A decision-making method and device for maintenance of rail transit vehicle running gear

Abstract

The present application discloses a decision-making method and device for the maintenance of the running gear of a rail transit vehicle, and relates to the field of rail transit. The method determines the sampling samples from all samples according to the preset sampling rules; determines the unqualified samples according to the comprehensive decision results of the health status of the sampling samples; determines whether the current sampling is passed according to the number of unqualified samples; if so, the current sampling is ended, and after a preset period of interval, the step of determining the sampling samples from all samples according to the preset sampling rules is entered; if not, the sampling width and severity rules are adjusted according to the number of sampling failures, and the step of determining the sampling samples from all samples according to the preset sampling rules is returned. It can be seen that when the method passes the sampling inspection, the detection is performed at intervals of preset periods without the need for real-time detection, and the operability is high. When the sampling inspection fails, the sampling width and severity rules are adjusted and the detection continues, which effectively guarantees the reliability of the sampling inspection and the safety of driving.

Description

Technical Field

[0001] The present application relates to the field of rail transit, and in particular to a decision-making method and device for maintenance of a rail transit vehicle running gear. Background Art

[0002] At present, subway companies generally adopt a maintenance mode that combines preventive maintenance with temporary maintenance, and generally focus on preventive maintenance such as daily inspection, biweekly inspection, monthly inspection, and scheduled maintenance. This traditional regular scheduled maintenance mode is prone to excessive maintenance. Frequent train parking and maintenance directly reduce the train's online rate, increase train operating costs, and also consume huge manpower costs. Condition maintenance refers to a method of judging equipment abnormalities, predicting equipment failures, and performing maintenance before failures occur based on equipment status information provided by advanced condition monitoring and diagnostic technologies. That is, arranging maintenance plans and implementing equipment maintenance based on the health status of the equipment. It is a new maintenance mode that is different from traditional regular scheduled maintenance. Condition maintenance based on big data support is the current development trend of the rail transit industry.

[0003] The current condition maintenance is to install fixed sensors and diagnostic instruments on the train, collect a large amount of data generated by each subsystem during operation in real time, analyze the data to determine the train operation status, and guide maintenance according to the train operation status. It can be seen that the current condition maintenance of trains on the entire line requires real-time monitoring of all trains on the entire line, and real-time collection of a large amount of data generated by each subsystem of each vehicle during operation. Since it is very difficult to install fixed monitoring equipment on existing vehicles, this method is not feasible for existing vehicles and has low economic efficiency.

[0004] It can be seen that how to make maintenance decisions for trains more conveniently, efficiently and economically is an urgent problem to be solved by technical personnel in this field. Summary of the invention

[0005] The purpose of this application is to provide a decision-making method and device for maintenance of the running gear of a rail transit vehicle, which are used to make maintenance decisions for trains.

[0006] In order to solve the above technical problems, the present application provides a decision-making method for maintenance of a rail transit vehicle running gear, comprising:

[0007] Determine the sampling samples from all samples according to the preset sampling rules; wherein all samples include all the inspected parts of the train to be inspected;

[0008] Obtain the comprehensive decision results of the health status of the sampled samples, and determine the unqualified samples among the sampled samples according to the comprehensive decision results of the health status of the sampled samples;

[0009] Determine whether the current sampling is passed based on the number of unqualified samples;

[0010] If yes, the current sampling is terminated, and after a preset period, the process returns to the step of determining the sampling sample from all samples according to the preset sampling rule;

[0011] If not, adjust the inspection strictness rules according to the number of times the inspection fails, and return to the step of determining the inspection sample from all samples according to the preset inspection rules.

[0012] Preferably, determining the sample for sampling from all samples according to the preset sampling rule includes:

[0013] If all the inspected parts of the current train are qualified, then select random inspection samples from the inspected parts of the remaining trains except the current train; wherein the current train is one of all the trains to be inspected;

[0014] If there are unqualified parts among the inspected parts of the current train, the unqualified parts in the current train will be added to the random inspection sample.

[0015] Preferably, obtaining the comprehensive decision result of the health status of the sample includes:

[0016] The weights of the evaluation results obtained by evaluating the random inspection samples by each inspection system, as well as the weights of the various inspection values ​​obtained by the inspection system for the random inspection samples and the historical record data of the random inspection samples recorded by the maintenance information system are allocated according to the preset weight allocation rules; wherein each inspection system includes a mobile vehicle-mounted running gear inspection system, a trackside wheelset geometry inspection system and a trackside flat wheel inspection system;

[0017] The evaluation results of each detection system are integrated according to the weights of the evaluation results of each detection system, and the detection values ​​and historical record data of each detection system are integrated according to the weights of each detection value and the historical record data of the maintenance information system;

[0018] Based on the comprehensive evaluation of the fusion results, the comprehensive decision result on the health status of the sampled samples is obtained.

[0019] Preferably, the rules for adjusting the strictness of random inspection according to the number of failed random inspections include:

[0020] If the current sampling inspection fails, the sampling inspection method will be adjusted from the normal sampling inspection method to the strict sampling inspection method;

[0021] Correspondingly, if the random inspection is passed for multiple times in a row, the random inspection method will be adjusted from the stricter random inspection method to the normal random inspection method.

[0022] Preferably, if the sampling method is a normal sampling method, judging whether the current sampling is passed according to the number of unqualified samples includes:

[0023] comparing the number of unqualified samples with a preset first unqualified threshold;

[0024] If the number of unqualified samples is less than the first unqualified threshold, it is determined that the current random inspection has passed;

[0025] If the number of unqualified samples is not less than the first unqualified threshold, it is determined that the current random inspection has failed.

[0026] Preferably, if the sampling inspection method is a strict sampling inspection method, judging whether the current sampling inspection is passed according to the number of unqualified samples includes:

[0027] Comparing the number of unqualified samples with a preset second unqualified threshold; wherein the second unqualified threshold is less than the first unqualified threshold;

[0028] If the number of unqualified samples is less than the second unqualified threshold, it is determined that the current random inspection has passed;

[0029] If the number of unqualified samples is not less than the second unqualified threshold, it is determined that the current random inspection has failed.

[0030] Preferably, if the current random inspection fails, before adjusting the random inspection leniency and strictness rules according to the number of random inspection failures, the following is further included:

[0031] Detect trains containing unacceptable samples;

[0032] Correspondingly, in the next random inspection, the samples to be inspected are determined from all samples according to the preset random inspection rules, including:

[0033] The repaired unqualified samples are added to the random inspection samples, and the remaining samples of the random inspection samples are selected from trains that do not contain unqualified samples.

[0034] Preferably, if the current random inspection fails, it also includes:

[0035] If a train fails the random inspection twice in a row or fails to pass the random inspection multiple times in total, the random inspection will be suspended and all trains that have not been maintained will be inspected.

[0036] Preferably, before detecting all trains that have not been repaired, the method further comprises:

[0037] Determine the fault level of the detected component in the unmaintained train according to the comprehensive decision result of the health status of the detected component in the unmaintained train;

[0038] Calculating the failure rate of the inspected components according to the failure levels of the inspected components in the train that has not been repaired;

[0039] If the failure rate is greater than the preset failure rate threshold, the preset cycle is reset.

[0040] In order to solve the above technical problems, the present application also provides a decision-making device for maintenance of a running gear of a rail transit vehicle, comprising:

[0041] A determination module, used to determine the sampling samples from all samples according to the preset sampling rules; wherein all samples include all the inspected parts of the train to be inspected;

[0042] An acquisition module is used to obtain the comprehensive decision results of the health status of the sampled samples, and determine the unqualified samples in the sampled samples according to the comprehensive decision results of the health status of the sampled samples;

[0043] The judgment module is used to judge whether the current sampling has passed according to the number of unqualified samples; if so, the end module is triggered; if not, the adjustment module is triggered;

[0044] The end module is used to end the current sampling inspection and trigger the determination module after a preset period of time;

[0045] The adjustment module is used to adjust the inspection strictness rules according to the number of failed inspections and trigger the determination module.

[0046] The present application provides a decision-making method for the maintenance of the running gear of a rail transit vehicle, which method determines the sampling samples from all samples according to the preset sampling rules; wherein all samples include the inspected components of all trains to be inspected; obtains the comprehensive decision results of the health status of the sampling samples, and determines the unqualified samples in the sampling samples according to the comprehensive decision results of the health status of the sampling samples; judges whether the current sampling is passed according to the number of unqualified samples; if so, ends the current sampling, and after a preset interval period, enters the step of determining the sampling samples from all samples according to the preset sampling rules; if not, adjusts the sampling width and severity rules according to the number of sampling failures, and returns to the step of determining the sampling samples from all samples according to the preset sampling rules. It can be seen that the method selects the sampling samples for detection, does not need to detect the inspected components of all trains to be detected, and makes decisions according to the number of unqualified samples. When the current sampling is passed, the detection is performed at intervals of preset periods, and real-time detection is not required. The operability is high, and when the current sampling is not passed, the sampling width and severity rules are adjusted and the detection is continued, which effectively guarantees the reliability of the sampling and the safety of driving.

[0047] In addition, the decision-making device for maintenance of the running gear of a rail transit vehicle provided in the present application corresponds to the decision-making method for maintenance of the running gear of a rail transit vehicle, and the effects are as above. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0049] Figure 1 A flowchart of a decision-making method for maintenance of a rail transit vehicle running gear provided in this application;

[0050] Figure 2 A schematic diagram of a multi-source heterogeneous data fusion evaluation decision model provided in this application;

[0051] Figure 3 A flow chart for adjusting the sampling width and strictness provided for this application;

[0052] Figure 4 A flowchart of a decision-making method for maintenance of a rail transit vehicle running gear provided in this application;

[0053] Figure 5 A structural diagram of a decision-making device for maintenance of a rail transit vehicle running gear provided in this application. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0055] The core of this application is to provide a decision-making method, device and medium for the maintenance of the running gear of a rail transit vehicle, which are used to comprehensively evaluate the health status of the train and make maintenance decisions.

[0056] At present, most of the existing rail transit trains are not designed with sensors and diagnostic instruments installed in the running gear when they leave the factory. If they want to achieve status maintenance, they need to be modified. The modification of existing vehicles usually involves welding, punching and other operations, which will cause a certain degree of damage to the structural strength of the components. In addition, the modification of existing vehicles requires the installation of equipment inside and under the vehicle, involving a lot of wiring work. Due to the space limitations of the vehicle, the modification is very difficult. In order to overcome the above shortcomings, the present application uses a mobile vehicle-mounted running gear detection system, which includes: a train host, a train extension, a preprocessor and a composite sensor, wherein the train host and the train extension form a network through a train bus, and the preprocessor is connected to the composite sensor. If the components to be detected include bearings and wheels, a composite sensor can be installed at the bearing according to the detection requirements to detect the vibration impact and temperature information of each measuring point on the bearing. In the specific implementation, the surface of the detected component should be cleaned first, and then the composite sensor should be fixed by magnetic attraction and gluing. After the installation is completed, debugging and data collection should be carried out during the skylight time. After the collection is completed, the composite sensor should be removed, and the status maintenance decision should be made through temporary installation, without the need to modify the existing trains.

[0057] Figure 1 A flowchart of a decision-making method for maintenance of a rail transit vehicle running gear provided in this application, such as Figure 1 As shown, the method includes:

[0058] S1: Determine a sampling sample from all samples according to a preset sampling rule; wherein all samples include all inspected components of the train to be inspected.

[0059] All trains to be inspected refer to all trains on the same rail transit operating line, and the inspected parts refer to the running gear of the train, including bearings, wheels and other parts. The inspection samples refer to the inspected parts of the trains selected from all trains to be inspected according to the preset inspection rules. In a specific implementation, a certain number of trains can be randomly selected from all trains to be inspected, and the inspected parts of the selected trains can be used as inspection samples. Alternatively, a train can be firstly selected and the inspection can be carried out according to the comprehensive decision result of the health status of the train. For example, if the inspected parts of the selected trains are all qualified, the inspection samples are selected from the remaining trains, and if there are unqualified parts among the inspected parts of the selected trains, the unqualified parts of the train are added to the inspection samples. This embodiment does not limit the inspection rules, and the inspection rules can be defined by oneself in the specific implementation.

[0060] S2: Obtain the comprehensive decision results of the health status of the sampled samples, and determine the unqualified samples in the sampled samples according to the comprehensive decision results of the health status of the sampled samples.

[0061] The comprehensive decision result of the health status includes the health status of the detected components and the corresponding maintenance suggestions, among which the health status of the detected components can be divided into five levels: healthy, sub-healthy, minor fault, moderate fault and serious fault. The detected components in sub-healthy or minor fault state refer to the detected components that have fault characteristics but do not meet the definition of unqualified, that is, the detected components whose health status does not affect the safe operation for the time being, and do not need to be processed immediately. Therefore, if the health status of the sample is healthy, sub-healthy or minor fault, the sample can be judged as a qualified sample, otherwise the sample is judged as an unqualified sample. It is worth noting that since the detected components with sub-health and minor faults have the risk of developing into serious faults, in order to reduce the risk, it is necessary to track and test the above two types of detected components regularly, such as using a mobile vehicle-mounted running gear detection system to keep a focus on them, and carry out centralized maintenance in combination with the maintenance plan. It is understandable that in a specific implementation, users can classify the health level of the detected components by themselves. For example, the health level of the detected components can be divided into only three levels: healthy, sub-healthy and faulty. Obviously, if the divided health levels are different, the way of judging whether the detected components are qualified will be different.

[0062] S3: Determine whether the current sampling is passed based on the number of unqualified samples; if so, proceed to S4; if not, proceed to S5.

[0063] In a specific implementation, if the number of unqualified samples does not exceed the preset unqualified threshold, it can be determined that the current random inspection has been passed, otherwise it is determined that the current random inspection has not been passed. It should be noted that the unqualified threshold is positively correlated with the number of random inspection samples. The more random inspection samples there are, the larger the unqualified threshold is.

[0064] S4: End the current sampling inspection and return to S1 after a preset period.

[0065] The preset period can be a preset mileage or a preset time, which is not limited in this embodiment. If the current random inspection is passed, it can be considered that all trains to be inspected are in a healthy state at the current time, and the current random inspection can be ended. In addition, to ensure the safety of the train, after the preset period, it is necessary to enter step S1 again for a new round of random inspection.

[0066] S5: Adjust the inspection strictness rules according to the number of failed inspections and return to S1.

[0067] Adjusting the inspection width and strictness rules includes adjusting the unqualified threshold, the method of extracting samples, and the interval period of each round of inspection. For example, if the current inspection fails, then in the next inspection, in order to ensure the accuracy of the inspection, the inspection samples should be selected from the remaining samples except the qualified samples of the current inspection, and the unqualified threshold used should be less than the unqualified threshold of the current inspection, that is, relative to the current inspection, the next inspection should be carried out in a stricter manner. If multiple inspections fail cumulatively, the interval period of each round of inspections should be reduced to reduce the risk of train failure by increasing the number of inspections. It is understandable that if the current inspection fails, after the unqualified samples in the current inspection are repaired, in order to ensure driving safety, the next inspection should be started immediately until the inspection passes.

[0068] This embodiment provides a decision-making method for the maintenance of the running gear of a rail transit vehicle, which method determines the sampling samples from all samples according to the preset sampling rules; wherein all samples include the inspected components of all trains to be inspected; obtains the comprehensive decision results of the health status of the sampling samples, and determines the unqualified samples in the sampling samples according to the comprehensive decision results of the health status of the sampling samples; judges whether the current sampling is passed according to the number of unqualified samples; if so, the current sampling is ended, and after a preset period of interval, enters the step of determining the sampling samples from all samples according to the preset sampling rules; if not, the sampling width and severity rules are adjusted according to the number of sampling failures, and returns to the step of determining the sampling samples from all samples according to the preset sampling rules. It can be seen that the method selects the sampling samples for detection, does not need to detect the inspected components of all trains to be detected, and makes decisions according to the number of unqualified samples. When the current sampling is passed, the detection is performed at intervals of preset periods, and real-time detection is not required. The operability is high. When the current sampling fails, the sampling width and severity rules are adjusted and the detection continues, which effectively guarantees the accuracy of the decision.

[0069] On the basis of the above embodiment, in order to ensure the reliability of random inspection, this embodiment provides a preferred random inspection method, according to which the train to be inspected is determined from all trains to be inspected. This step includes:

[0070] If all the inspected parts of the current train are qualified, then select random inspection samples from the inspected parts of the remaining trains except the current train; wherein the current train is one of all the trains to be inspected;

[0071] If there are unqualified parts among the inspected parts of the current train, the unqualified parts in the current train will be added to the random inspection sample.

[0072] In this embodiment, the current train is the train currently being inspected by the mobile vehicle-mounted running gear inspection system. In a specific implementation, the mobile vehicle-mounted running gear inspection system can be used to inspect any one of all trains to be inspected. In addition, if the current random inspection fails, when the next random inspection is started, the unqualified samples of the current random inspection should be added to the random inspection samples of the next random inspection after maintenance, that is, the random inspection samples of the next random inspection include the unqualified samples of the current random inspection and the remaining samples, wherein the remaining samples are randomly selected from the trains that do not contain the unqualified samples of the current random inspection.

[0073] On the basis of the above-mentioned embodiment, in order to ensure the accuracy of the comprehensive decision result of health status, this embodiment integrates the evaluation results obtained by evaluating the sampled samples by multiple detection systems, and integrates the detection values ​​of multiple detection systems and the historical record data of the maintenance information system, so as to obtain the comprehensive decision result of health status of the sampled samples based on the comprehensive evaluation of the fusion results. This step includes:

[0074] The weights of the evaluation results obtained by evaluating the random inspection samples by each inspection system, as well as the weights of the various inspection values ​​obtained by the inspection system for the random inspection samples and the historical record data of the random inspection samples recorded by the maintenance information system are allocated according to the preset weight allocation rules; wherein each inspection system includes a mobile vehicle-mounted running gear inspection system, a trackside wheelset geometry inspection system and a trackside flat wheel inspection system;

[0075] The evaluation results of each detection system are integrated according to the weights of the evaluation results of each detection system, and the detection values ​​and historical record data of each detection system are integrated according to the weights of each detection value and the historical record data of the maintenance information system;

[0076] Based on the comprehensive evaluation of the fusion results, the comprehensive decision result on the health status of the sampled samples is obtained.

[0077] Figure 2 A schematic diagram of a multi-source heterogeneous data fusion evaluation decision model provided for this application, such as Figure 2As shown, the trackside running gear detection system includes a trackside wheelset geometry detection system and a trackside flat wheel detection system. The detection values ​​of each detection system and the historical record data of the maintenance information system are the input data of the model. Each detection system can obtain an evaluation result by judging its own detection value respectively, and then the weights of the evaluation results of each detection system and the weights of the detection values ​​of each detection system and the weights of the historical record data recorded in the maintenance information system are allocated according to the preset weight allocation principle. At this time, the evaluation results of each detection system can be fused according to the weights of the evaluation results of each detection system, and the detection values ​​and historical record data of each detection system can be fused according to the weights of the detection values ​​of each detection system and the historical record data of the maintenance information system. Therefore, a comprehensive judgment can be made based on the data fusion result, and a comprehensive decision result of the health status can be output. In addition, in Figure 2 In the data, wheel tread failure includes peeling, abrasion, loss of roundness and internal defects. The QR value refers to the horizontal distance between the intersection of the vertical line 10mm above the rolling circle tread reference line and the inner measure of the wheel rim and the intersection of the vertical line 2mm below the top of the wheel rim and the inner measure of the wheel rim. The impact value and trend, vibration effective value and trend, vibration peak-to-peak value and trend and vibration average value and trend measured by the mobile vehicle-mounted running gear detection system are all measured from the bearings, while the various detection values ​​measured by the trackside flat wheel detection system are all measured from the wheels.

[0078] The trackside wheelset geometry detection system is used to complete the measurement of parameters such as wheel flange height, wheel flange width, tread circumferential wear, vertical wear, rim thickness, wheel diameter, QR value, wheelset inner distance, etc. during the operation of the train. The trackside flat wheel detection system is used to detect the running gear bearing temperature, tread abrasion, motor and gearbox temperature and flat wheel detection of the train passing through the detection area. The maintenance information system is used to record the relevant data of each maintenance of the train, that is, the historical record data. The trackside wheelset geometry detection system, the trackside flat wheel detection system and the maintenance information system are all existing systems, so this embodiment does not make specific descriptions for them.

[0079] In the specific implementation, the detection data of each detection system should be cleaned first, including cleaning irrelevant data, repeated data and smooth noise data in the detection data, and eliminating data in the detection data that affect the data fusion effect, such as outliers, missing values, etc. Then, the weights of the evaluation results and detection values ​​of each system reflecting the health status of the components are allocated according to the preset weight allocation rules. Specifically, the weights of the evaluation results of each system can be allocated according to the application effect of each system. For example, the weight of the evaluation results of the system with higher accuracy of the measured data should be greater than the weight of the evaluation results of the system with lower accuracy of the measured data; and the weights of each detection value of each system can be allocated according to the application experience of each system. For example, if the cumulative application data of the mobile vehicle-mounted running gear detection system is greater than that of the trackside running gear detection system nationwide, the weight of the detection value of the mobile vehicle-mounted running gear detection system should be greater than the weight of the detection value of the trackside running gear detection system. In order to improve the accuracy of the comprehensive evaluation of the model, in addition to the detection values ​​of each system, the historical record data of the random inspection samples in the maintenance information system can also be integrated. Therefore, the weight allocation includes not only the allocation of the detection values ​​of each system, but also the weight allocation of the historical record data of the random inspection samples.

[0080] It should be noted that in the specific implementation, in order to further improve the accuracy of the comprehensive evaluation of the model, a multi-level weighting method can be adopted, and the sum of the weights of each level is 1, including weighting the evaluation results of each system, weighting the different types of detection values ​​in each system, and extracting and weighting the different features of the same type of detection values ​​in each system, which can effectively solve the problem of data structure alienation caused by differences between systems. In addition, the weight allocation can be dynamically adjusted according to the actual application effect, the working status of each system can be supervised, and the accuracy of the detection of each system can be examined.

[0081] After the weight allocation is completed, data fusion can be performed according to the allocated weights to obtain the comprehensive decision result of the health status. Among them, the calculation formula of the comprehensive decision result K of the health status should satisfy formula (1):

[0082]

[0083] In formula (1), Pi is the weight of the evaluation result of each system, Pij is the weight of the detection value or historical record data of each system, Aij is the state data and historical record data of the detected components of the train to be detected corresponding to each system; i is the number of systems. In this embodiment, since the system includes four systems, namely, a mobile on-board running gear detection system, a trackside flat wheel detection system, a trackside wheelset geometry detection system and a maintenance information system, m is taken as 4; j is the type of detection value of each system. For example, in FIG2 , the mobile on-board running gear detection system data has five detection values. Therefore, for this system, n can be taken as 5.

[0084] This embodiment provides a preferred data fusion method, which combines multi-source data such as mobile vehicle-mounted running gear detection system data, trackside running gear detection system data and historical record data of the maintenance information system, solves the problems of "information islands" and "inability to fully utilize data", and fully guarantees the accuracy of the comprehensive decision-making results of the health status of the random inspection samples.

[0085] On the basis of the above embodiment, in order to ensure the reliability and economy of random inspection, this embodiment provides a supplementary explanation of the random inspection method. This step includes:

[0086] If the current random inspection fails, the random inspection method will be adjusted from normal random inspection to strict random inspection;

[0087] Correspondingly, if the random inspection is passed for multiple consecutive times, the random inspection method will be adjusted from stricter random inspection to normal random inspection.

[0088] In the specific implementation, if the current random inspection fails, the inspection method should be adjusted to stricter inspection to improve the reliability of the random inspection. When stricter inspection is adopted, after the number of passed inspections reaches a certain number, in order to save the inspection costs, the inspection method can be relaxed, that is, the normal inspection method can be adopted again to determine whether the inspection has passed.

[0089] Based on the above embodiment, this embodiment provides a supplementary explanation of the normal sampling method. This step includes:

[0090] comparing the number of unqualified samples with a preset first unqualified threshold;

[0091] If the number of unqualified samples is less than the first unqualified threshold, it is determined that the current random inspection has passed;

[0092] If the number of unqualified samples is not less than the first unqualified threshold, it is determined that the current random inspection has failed.

[0093] The first unqualified threshold is the unqualified threshold for normal random inspection. In specific implementation, the size of the first unqualified threshold can be set according to actual conditions. In addition, to ensure the reliability of random inspection, the first unqualified threshold should be much smaller than the total number of random inspection samples.

[0094] This embodiment provides a method for determining whether a random inspection has passed. When the random inspection is carried out normally, this method can be used to determine whether the random inspection has passed, thereby ensuring the economy of the random inspection while also effectively ensuring the reliability of the random inspection.

[0095] Based on the above embodiment, this embodiment provides additional explanations on the strict sampling inspection method. This step includes:

[0096] Comparing the number of unqualified samples with a preset second unqualified threshold; wherein the second unqualified threshold is less than the first unqualified threshold;

[0097] If the number of unqualified samples is less than the second unqualified threshold, it is determined that the current random inspection has passed;

[0098] If the number of unqualified samples is not less than the second unqualified threshold, it is determined that the current random inspection has failed.

[0099] The second unqualified threshold is the unqualified threshold for the stricter random inspection. Therefore, the second unqualified threshold should be less than the first unqualified threshold and can be defined in the specific implementation. This embodiment provides another way to determine whether the random inspection has passed. When the random inspection is stricter, this method can be used to determine whether the random inspection has passed, thereby improving the reliability of the random inspection and effectively ensuring the safety of the train.

[0100] Based on the above embodiment, if the current random inspection fails, in order to ensure driving safety, the unqualified samples in the current random inspection need to be inspected and repaired so as to conduct the next random inspection. This step includes:

[0101] Detect trains containing unacceptable samples;

[0102] Correspondingly, in the next random inspection, the samples to be inspected are determined from all samples according to the preset random inspection rules, including:

[0103] The repaired unqualified samples are added to the random inspection samples, and the remaining samples of the random inspection samples are selected from trains that do not contain unqualified samples.

[0104] Specifically, if the current random inspection fails, it means that the current number of unqualified samples is greater than the unqualified threshold. In order to ensure driving safety, the train containing the unqualified samples needs to be inspected and repaired; in the next random inspection, the unqualified samples after maintenance will be added to the random inspection samples, and the remaining samples of the random inspection samples will be selected from the trains that do not contain unqualified samples to further confirm whether the unqualified samples after maintenance are qualified samples in this random inspection, thereby improving the reliability of the random inspection.

[0105] Based on the above embodiment, if the current random inspection fails, in order to ensure the safety of the vehicle, it can be further determined whether the random inspection needs to be suspended. This step includes:

[0106] If a train fails the random inspection twice in a row or fails to pass the random inspection multiple times in total, the random inspection will be suspended and all trains that have not been maintained will be inspected.

[0107] In the specific implementation, if the first random inspection fails, the second random inspection will adopt a stricter inspection method. If the second random inspection also fails, that is, two consecutive random inspections fail, in order to ensure the safety and reliability of train operation, the random inspection should be suspended and all untested trains should be inspected. In addition, if the number of random inspection failures within a preset period of multiple consecutive rounds of trains reaches a certain value, the random inspection should also be suspended and a full inspection of all untested trains should be carried out.

[0108] Figure 3 A flow chart for adjusting the sampling width and strictness provided for this application is as follows: Figure 3 As shown in the figure, after the full inspection of all uninspected trains on the entire line is completed, the random inspection samples can be re-drawn, and the comprehensive decision results of the health status of the random inspection samples can be obtained for evaluation and analysis. If the evaluation is qualified, it means that it has been approved and can be re-inspected. To ensure the safety of the train, a stricter sampling method should be used to determine whether it has passed the random inspection; after passing the stricter inspection twice in a row, in order to save the inspection fee, the random inspection can be relaxed and the normal sampling method can be used for random inspection. It is worth noting that Figure 3 The number of random inspections involved is the number of random inspections during the entire life cycle of the train.

[0109] This embodiment provides additional explanation for the situation where the vehicle fails two consecutive inspections or multiple inspections cumulatively, and can ensure driving safety while taking into account economy.

[0110] Based on the above embodiment, if the sample failure rate is high, that is, the sampling inspection according to the current preset period cannot effectively ensure the safety of the train, then the preset period can be adjusted according to the failure rate. This step includes:

[0111] Determine the fault level of the detected component in the unmaintained train according to the comprehensive decision result of the health status of the detected component in the unmaintained train;

[0112] Calculating the failure rate of the inspected components according to the failure levels of the inspected components in the train that has not been repaired;

[0113] If the failure rate is greater than the preset failure rate threshold, the preset cycle is reset.

[0114] In a specific implementation, the unqualified samples of the detected components in the unrepaired train can be determined according to the fault level of the detected components in the unrepaired train, and the unqualified rate can be calculated according to the number of unqualified samples of the detected components and the total number of random inspection samples. If the unqualified rate is greater than the preset unqualified rate threshold, it is necessary to re-formulate the preset cycle, which is specifically manifested as reducing the preset cycle. Among them, the preset unqualified rate threshold is set according to the actual situation and can be between 0 and 5%.

[0115] Figure 4 The present application provides a flow chart of a decision-making method for maintenance of a rail transit vehicle running gear, such as Figure 4 As shown, the total mileage is 1.5 million kilometers, and the preset period is 150,000 kilometers. Specifically, if a medium fault or serious fault alarm occurs before 1.5 million kilometers, such as a bearing level 2 alarm, the whole line random inspection is directly started; and if no medium fault or serious fault alarm occurs before 1.5 million kilometers, the random inspection is carried out according to the preset period. It is understandable that the mileage of the first random inspection can be flexibly adjusted according to the historical random inspection results. For example, if the historical random inspection results are good, in order to save the random inspection costs, the mileage of the first random inspection can be increased to reduce the number of random inspections. In addition, in order to ensure the reliability of the random inspection, the number of random inspections can be increased. For example, when the first train on the entire line reaches 1.35 million kilometers, it can be inspected multiple times to ensure that all trains on the entire line are in normal condition when the next execution is carried out.

[0116] This embodiment calculates the unqualified rate of the inspected components. When the unqualified rate is greater than a preset unqualified rate threshold, the preset period is re-established, specifically by reducing the preset period, that is, ensuring the safety of the train by increasing the number of random inspections.

[0117] In the above embodiments, the decision-making method for the maintenance of the running gear of a rail transit vehicle is described in detail, and the present application also provides an embodiment corresponding to the decision-making device for the maintenance of the running gear of a rail transit vehicle. It should be noted that the present application describes the embodiments of the device part from two perspectives, one is based on the perspective of the functional module, and the other is based on the perspective of the hardware.

[0118] Figure 5 A structural diagram of a decision-making device for maintenance of a rail transit vehicle running gear provided in this application, such as Figure 5 As shown, the device comprises:

[0119] The determination module 10 is used to determine the sample to be inspected from all samples according to the preset inspection rule; wherein all samples include all the inspected components of the train to be inspected;

[0120] The acquisition module 11 is used to obtain the comprehensive decision results of the health status of the sampled samples, and determine the unqualified samples in the sampled samples according to the comprehensive decision results of the health status of the sampled samples;

[0121] The judging module 12 is used to judge whether the current sampling has passed according to the number of unqualified samples; if so, the ending module 13 is triggered; if not, the adjusting module 14 is triggered;

[0122] Ending module 13, used to end the current sampling inspection and trigger the determination module 10 after a preset period;

[0123] The adjustment module 14 is used to adjust the inspection strictness rules according to the number of failed inspections and trigger the determination module 10.

[0124] Since the embodiments of the apparatus part correspond to the embodiments of the method part, please refer to the description of the embodiments of the method part for the embodiments of the apparatus part, which will not be repeated here.

[0125] The decision-making device for the maintenance of the running parts of rail transit vehicles provided in this embodiment determines the sampling samples from all samples according to the preset sampling rules through the determination module; wherein all samples include the inspected components of all trains to be inspected; through the acquisition module, the comprehensive decision results of the health status of the sampling samples are obtained, and the unqualified samples in the sampling samples are determined according to the comprehensive decision results of the health status of the sampling samples; through the judgment module, it is judged whether the current sampling inspection is passed according to the number of unqualified samples; if so, the end module is triggered; if not, the adjustment module is triggered; through the end module, the current sampling inspection is ended, and after the interval preset period, the determination module is triggered; through the adjustment module, the sampling inspection width and severity rules are adjusted according to the number of sampling failures, and the determination module is triggered. It can be seen that the device selects sampling samples for detection, does not need to detect the inspected components of all trains to be detected, and makes decisions according to the number of unqualified samples. When the current sampling inspection is passed, the detection is performed at intervals of preset periods, and real-time detection is not required. The operability is high, and when the current sampling inspection is not passed, the sampling inspection width and severity rules are adjusted and the detection is continued, which effectively guarantees the reliability of the sampling inspection and the safety of driving.

[0126] The above is a detailed introduction to the decision-making method and device for the maintenance of the running gear of a rail transit vehicle provided by the present application. The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same and similar parts between the embodiments can be referenced to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.

[0127] It should also be noted that, in this specification, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device including the element.

Claims

1. A decision-making method for maintenance of rail transit vehicle running gear, characterized in that: include: Determine the sampling samples from all samples according to the preset sampling rules; wherein all the samples include all the inspected parts of the train to be inspected; Obtaining a comprehensive decision result of the health status of the sampled samples, and determining unqualified samples among the sampled samples according to the comprehensive decision result of the health status of the sampled samples; Determine whether the current sampling is passed according to the number of the unqualified samples; If yes, then the current sampling inspection is terminated, and after a preset period, the process returns to the step of determining the sampling inspection sample from all samples according to the preset sampling inspection rule; If not, adjust the inspection strictness rule according to the number of failed inspections, and return to the step of determining the inspection sample from all samples according to the preset inspection rule; Correspondingly, determining the sampling sample from all samples according to the preset sampling rule includes: If all the inspected parts of the current train are qualified, the sampling samples are selected from the inspected parts of the remaining trains except the current train; wherein the current train is one of all the trains to be inspected; If there are unqualified parts among the inspected parts of the current train, the unqualified parts in the current train are added to the random inspection sample; Correspondingly, the obtaining of the comprehensive decision result of the health status of the sample includes: The weights of the evaluation results obtained by evaluating the sampled samples by each detection system, as well as the weights of the detection values ​​obtained by detecting the sampled samples by each detection system and the historical record data of the sampled samples recorded by the maintenance information system are allocated according to the preset weight allocation rules; wherein each detection system includes a mobile vehicle-mounted running gear detection system, a trackside wheelset geometry detection system and a trackside flat wheel detection system; fusing the evaluation results of each of the detection systems according to the weights of the evaluation results of each of the detection systems, and fusing the detection values ​​of each of the detection systems and the historical record data of the maintenance information system according to the weights of the detection values ​​of each of the detection systems and the historical record data; Obtaining a comprehensive decision result of the health status of the sample according to a comprehensive evaluation of the fusion result; Correspondingly, the rules for adjusting the severity of random inspections according to the number of failed random inspections include: If the current sampling inspection fails, the sampling inspection method will be adjusted from the normal sampling inspection method to the strict sampling inspection method; Correspondingly, if the random inspection is passed for multiple times in succession, the random inspection mode is adjusted from the stricter random inspection mode to the normal random inspection mode.

2. The decision-making method for rail transit vehicle running gear maintenance according to claim 1, characterized in that: If the sampling inspection method is the normal sampling inspection method, then judging whether the current sampling inspection is passed according to the number of unqualified samples includes: Comparing the number of the unqualified samples with a preset first unqualified threshold; If the number of unqualified samples is less than the first unqualified threshold, it is determined that the current random inspection has passed; If the number of unqualified samples is not less than the first unqualified threshold, it is determined that the current random inspection has failed.

3. The decision-making method for rail transit vehicle running gear maintenance according to claim 2, characterized in that: If the sampling inspection method is the strict sampling inspection method, then judging whether the current sampling inspection is passed according to the number of unqualified samples includes: Comparing the number of the unqualified samples with a preset second unqualified threshold; wherein the second unqualified threshold is less than the first unqualified threshold; If the number of unqualified samples is less than the second unqualified threshold, it is determined that the current random inspection has passed; If the number of the unqualified samples is not less than the second unqualified threshold, it is determined that the current random inspection has failed.

4. The decision-making method for rail transit vehicle running gear maintenance according to claim 1, characterized in that: If the current random inspection fails, before the random inspection leniency and strictness rules are adjusted according to the number of random inspection failures, the following are also included: detecting a train containing the unqualified sample; Correspondingly, during the next random inspection, determining the random inspection sample from all samples according to the preset random inspection rule includes: The unqualified samples after maintenance are added to the random inspection samples, and the remaining samples of the random inspection samples are selected from trains that do not contain the unqualified samples.

5. The decision-making method for rail transit vehicle running gear maintenance according to claim 4, characterized in that: If the current random inspection fails, it also includes: If the train fails the random inspection for two consecutive times or fails the random inspection for multiple times in total, the random inspection will be suspended and all trains that have not been repaired will be inspected.

6. The decision-making method for rail transit vehicle running gear maintenance according to claim 5, characterized in that: Before the detection of all trains that have not been repaired, the method further includes: Determining the fault level of the detected component in the train that has not been maintained according to the comprehensive decision result of the health status of the detected component in the train that has not been maintained; Calculating the failure rate of the detected components according to the fault levels of the detected components in the train that has not been repaired; If the failure rate is greater than a preset failure rate threshold, the preset period is reset.

7. A decision-making device for maintenance of rail transit vehicle running gear, characterized in that: include: A determination module, used to determine the sample to be inspected from all samples according to a preset inspection rule; wherein all the samples include all the inspected components of the train to be inspected; An acquisition module, used to acquire a comprehensive decision result of the health status of the sampled samples, and determine unqualified samples among the sampled samples according to the comprehensive decision result of the health status of the sampled samples; A judgment module, used to judge whether the current sampling is passed according to the number of the unqualified samples; if so, trigger the end module; if not, trigger the adjustment module; The ending module is used to end the current sampling inspection and trigger the determining module after a preset period; The adjustment module is used to adjust the inspection strictness rule according to the number of failed inspections and trigger the determination module; Correspondingly, determining the sampling sample from all samples according to the preset sampling rule includes: If all the inspected parts of the current train are qualified, the sampling samples are selected from the inspected parts of the remaining trains except the current train; wherein the current train is one of all the trains to be inspected; If there are unqualified parts among the inspected parts of the current train, the unqualified parts in the current train are added to the random inspection sample; Correspondingly, the obtaining of the comprehensive decision result of the health status of the sample includes: The weights of the evaluation results obtained by evaluating the sampled samples by each detection system, as well as the weights of the detection values ​​obtained by detecting the sampled samples by each detection system and the historical record data of the sampled samples recorded by the maintenance information system are allocated according to the preset weight allocation rules; wherein each detection system includes a mobile vehicle-mounted running gear detection system, a trackside wheelset geometry detection system and a trackside flat wheel detection system; fusing the evaluation results of each of the detection systems according to the weights of the evaluation results of each of the detection systems, and fusing the detection values ​​of each of the detection systems and the historical record data of the maintenance information system according to the weights of the detection values ​​of each of the detection systems and the historical record data; Obtaining a comprehensive decision result of the health status of the sample according to a comprehensive evaluation of the fusion result; Correspondingly, the rules for adjusting the severity of random inspections according to the number of failed random inspections include: If the current sampling inspection fails, the sampling inspection method will be adjusted from the normal sampling inspection method to the strict sampling inspection method; Correspondingly, if the random inspection is passed for multiple times in succession, the random inspection mode is adjusted from the stricter random inspection mode to the normal random inspection mode.

Images