A health evaluation method and system for wheel size detection equipment of a motor vehicle
By collecting and processing functional parameters of the EMU wheel size detection equipment, a comprehensive evaluation model is built, combining subjective and objective evaluation, multiple algorithms are used to evaluate the equipment health status, and visually display it through digital twin models, the equipment health status evaluation is solved The problems of strong subjectivity and high cost achieve more accurate and efficient equipment health assessment and maintenance.
Patent Information
- Application Number
- CN202410825720.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-25
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2044-06-25
AI Technical Summary
The prior art has problems of strong subjectivity and high cost in the evaluation of equipment health status. The sensor configuration requirements are high and cannot fully reflect the equipment status, resulting in lagging maintenance timing, low efficiency and high cost.
The health evaluation method of the EMU wheel size detection equipment is adopted, and functional parameters are collected, pre-processed and dimensional reduction are performed, and a comprehensive evaluation model is constructed. Combined subjective and objective evaluation, multiple algorithms are used to evaluate the equipment health status, and visualize it through the digital twin model. exhibit.
It improves the accuracy and efficiency of equipment health status assessment, reduces maintenance costs, and improves equipment reliability and railway safety.
Smart Images

Figure CN118411153B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of wheel inspection equipment health evaluation, and in particular to a motor vehicle wheel size detection equipment health evaluation method and system. Background Art
[0002] Accelerate the integration and development of new-generation information technology and manufacturing technology, take intelligent manufacturing as the main direction of the deep integration of the two, focus on the development of intelligent equipment and intelligent products, promote the intelligent production process, cultivate new production methods, and comprehensively improve the intelligent level of enterprise R&D, production, management and services. Equipment health status assessment is the core business of industrial enterprises and an important part of equipment intelligence and equipment management intelligence. Based on the current status of equipment use, it is crucial to conduct a scientific and reasonable evaluation of the health status of the equipment and carry out reasonable and scientific overhaul, minor repair, planned repair and preventive repair. In addition, the documents
[2021] No. 114 and
[2021] No. 59 issued by the China Railway Corporation also require the implementation of standardized, automated and intelligent maintenance, and carry out planned preventive maintenance while ensuring the reliability of equipment. Equipment health status assessment is a management model that promotes the overall health of equipment, which can help operators accurately grasp the technical status and risk points of equipment, and reduce equipment health risks through accurate and effective equipment inspection and maintenance activities. Therefore, it is crucial to conduct a scientific and reasonable health evaluation of equipment.
[0003] There are currently two main directions for evaluating equipment: one is to use the operating indicators currently monitored by the equipment, diagnose the importance relationship of each indicator with experts, and use highly subjective methods such as hierarchical analysis to evaluate the health status of the equipment. This technology is highly subjective and does not fully reflect the health status of the equipment due to incomplete indicators. The second is to increase the number of monitoring sensors to monitor the devices that mainly affect the operation of the equipment, use machine learning and deep learning algorithms to identify the health status of the equipment, and determine the status of the equipment. However, some equipment has a more complex structure and requires more devices to be monitored. The model and configuration of the monitoring sensor are required to be high, the cost of equipment production is increased, and some devices cannot be monitored by sensors, and the monitored items cannot fully reflect the status information of the equipment. Summary of the invention
[0004] In view of this, the present invention provides a method and system for evaluating the health of a motor vehicle wheel size detection device to solve the above-mentioned problem.
[0005] In order to solve the above technical problems, the present invention provides a health evaluation method for a motor vehicle wheel size detection device, comprising:
[0006] Collect functional parameters of wheel size detection equipment;
[0007] After preprocessing the functional parameters to satisfy the normal distribution, the functional parameters are reduced in dimension;
[0008] A comprehensive evaluation model is constructed, and the data extracted after dimensionality reduction is used as the input parameter of the model. The health evaluation results of the wheel size detection equipment are fed back through the comprehensive evaluation model;
[0009] The health assessment results are displayed visually, including establishing a health assessment report and building a digital twin model.
[0010] As an optional method, the functional parameters include operation data, data quality data, equipment comprehensive evaluation data and fault data of the wheel size detection equipment.
[0011] As an optional method, preprocessing the function parameters is to standardize the function parameters;
[0012] Dimensionality reduction of functional parameters includes:
[0013] According to the preset, the feature whose contribution rate to the health of the equipment is greater than the threshold in the functional parameters is taken as the first main feature; then, in each first main feature, the functional parameter indicators whose absolute value of the influence of the linear combination relationship on the first main feature is greater than the preset value are summarized to obtain the second main feature; the second main feature is the input parameter of the comprehensive evaluation model.
[0014] As an optional method, when the comprehensive evaluation model is being evaluated, if the accuracy of the comprehensive evaluation model does not meet the requirements and / or the input parameters have not been accumulated, the output parameters of the model will be packaged for secondary construction and optimization of the model at this stage; until the accuracy of the comprehensive evaluation model reaches the expected level, its output parameters will be accumulated as the model data set.
[0015] As an optional method, the evaluation process of the comprehensive evaluation model includes:
[0016] After inputting the parameters, subjective evaluation and objective evaluation are performed respectively, and the results of subjective evaluation and objective evaluation are weighted to obtain the equipment evaluation weight, and then the results are summarized.
[0017] As an optional method, subjective evaluation includes:
[0018] Experts score the importance of the reduced-dimensional indicators based on the current data distribution and the current equipment operation status. Then, all the results are summarized and outliers are removed. The weight of each indicator is calculated based on the mean of each feature.
[0019] Objective evaluation includes:
[0020] The entropy weight method is used to weight the current data features, and the objective weights of each device feature are obtained from the perspective of data distribution.
[0021] As an optional method, after the dimensionality reduction data is input, the new data and previous data of the selected features are used to calculate the comprehensive weight through the entropy weight method. When the result calculated by combining the comprehensive weight with the subjective weight does not meet the verification requirements, the weighting formula is adjusted until the parameters meet the verification conditions.
[0022] As an optional method, random forest, GBDT and XGBoost algorithms are used as base learners, and the stacking method is used to calculate the comprehensive result, which will be used as the final result of health evaluation.
[0023] As an optional method, the health evaluation report is used to display the score and status of the equipment health evaluation, the recent status trend curve of the equipment, the detailed data of the equipment health evaluation, and the horizontal comparison of the status change and development of the equipment with similar equipment;
[0024] The digital twin model is used to achieve synchronization and interactivity between the physical model and virtual model of the equipment. It is configured with BIM models, charts, indicators and alarm modules to quickly understand the operating status of equipment and facilities and take timely measures. After taking measures, the abnormal equipment or component marks will be cancelled.
[0025] On the other hand, the present invention also provides a motor vehicle wheel size detection equipment health evaluation system, which is used to execute the above motor vehicle wheel size detection equipment health evaluation method, comprising:
[0026] A data acquisition module, the data acquisition module is used to collect functional parameters of the wheel size detection equipment;
[0027] Feature extraction module: The feature extraction module is used to pre-process the function parameters so that they meet the normal distribution and then reduce the dimension of the function parameters;
[0028] The model building module is used to build a comprehensive evaluation model, and the data extracted after dimensionality reduction is used as the input parameter of the model. The health evaluation result of the wheel size detection equipment is fed back through the comprehensive evaluation model;
[0029] A health evaluation module, which is used to convert the accumulated model output parameters into final health evaluation results according to a preset method;
[0030] Health evaluation visualization module: The health evaluation visualization module visualizes the health evaluation results. The visualization includes establishing a health evaluation report and building a digital twin model.
[0031] The beneficial effects of the present invention are:
[0032] In view of the problems of delayed maintenance, low efficiency and high cost caused by planned maintenance and post-maintenance of wheel detection equipment of wheel group, the present invention proposes a health evaluation method and system for EMU wheel size detection equipment, which can evaluate and predict the health status of the equipment, improve the maintenance mode of EMU detection equipment, solve the current problems of high maintenance cost and low efficiency, improve the reliability of EMU wheel detection equipment, and ensure railway safety. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] Figure 1 A schematic diagram of the application process of the health assessment method provided by an embodiment of the present invention;
[0034] Figure 2 A flow chart of a health evaluation algorithm for a wheel size detection device provided in an embodiment of the present invention;
[0035] Figure 3 A schematic diagram of a main component analysis and feature extraction process provided by an embodiment of the present invention;
[0036] Figure 4 A flow chart of model construction provided by an embodiment of the present invention;
[0037] Figure 5 A schematic diagram of an application of a health assessment model provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0038] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention is further described in detail below in conjunction with specific implementation methods.
[0039] Example
[0040] There are currently two main directions for evaluating equipment: one is to use the operating indicators currently monitored by the equipment, diagnose the importance relationship of each indicator with experts, and use highly subjective methods such as hierarchical analysis to evaluate the health status of the equipment. This technology is highly subjective and does not fully reflect the health status of the equipment due to incomplete indicators. The second is to increase the number of monitoring sensors to monitor the devices that mainly affect the operation of the equipment, use machine learning and deep learning algorithms to identify the health status of the equipment, and determine the status of the equipment. However, some equipment has a more complex structure and requires more devices to be monitored. The model and configuration of the monitoring sensor are required to be high, the cost of equipment production is increased, and some devices cannot be monitored by sensors, and the monitored items cannot fully reflect the status information of the equipment.
[0041] See also Figure 1, which is the application process of the health assessment method of this embodiment. In order to cope with the current status of technological development, innovations and improvements are made in data acquisition, health assessment algorithms and visualization of results. This embodiment deeply understands and grasps the operation and structure of the equipment. In order to ensure the comprehensiveness of the data, this embodiment adds an Internet of Things gateway, improves the data access method, and realizes data access in multiple formats of the equipment. Then, comprehensive data mining and analysis are carried out, and combined with the opinions of relevant experts and technicians, it is summarized that the health status of the equipment can be evaluated from the stability of the equipment, the performance of the equipment, and the maintenance level of the equipment. The indicators of the three dimensions can more comprehensively reflect the health status of the equipment. In addition, the evaluation method combines the opinions of subjective and objective factors to evaluate the equipment, and combines multiple algorithms to predict the status of the equipment. Finally, most of the current equipment health research only studies the algorithm, without considering the convenience and operability of the work of on-site application personnel. Therefore, this embodiment integrates digital twin technology, maps real equipment to virtual space, and visualizes the status of the equipment, which is convenient for operators to view and use.
[0042] See also Figure 2-Figure 5 , this embodiment is implemented as follows:
[0043] A health evaluation method for wheel size detection equipment of a motor vehicle comprises: collecting functional parameters of the wheel size detection equipment; preprocessing the functional parameters to satisfy a normal distribution, and then reducing the dimensions of the functional parameters; constructing a comprehensive evaluation model, using the data extracted after the dimension reduction as input parameters of the model, and feeding back the health evaluation results of the wheel size detection equipment through the comprehensive evaluation model; and visually displaying the health evaluation results, wherein the visual display comprises establishing a health evaluation report and constructing a digital twin model.
[0044] When collecting data, the wheel size detection equipment will generate data (functional parameters) in various formats in real time, such as image data, ultrasonic data, status data, etc. The amount of data is large, and the previous data access method cannot efficiently and completely access it to accumulate data. This embodiment adds an Internet of Things gateway to access the data, through which the operation data, data quality data, equipment comprehensive evaluation data, fault data, etc. of the wheel size detection equipment are collected. These data mainly include equipment stability, equipment data quality, and data maintenance and repair dimensions. There are some corresponding indicators under each dimension, and each dimension has 11 indicators, 7 indicators, and 5 indicators, respectively, with a total of 23 indicators. The details of the indicators are shown in Table 1. These indicators can more comprehensively reflect the health status of the equipment.
[0045] Table 1 Equipment health assessment index table
[0046] Equipment stability Data quality Maintenance Level Data collection box 1 Accuracy satisfaction rate Mean time to recover from failure Data collection box 2 Equipment effective operation rate Software Uptime Data collection box 3 Equipment utilization Hardware Uptime Data collection box 4 Time utilization Ratio of maintenance personnel per device Data collection box 5 Failure downtime rate Equipment monthly maintenance frequency Data collection box 6 Equipment failure rate Data collection box 7 Time availability Data collection box 8 Precision trigger sensor Trackside integrated control box Trackside Environmental Control Box
[0047] After that, the above data is subjected to feature extraction. The feature extraction module first standardizes the data. Since the principal component analysis is to be performed to extract indicators, the data needs to satisfy the normal distribution, so z-score standardization is used. The formula is: (x-mean) / std, where x is the original data, mean and std are the mean and standard deviation of the data.
[0048] The process of this embodiment is as follows Figure 3 As shown, this embodiment reduces the dimension of the data, extracts the first main feature as the principal component, finds out the index with greater influence on the contribution rate as the second main feature, and uses the dimension reduction result as the model input.
[0049] Optionally, three principal components are selected as the result of taking the components with a principal component contribution rate greater than 80%. In addition, since the selected principal components are combined variables, in order to ensure the completeness of the indicators, this embodiment makes a simple modification to the principal component method, and selects the indicators with an absolute value of influence on the principal component greater than 0.6 in the linear combination relationship of each principal component for summary statistics. The summarized indicators are used as indicators for the next step of health evaluation. A total of 13 indicators are selected, including 6, 4, and 3 indicators in the equipment stability, equipment data quality, and data maintenance and repair dimensions, as shown in Table 2.
[0050] Table 2 Principal component extraction feature results
[0051] Equipment stability Data quality Maintenance Level Data collection box 1 Accuracy satisfaction rate Mean time to recover from failure Data collection box 4 Equipment effective operation rate Software Uptime Data collection box 6 Failure downtime rate Hardware Uptime Data collection box 8 Equipment failure rate Trackside integrated control box Precision trigger sensor
[0052] In particular, in this embodiment, according to the stage of health assessment, it is divided into two stages, namely, the debugging and optimization stage and the model maturity stage. In the debugging and optimization stage of the health assessment model, when the accuracy of the health assessment algorithm is not high or the health assessment data set is not accumulated or the accumulated data amount is small, the data after dimension reduction in this stage will be input into the model construction module to build the evaluation model and optimize and adjust it. After the accuracy of the health assessment model reaches the expected level, the data after dimension reduction will be input into the health assessment module as the accumulation of the health assessment model data set.
[0053] The model building module of this embodiment is as follows Figure 4 As shown in the figure, it is mainly divided into two parts. The first part is the comprehensive evaluation model, and the second part is the adjustment and optimization of the comprehensive evaluation model. After the reduced-dimensional data is input, the health evaluation results will be calculated through subjective and objective evaluations.
[0054] The subjective evaluation part of this embodiment adopts the expert evaluation method. Some relevant experts and technicians familiar with the equipment score the importance of the reduced-dimensional indicators based on the current data distribution and the current equipment operation status. Then, all the results are summarized and outliers are removed. The weight of each indicator is calculated based on the mean value of each feature. ,in It represents the subjective weight of the ith indicator. The input data will be used to calculate the subjective result of equipment health evaluation according to the obtained weight.
[0055] The objective evaluation part of this embodiment adopts the entropy weight method. The entropy weight method is used to weight the current data features and obtain the objective weight of each feature of the device from the perspective of data distribution. ,in It represents the objective weight of the ith indicator. The input data will be used to calculate the objective result of the equipment health evaluation based on the obtained weight.
[0056] Afterwards, the weighted average method is used to combine the subjective and objective evaluations mentioned above. The weights of the indicators obtained by the two evaluation methods are weighted, and the result is used as the final weight of the equipment indicator. The weighted formula is: , It represents the subjective and objective weighted weight of the ith indicator, k is the weighting factor, and the value of k will be adjusted and optimized according to the practical effect. The basis for adjusting the k value is the accuracy of the previous equipment health evaluation results and their feedback.
[0057] Finally, the results are summarized. The comprehensive weight results and characteristic values are summarized as the health evaluation results of the equipment. The formula is: , where F is the device health score, represents the indicator weight, Indicates the index score. As well as on-site feedback, on-site technicians will track the equipment every day, check the results of each indicator, and feedback the true status of the equipment; algorithm personnel will continuously adjust the evaluation-related algorithms based on the feedback data at the current stage, and use the accuracy of the algorithm results and the actual results as the basis for verifying the model. After the accuracy reaches the set threshold, the model is tried for a period of time without obvious problems, and the model is considered mature.
[0058] Optionally, in addition to simple training, this embodiment also includes comprehensive evaluation model adjustment and optimization. After the dimensionality reduction data is input, the new data and previous data of the selected features are weighted by the entropy weight method. The weights are combined with the subjective weights. If the calculated result does not meet the verification, the weighting formula will be automatically adjusted. Subjective and objective weighted parameters until the parameters meet the validation criteria.
[0059] After the above model is built and when the model is in the mature stage, the health evaluation module will use the random forest, GBDT, and XGBoost algorithms as base learners based on the currently accumulated data, and use the stacking method to calculate the comprehensive results of the three methods. The results will be used as the final results of the health evaluation, such as Figure 5 As shown, only the parameters and algorithms of the model need to be updated regularly afterwards. Currently, this module considers the function of automatic training. Before each data input, multiple base learners (random forest, GBDT, XGBoost algorithm) will be trained according to the data before the calculation date, and then the stacking method will be used to train the comprehensive model. The model will automatically train according to the threshold (the accuracy of previous feedback data), adjust the parameters related to the model, and output the local optimal parameters. Finally, the results of multiple models will be combined to obtain the comprehensive results of multiple models as the final health evaluation score, which is used to guide production and application practice. If a model training does not reach the threshold, the current optimal training result will be selected for output, and the current date will be marked for subsequent verification and adjustment optimization. After the model matures, the results of the feature extraction module will be directly input into the health evaluation module for subsequent prediction of the health status of the equipment.
[0060] Finally, in order to facilitate the use of technical personnel, this embodiment will also visualize the health assessment results, which are mainly divided into two parts. The first part is the health assessment report, which shows the score and status of the equipment health assessment, the recent status trend curve of the equipment, the detailed data of the equipment health assessment, and the horizontal comparison of the status changes and development of the equipment with the same type of equipment; the second part is the digital twin model, which realizes the synchronization and interactivity of the physical model and the virtual model of the equipment. Users can quickly understand the operation of equipment and facilities through BIM models, charts, indicators and alarm information, and take timely measures. After taking corresponding measures, the abnormal equipment or component mark will be cancelled. In addition, users can also use it to quickly understand the equipment, train new employees and carry out related technical demonstration research.
[0061] Based on the above scheme, in order to solve the problems in the field of equipment health status assessment that the evaluation methods are highly subjective, the mining of equipment is not in-depth and comprehensive enough, and the portability and practicality are not strong, this embodiment starts from the multi-source data access of equipment, algorithm optimization, and result visualization innovation. By constructing a multi-dimensional equipment health assessment indicator system, this embodiment proposes an evaluation method for in-depth mining and analysis of equipment conditions, a health assessment multi-model combination prediction method, and a digital twin display of health assessment results.
[0062] This embodiment comprehensively evaluates the health status of the equipment from three dimensions: stability, performance, and maintenance level. The indicators involved can fully reflect the health status of the equipment. The evaluation method not only takes into account the opinions of relevant experts and technicians, but also takes into account the distribution of data. It calculates the weights of each equipment indicator from a practical and objective perspective, and evaluates the equipment by comprehensively considering subjective and objective factors. The multi-model combination prediction method can greatly improve the accuracy and practicality of equipment health status assessment, increase the efficiency of evaluation, and provide support for the realization of intelligent manufacturing and equipment reliability.
[0063] In addition, this embodiment also provides a motor vehicle wheel size detection equipment health evaluation system, which is used to execute the above motor vehicle wheel size detection equipment health evaluation method, including:
[0064] A data acquisition module, the data acquisition module is used to collect functional parameters of the wheel size detection equipment;
[0065] Feature extraction module: The feature extraction module is used to pre-process the function parameters so that they meet the normal distribution and then reduce the dimension of the function parameters;
[0066] The model building module is used to build a comprehensive evaluation model, and the data extracted after dimensionality reduction is used as the input parameter of the model. The health evaluation result of the wheel size detection equipment is fed back through the comprehensive evaluation model;
[0067] A health evaluation module, which is used to convert the accumulated model output parameters into final health evaluation results according to a preset method;
[0068] Health evaluation visualization module: The health evaluation visualization module visualizes the health evaluation results. The visualization includes establishing a health evaluation report and building a digital twin model.
[0069] In summary, the application field of this embodiment is the wheel size detection equipment of the wheel set, and the comprehensiveness of the data is considered from the data perspective. Based on the operation of the equipment, the operation law and failure law of the equipment are analyzed, and then the equipment is evaluated in multiple dimensions in combination with subjective factors and objective factors, and the evaluation results are applied to the site for practical verification. In addition, this embodiment also integrates digital twin technology and Internet of Things technology. Based on the health status and usage of the equipment, abnormal equipment or components are marked and displayed to the user, realizing the synchronization and interactivity of the physical entity and virtual entity of the equipment, improving the real-time monitoring, simulation and prediction capabilities of the physical entity, and promoting the intelligent management and operation and maintenance of the equipment. Users can quickly understand the operation of equipment and facilities through BIM models, charts, indicators and alarm information, and take timely measures. After taking corresponding measures, the abnormal equipment or component mark will be cancelled. The combined application of digital twin technology and health assessment can quickly and intuitively display the health status of the equipment, which is convenient for the rapid resolution of problems.
[0070] The above are only preferred embodiments of the present invention. It should be noted that the above preferred embodiments should not be regarded as limiting the present invention, and the protection scope of the present invention should be based on the scope defined by the claims. For ordinary technicians in this technical field, several improvements and modifications can be made without departing from the spirit and scope of the present invention, and these improvements and modifications should also be regarded as the protection scope of the present invention.
Claims
1. A health evaluation method for wheel size detection equipment of a motor vehicle. It is characterized in that include: Collect functional parameters of wheel size detection equipment; After preprocessing the functional parameters to satisfy normal distribution, dimension reduction is performed on the functional parameters; The functional parameters include operation data, data quality data, equipment comprehensive evaluation data and fault data of the wheel size detection equipment; The dimension reduction includes: according to the preset, the features in the functional parameters whose contribution rate to the health of the equipment is greater than 80% are taken as the first main features, and then the functional parameter indicators in each first main feature whose absolute value of the linear combination relationship on the first main feature is greater than 0.6 are summarized to obtain the second main feature; the second main feature is the input parameter of the comprehensive evaluation model; wherein the input parameter includes at least three dimensions: equipment stability, equipment data quality, and data maintenance and repair; A comprehensive evaluation model is constructed, the data extracted after dimensionality reduction is used as input parameters of the model, and the health evaluation result of the wheel size detection equipment is fed back through the comprehensive evaluation model; The health assessment results are visualized, and the visualization includes establishing a health assessment report and building a digital twin model.
2. A method for evaluating the health of a motor vehicle wheel size detection device according to claim 1, It is characterized in that When the comprehensive evaluation model is being evaluated, if the accuracy of the comprehensive evaluation model does not meet the requirements, the output parameters of the model will be packaged for secondary construction and optimization of the model at this stage; Until the accuracy of the comprehensive evaluation model reaches the expected level, its output parameters will be accumulated as the model data set; wherein, judging whether the accuracy meets the requirements includes: judging whether the input parameters are accumulated and / or whether the data volume reaches a preset threshold.
3. A method for health assessment of a motor vehicle wheel size detection device according to claim 1, It is characterized in that The evaluation process of the comprehensive evaluation model includes: After the parameters are input, subjective evaluation and objective evaluation are performed respectively, and after the results of the subjective evaluation and objective evaluation are weighted to obtain the equipment evaluation weight, the results are summarized.
4. A method for health assessment of a motor vehicle wheel size detection device according to claim 3, It is characterized in that The subjective evaluation includes: Experts score the importance of the reduced-dimensional indicators based on the current data distribution and the current equipment operation status. Then, all the results are summarized and outliers are removed. The weight of each indicator is calculated based on the mean of each feature. The objective evaluation includes: The entropy weight method is used to weight the current data features, and the objective weights of each device feature are obtained from the perspective of data distribution.
5. A method for health assessment of a motor vehicle wheel size detection device according to claim 4, It is characterized in that After inputting the reduced-dimensional data, the new and previous data of the selected features are used to calculate the comprehensive weights through the entropy weight method. When the result calculated by combining the comprehensive weights with the subjective weights does not meet the verification requirements, the weighting formula is adjusted until the parameters meet the verification requirements.
6. A method for health assessment of a motor vehicle wheel size detection device according to claim 1, It is characterized in that Using random forest, GBDT and XGBoost algorithms as base learners, the stacking method is used to calculate the comprehensive result, which will be used as the final result of health evaluation.
7. A method for health assessment of a motor vehicle wheel size detection device according to claim 1, It is characterized in that The health evaluation report is used to display the score and status of the equipment health evaluation, the recent status trend curve of the equipment, detailed data related to the equipment health evaluation, and a horizontal comparison of the status changes and development of the equipment with similar equipment; The digital twin model is used to achieve synchronization and interactivity between the physical model and virtual model of the equipment. It is configured with BIM models, charts, indicators and alarm modules to quickly understand the operating status of equipment and facilities and take timely measures. After taking measures, the abnormal equipment or component marks will be cancelled.
8. A motor vehicle wheel size detection equipment health evaluation system, used to execute the motor vehicle wheel size detection equipment health evaluation method as claimed in any one of claims 1 to 7, It is characterized in that include: A data acquisition module, wherein the data acquisition module is used to acquire functional parameters of the wheel size detection device; A feature extraction module, wherein the feature extraction module is used to pre-process the function parameters so that they satisfy a normal distribution and then reduce the dimension of the function parameters; The functional parameters include operation data, data quality data, equipment comprehensive evaluation data and fault data of the wheel size detection equipment; The dimension reduction includes: according to the preset, the features in the functional parameters whose contribution rate to the health of the equipment is greater than 80% are taken as the first main features, and then the functional parameter indicators in each first main feature whose absolute value of the linear combination relationship on the first main feature is greater than 0.6 are summarized to obtain the second main feature; the second main feature is the input parameter of the comprehensive evaluation model; wherein the input parameter includes at least three dimensions: equipment stability, equipment data quality, and data maintenance and repair; A model building module, the model building module is used to build a comprehensive evaluation model, use the data extracted after dimensionality reduction as input parameters of the model, and feed back the health evaluation result of the wheel size detection equipment through the comprehensive evaluation model; A health evaluation module, which is used to convert the accumulated model output parameters into final health evaluation results according to a preset method; A health assessment visualization module, wherein the health assessment visualization module visualizes the health assessment results, and the visualization includes establishing a health assessment report and constructing a digital twin model.
