定过的数据作为训练数据，对振动数据进行归一化处理，将其缩放到标准范围内，确保模型的稳定性和收敛速度。The calibrated data is used as training data, and the vibration data is normalized and scaled to the standard range to ensure the stability and convergence speed of the model.
使用训练数据对Anomaly Transformer模型进行训练。定义损失函数，均方误差(MSE)异常损失函数。Train the Anomaly Transformer model using the training data. Define the loss function, the mean squared error (MSE) anomaly loss function.
定义优化器，随机梯度下降(SGD)，设置学习率和其他超参数进行模型训练。迭代训练模型，通过反向传播优化模型参数，使其能够从振动数据中提取关键特征并预测异常。将训练好的Anomaly Transformer模型，通过convert_tflite_model函数将模型转换为TensorFlow Lite模型。将经过训练和转换的Anomaly Transformer模型部署到VCU，将处理后的振动数据转换成张量的形式输入模型。在VCU上设置实时推理任务，以连续地接收和处理传感器数据，并输出各个部件损坏检测结果。Define the optimizer, stochastic gradient descent (SGD), set the learning rate and other hyperparameters for model training. Iterate the model training and optimize the model parameters through back propagation so that it can extract key features from vibration data and predict anomalies. Convert the trained Anomaly Transformer model to a TensorFlow Lite model through the convert_tflite_model function. Deploy the trained and converted Anomaly Transformer model to the VCU and convert the processed vibration data into tensors for input into the model. Set up real-time inference tasks on the VCU to continuously receive and process sensor data and output the damage detection results of each component.
(6)数据记录和故障诊断：(6) Data recording and fault diagnosis:
系统将处理后的振动数据记录在安全可靠的服务器数据库中。这包括原始振动信号、降噪处理后的数据、时频分析结果、振动传感器矩阵实时损坏检测结果。数据库的结构允许高效地存储和检索历史数据。远程访问数据以进行故障诊断，提供更准确的信息，辅助维修工作。故障诊断用采用孤立森林。将经过预处理的振动数据作为输入，创建一个孤立森林模型。为了构建孤立森林模型，设定了相应的算法参数，并使用这一模型对数据进行训练。基于孤立森林模型的训练结果，将振动数据进行分类，分为两个类别：正常和损坏。正常的振动数据将被标记为类别0，而损坏的数据则被标记为类别1。最后，将这些标记过的数据重新存储，做备份保留。The system records the processed vibration data in a secure and reliable server database. This includes the original vibration signal, the data after noise reduction, the time-frequency analysis results, and the real-time damage detection results of the vibration sensor matrix. The structure of the database allows efficient storage and retrieval of historical data. Remote access to data for fault diagnosis provides more accurate information and assists maintenance work. Fault diagnosis uses isolation forests. The pre-processed vibration data is used as input to create an isolation forest model. In order to build the isolation forest model, the corresponding algorithm parameters are set and the data is trained using this model. Based on the training results of the isolation forest model, the vibration data is classified into two categories: normal and damaged. Normal vibration data will be marked as category 0, while damaged data will be marked as category 1. Finally, the labeled data is stored again for backup.
基于上述描述，本发明提供的ATV全地形车整车维护管理系统的整个实施流程如图2所示。这种ATV全地形车整车维护管理系统相较于现有技术在维护方面的优势显著，其通过精准布置振动传感器，系统实现全面监测和实时检测底盘关键连接处的振动情况，提高了维护的及时性。智能维护提醒系统通过异常频率和特征频率计算，生成详细的故障信息，并通过多渠道通知维护人员。同时，远程故障诊断和数据记录功能使得维护人员能够远程获取车辆振动数据，降低了维护成本，提高了维护效率。系统的降噪和准确性提高措施，进一步保障了振动信号的稳定性。综合而言，这一系统在全面性、实时性、智能性等方面为整车维护带来了显著的优势，助力降低成本和提高效率。Based on the above description, the entire implementation process of the ATV all-terrain vehicle maintenance management system provided by the present invention is shown in Figure 2. This ATV all-terrain vehicle maintenance management system has significant advantages over the prior art in terms of maintenance. By precisely arranging vibration sensors, the system can achieve comprehensive monitoring and real-time detection of vibration conditions at key connections of the chassis, thereby improving the timeliness of maintenance. The intelligent maintenance reminder system generates detailed fault information through abnormal frequency and characteristic frequency calculations, and notifies maintenance personnel through multiple channels. At the same time, remote fault diagnosis and data recording functions enable maintenance personnel to remotely obtain vehicle vibration data, reducing maintenance costs and improving maintenance efficiency. The system's noise reduction and accuracy improvement measures further ensure the stability of vibration signals. In summary, this system brings significant advantages to vehicle maintenance in terms of comprehensiveness, real-time, and intelligence, helping to reduce costs and improve efficiency.
并且，本发明提欧根的提供涵盖了传感器在底盘轴与轴连接处的安装，覆盖轮轴10、驱动轴8、传动轴9等关键部位。通过高采样频率实时捕捉并记录振动信号，传感器数据经A/D转换模块转为数字信号，采用CAN协议传输至车辆控制器5。无线通讯采用高效协议，确保稳定传输到服务器，进行数据处理与降噪，时频分析，计算振动特征频率，实现对异常频率的检测、振动传感器矩阵实时异常检测。通过生成维护提醒和通知，车主或维修人员能够及时采取维护措施。系统记录处理后的振动数据在服务器数据库，可通过远程访问进行故障诊断。Moreover, the provision of the TIOGEN of the present invention covers the installation of sensors at the connection between the chassis shaft and the shaft, covering key parts such as the wheel axle 10, the drive shaft 8, and the transmission shaft 9. Vibration signals are captured and recorded in real time through high sampling frequency, and the sensor data is converted into digital signals through the A/D conversion module and transmitted to the vehicle controller 5 using the CAN protocol. Wireless communication uses an efficient protocol to ensure stable transmission to the server, perform data processing and noise reduction, time-frequency analysis, calculate vibration characteristic frequencies, and realize detection of abnormal frequencies and real-time abnormality detection of vibration sensor matrices. By generating maintenance reminders and notifications, car owners or maintenance personnel can take maintenance measures in a timely manner. The system records the processed vibration data in the server database, which can be accessed remotely for fault diagnosis.
进一步，本发明还提供了一种ATV全地形车整车维护管理方法，该方法采用上述提供的ATV全地形车整车维护管理系统实现。该方法包括：Furthermore, the present invention also provides an ATV all-terrain vehicle maintenance management method, which is implemented using the ATV all-terrain vehicle maintenance management system provided above. The method comprises:
步骤1、获取原始振动信号。Step 1: Get the original vibration signal.
步骤2、将原始振动信号转化为数字振动数据。Step 2: Convert the original vibration signal into digital vibration data.
步骤3、对数字振动数据进行处理，并基于处理后的数字振动数据确定振动特征频率，具体包括：Step 3: Processing the digital vibration data and determining the vibration characteristic frequency based on the processed digital vibration data, specifically including:
使用fir1函数建立滤波器，并使用汉明窗口作为滤波器的窗口函数，对数字振动数据进行滤波处理，得到滤波信号。The fir1 function is used to establish a filter, and the Hamming window is used as the window function of the filter to filter the digital vibration data to obtain a filtered signal.
采用滑动平均法对滤波信号进行平滑处理，得到平滑信号。The sliding average method is used to smooth the filtered signal to obtain a smoothed signal.
对平滑信号进行快速傅里叶变换，得到频域信号。Perform fast Fourier transform on the smoothed signal to obtain the frequency domain signal.
基于频域信号得到振动特征频率。The vibration characteristic frequency is obtained based on the frequency domain signal.
步骤4、基于振动特征频率生成底盘部件检测结果。Step 4: Generate chassis component detection results based on the vibration characteristic frequency.
步骤5、当底盘部件检测结果存在异常时，服务器还用于生成维护提醒和通知。Step 5: When there is an abnormality in the chassis component detection result, the server is also used to generate maintenance reminders and notifications.
步骤6、当底盘部件检测结果无异常时，不做处理。Step 6: If there is no abnormality in the chassis component inspection results, no processing is performed.
进一步，作为本发明的一个实施例，上述步骤4中基于振动特征频率生成底盘部件检测结果的实施过程可以包括：Further, as an embodiment of the present invention, the implementation process of generating chassis component detection results based on the vibration characteristic frequency in the above step 4 may include:
基于振动特征频率生成频谱数据。Generate spectral data based on the vibration characteristic frequencies.
采用Butterworth高通滤波器对频谱数据进行滤波处理，得到正常状态频谱和异常状态频谱。The Butterworth high-pass filter is used to filter the spectrum data to obtain the normal state spectrum and the abnormal state spectrum.
对正常状态频谱和异常状态频谱进行差异分析，得到差异结果。The difference results are obtained by performing difference analysis on the normal state spectrum and the abnormal state spectrum.
当差异结果超过设定阈值时，确定ATV全地形车存在初步局部异常情况。当差异结果不超过设定阈值时，确定ATV全地形车无异常。When the difference result exceeds the set threshold, it is determined that the ATV all-terrain vehicle has a preliminary local abnormality. When the difference result does not exceed the set threshold, it is determined that the ATV all-terrain vehicle has no abnormality.
根据正常工作状态下的频谱数据构建的统计模型。A statistical model constructed based on spectrum data under normal working conditions.
检测实际频谱数据与统计模型得到的频谱数据间的偏差。Detect the deviation between the actual spectrum data and the spectrum data obtained by the statistical model.
判断偏差是否超出一定阈值，得到第一判断结果。It is determined whether the deviation exceeds a certain threshold value to obtain a first determination result.
当第一判断结果为是时，确定ATV全地形车存在存在同一部件的整体异常情况。When the first determination result is yes, it is determined that the ATV all-terrain vehicle has an overall abnormality of the same component.
当第一判断结果为否时，不做处理。When the first judgment result is no, no processing is performed.
再进一步，作为本发明的另一个实施例，上述步骤4中基于振动特征频率生成底盘部件检测结果的实施过程可以包括：Furthermore, as another embodiment of the present invention, the implementation process of generating chassis component detection results based on the vibration characteristic frequency in the above step 4 may include:
基于振动特征频率生成频谱数据。Generate spectral data based on the vibration characteristic frequencies.
采用Butterworth高通滤波器对频谱数据进行滤波处理，得到正常状态频谱和异常状态频谱。The Butterworth high-pass filter is used to filter the spectrum data to obtain the normal state spectrum and the abnormal state spectrum.
基于异常状态频谱得到异常振动频率。The abnormal vibration frequency is obtained based on the abnormal state spectrum.
判断单位时间段内异常振动频率出现的次数是否在第一设定阈值范围内，得到第二判断结果。It is determined whether the number of occurrences of the abnormal vibration frequency within a unit time period is within a first set threshold range to obtain a second determination result.
当第二判断结果为是，则确定ATV全地形车底盘轮轴10连接位置处或传动轴9、差速器7或驱动轴8结合位置处的健康状态为正常状态或轻微告警状态。When the second judgment result is yes, it is determined that the health status of the ATV all-terrain vehicle chassis wheel axle 10 connection position or the transmission shaft 9, differential 7 or drive shaft 8 combination position is normal or slightly alarmed.
当第二判断结果为否，则判断单位时间段内异常振动频率出现的次数是否在第二设定阈值范围内，得到第三判断结果。When the second judgment result is no, it is determined whether the number of occurrences of the abnormal vibration frequency within the unit time period is within the second set threshold range to obtain a third judgment result.
当第三判断结果为是时，则确定ATV全地形车底盘轮轴10连接位置处或传动轴9、差速器7或驱动轴8结合位置处的健康状态的为中度告警状态。When the third judgment result is yes, it is determined that the health status of the ATV all-terrain vehicle chassis wheel axle 10 connection position or the transmission shaft 9, differential 7 or drive shaft 8 combination position is a moderate warning state.
当第三判断结果为否时，则判断单位时间段内异常振动频率出现的次数是否在第三设定阈值范围内，得到第四判断结果。When the third judgment result is no, it is determined whether the number of occurrences of the abnormal vibration frequency within the unit time period is within the third set threshold range to obtain a fourth judgment result.
当第四判断结果为是时，则确定ATV全地形车底盘轮轴10连接位置处或传动轴9、差速器7或驱动轴8结合位置处的健康状态为严重告警状态。When the fourth judgment result is yes, it is determined that the health status of the ATV all-terrain vehicle chassis wheel shaft 10 connection position or the transmission shaft 9, differential 7 or drive shaft 8 combination position is a serious warning state.
进一步，作为本发明的又一个实施例，上述步骤4中基于振动特征频率生成底盘部件检测结果的实施过程可以包括：Further, as another embodiment of the present invention, the implementation process of generating chassis component detection results based on the vibration characteristic frequency in the above step 4 may include:
使用预训练分类模型进行特征提取，使用SUOD集成异常检测加速器集成多种异常检测加权方法。Use pre-trained classification models for feature extraction, and use the SUOD integrated anomaly detection accelerator to integrate multiple anomaly detection weighted methods.
采用每种异常检测方法基于历史振动数据得到异常分数，并对得到的异常分数进行加权计算，获得最终的综合异常分数。Each anomaly detection method is used to obtain an anomaly score based on historical vibration data, and the obtained anomaly scores are weighted to obtain the final comprehensive anomaly score.
对综合异常分数的历史振动数据进行人工标定。The historical vibration data of the comprehensive anomaly score was manually calibrated.
将标定过的数据作为训练数据，并对训练数据进行归一化处理。The calibrated data is used as training data and the training data is normalized.
使用归一化处理后的训练数据对Anomaly Transformer模型进行训练，得到训练好的Anomaly Transformer模型。The normalized training data is used to train the Anomaly Transformer model to obtain a trained Anomaly Transformer model.
通过convert_tflite_model函数将训练好的Anomaly Transformer模型转换为TensorFlow Lite模型。Convert the trained Anomaly Transformer model to a TensorFlow Lite model using the convert_tflite_model function.
将振动特征频率转换成张量的形式输入TensorFlow Lite模型得到底盘部件检测结果。The vibration characteristic frequency is converted into a tensor and input into the TensorFlow Lite model to obtain the chassis component detection results.
基于上述描述，本发明旨在解决传统振动监测系统在全面监测ATV底盘结构健康状况方面存在的不足，其通过采用先进的振动信号采集和处理技术，系统能够实时监测底盘关键部位的振动情况，进行高效的时频分析和振动特征频率计算，从而提供更准确的异常检测和故障诊断。通过生成维护提醒和通知，系统使车主或维修人员能够迅速采取适当的措施，提高了车辆的维护效率和及时性。因此，该系统能够在ATV的全地形使用场景中，提升车辆的可靠性和寿命，减少维修成本。Based on the above description, the present invention aims to solve the shortcomings of traditional vibration monitoring systems in comprehensively monitoring the health status of ATV chassis structures. By adopting advanced vibration signal acquisition and processing technology, the system can monitor the vibration conditions of key parts of the chassis in real time, perform efficient time-frequency analysis and vibration characteristic frequency calculation, thereby providing more accurate abnormality detection and fault diagnosis. By generating maintenance reminders and notifications, the system enables vehicle owners or maintenance personnel to take appropriate measures quickly, improving the efficiency and timeliness of vehicle maintenance. Therefore, the system can improve the reliability and life of the vehicle and reduce maintenance costs in the all-terrain use scenario of the ATV.
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，本发明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 RandomAccess Memory，MRAM)、铁电存储器(Ferroelectric RandomAccess 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(RandomAccess 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 RandomAccessMemory，SRAM)或动态随机存取存储器(Dynamic RandomAccessMemory，DRAM)等。本发明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。本发明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 skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided by the present invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided by the present invention may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited thereto. The processor involved in each embodiment provided by the present invention may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited thereto.
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 may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
本文中应用了具体个例对本发明的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本发明的方法及其核心思想，各个实施例之间相同相似部分互相参见即可；同时，对于本领域的一般技术人员，依据本发明的思想，在具体实施方式及应用范围上均会有改变之处。综上所述，本说明书内容不应理解为对本发明的限制。The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core ideas of the present invention. The same and similar parts between the embodiments can be referred to each other. At the same time, for those skilled in the art, according to the ideas of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410781045.0ACN118790172A (en) | 2024-06-18 | 2024-06-18 | ATV all-terrain vehicle maintenance management system and method |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410781045.0ACN118790172A (en) | 2024-06-18 | 2024-06-18 | ATV all-terrain vehicle maintenance management system and method |
| Publication Number | Publication Date |
|---|---|
| CN118790172Atrue CN118790172A (en) | 2024-10-18 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410781045.0APendingCN118790172A (en) | 2024-06-18 | 2024-06-18 | ATV all-terrain vehicle maintenance management system and method |
| Country | Link |
|---|---|
| CN (1) | CN118790172A (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1926413A (en)* | 2004-10-18 | 2007-03-07 | 日本精工株式会社 | Abnormality diagnosis system for machinery |
| CN107490488A (en)* | 2016-06-09 | 2017-12-19 | 通用汽车环球科技运作有限责任公司 | The vehicle health situation inspection carried out via noise and level of vibration |
| CN114270418A (en)* | 2019-08-15 | 2022-04-01 | 实耐宝公司 | vehicle health record |
| WO2022112920A1 (en)* | 2020-11-25 | 2022-06-02 | Continental Automotive | Method and device for detecting an anomaly in the operation of a vehicle, vehicle and data medium |
| CN117213856A (en)* | 2023-06-13 | 2023-12-12 | 贵州大学 | Bearing residual life prediction method based on Yun Bian double-data-source fusion |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1926413A (en)* | 2004-10-18 | 2007-03-07 | 日本精工株式会社 | Abnormality diagnosis system for machinery |
| CN107490488A (en)* | 2016-06-09 | 2017-12-19 | 通用汽车环球科技运作有限责任公司 | The vehicle health situation inspection carried out via noise and level of vibration |
| CN114270418A (en)* | 2019-08-15 | 2022-04-01 | 实耐宝公司 | vehicle health record |
| WO2022112920A1 (en)* | 2020-11-25 | 2022-06-02 | Continental Automotive | Method and device for detecting an anomaly in the operation of a vehicle, vehicle and data medium |
| CN117213856A (en)* | 2023-06-13 | 2023-12-12 | 贵州大学 | Bearing residual life prediction method based on Yun Bian double-data-source fusion |
| Publication | Publication Date | Title |
|---|---|---|
| DE102019108446A1 (en) | Method and apparatus for isolating a vehicle-side fault | |
| US20220118987A1 (en) | Method and system for fault diagnoses of intelligent vehicles | |
| CN103115789A (en) | Second generation small-wave support vector machine assessment method for damage and remaining life of metal structure | |
| CN114973155B (en) | Intelligent monitoring, analyzing and managing system based on AI image recognition behaviors | |
| CN104571079A (en) | Wireless long-distance fault diagnosis system based on multiple-sensor information fusion | |
| DE102017111505A1 (en) | Systems and methods for retrieving data from a remote system | |
| CN110084534A (en) | A kind of driving risks and assumptions quantization method based on driving behavior portrait | |
| CN114861719A (en) | High-speed train bearing fault diagnosis method based on ensemble learning | |
| CN109765879A (en) | A kind of remote monitoring system of new-energy automobile | |
| CN107563403B (en) | A working condition identification method for high-speed train operation | |
| CN118711279A (en) | A data-driven RV safety monitoring system and method | |
| CN114202198A (en) | Cloud-edge collaborative computing smart drive axle health monitoring system and method | |
| CN117369418A (en) | Fault diagnosis method, system, storage medium and equipment for feeding system of numerical control machine tool | |
| CN118790172A (en) | ATV all-terrain vehicle maintenance management system and method | |
| CN107539038A (en) | A kind of vehicle tyre pressure method for monitoring state and device | |
| CN114595721A (en) | Vehicle part strain load prediction method based on road condition recognition | |
| CN118551258A (en) | Unmanned mining card health diagnosis analysis method and system based on federal learning | |
| CN117392857B (en) | Truck type recognition system and recognition method based on Bluetooth network | |
| CN118230553A (en) | A traffic flow data analysis system based on highway | |
| CN117131450A (en) | Electric automobile charging equipment fault detection method, system, equipment and medium | |
| CN114357893A (en) | Vehicle-mounted antenna transmission system health assessment method based on multi-feature fusion SOM | |
| CN112850408A (en) | Elevator emergency stop trapped person fault detection method based on multi-model fusion | |
| Zhao et al. | Research on fault diagnosis method of electric vehicle battery system based on wavelet-RBF neural network | |
| CN114662219B (en) | Wheel wear prediction network model training method based on wavelet network model | |
| CN119066370B (en) | Monitoring and early warning method, system and storage medium for working state of hexahedral top press |
| 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 | ||
| RJ01 | Rejection of invention patent application after publication | Application publication date:20241018 |