Method for constructing a model for evaluating the braking performance of a railway vehicle and related device

Abstract

The application provides a method for constructing a rail vehicle braking performance evaluation model and related equipment. The method includes: collecting a training sample data set for training the braking performance evaluation model, the training sample data set including a plurality of deceleration process sub-data sets, each deceleration process sub-data set including a plurality of characteristic parameters of the rail vehicle in a braking process; labeling the plurality of deceleration process sub-data sets in the training sample data set using the deceleration of each deceleration process sub-data set, the label including normal braking and abnormal braking; and training the braking performance evaluation model based on the labeled training sample data set, the trained braking performance evaluation model being used to evaluate whether the braking performance of the rail vehicle is abnormal. The application realizes evaluation and judgment of the braking capability of the whole rail vehicle by analyzing the characteristics of the braking system and constructing the braking performance evaluation model, meets the daily vehicle operation and maintenance requirements, and improves the normal operation rate of the braking system.

Description

Technical Field

[0001] This invention relates to the field of rail vehicle safety inspection technology, and in particular to a method for constructing a rail vehicle braking performance evaluation model, an evaluation system, electronic equipment, and a readable storage medium. Background Technology

[0002] As a core system of rail transit vehicles, the braking system plays a vital role in ensuring vehicle operation safety and passenger safety and comfort. Therefore, research on the health status assessment of the braking system and the status monitoring of train braking capacity is particularly important.

[0003] Currently, braking systems are mainly tested through appearance and functionality. This method relies on human observation and simple threshold judgment, without fully considering the influence of external environment and weather factors, and without evaluating the braking capacity of trains, thus lacking strong applicability.

[0004] In view of this, the present invention is hereby proposed. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome at least some of the shortcomings of the prior art and provide a method for constructing a rail vehicle braking performance evaluation model. By conducting characteristic analysis of the rail vehicle braking system and constructing a rail vehicle braking performance evaluation model, the braking capacity of the entire rail vehicle can be evaluated and judged, so as to meet the daily vehicle operation and maintenance needs and improve the normal operation rate of the braking system.

[0006] To solve the above-mentioned technical problems, the basic concept of the technical solution adopted by the present invention is as follows:

[0007] A method for constructing a braking performance evaluation model for rail vehicles, comprising:

[0008] A training sample dataset is collected for training the braking performance evaluation model. The training sample dataset includes multiple deceleration process sub-datasets, and each deceleration process sub-dataset includes multiple feature parameters of the rail vehicle during the braking process.

[0009] The deceleration of each deceleration process subset is used to label multiple deceleration process subsets in the training sample dataset, where the labels include normal braking and abnormal braking; and

[0010] A braking performance evaluation model is trained based on a labeled training sample dataset. The trained braking performance evaluation model is used to evaluate whether the braking performance of rail vehicles is abnormal.

[0011] In some implementations, the step of labeling multiple deceleration process subsets in the training sample dataset using the deceleration of each deceleration process subset includes:

[0012] The multiple deceleration process subsets are divided into several level data groups according to the braking level of the rail vehicle, and the average speed in each deceleration process subset in each level data group is calculated.

[0013] Based on the preset speed intervals and the calculated average speed, the multiple deceleration process subsets in each stage data group are divided into several working condition data groups, and the deceleration in each deceleration process subset within each working condition data group in each stage data group is calculated; and

[0014] Statistical analysis is performed on the calculated deceleration, and each deceleration process subset in the multiple deceleration process subsets is labeled based on the results of the statistical analysis.

[0015] In some implementations, the step of performing statistical analysis on the calculated deceleration and labeling each deceleration process subset in the plurality of deceleration process subsets based on the results of the statistical analysis includes:

[0016] Based on the calculated deceleration, draw the deceleration box plot for different working condition data groups in each level data group;

[0017] According to the preset deceleration scoring system, the average score of deceleration in different working condition data groups in each level data group is calculated based on the drawn deceleration box plot.

[0018] A box plot of the average deceleration is drawn based on the average deceleration of different working condition data groups in each level data group; and

[0019] The average binning plots are used to label the multiple deceleration process subsets included in the training sample dataset.

[0020] In some implementations, the step of training a braking performance evaluation model based on a labeled training sample dataset includes:

[0021] Based on the labeled training sample dataset, different types of classifiers are trained separately to obtain multiple classifiers whose training performance meets the preset requirements.

[0022] The multiple classifiers are integrated using a voting method to obtain an initial braking performance evaluation model; and

[0023] The initial braking performance evaluation model is trained based on the labeled training sample data to obtain the braking performance evaluation model.

[0024] In some implementations, the different types of classifiers include logistic regression classifiers, support vector machine classifiers, K-nearest neighbor classifiers, Bagging classifiers, random forest classifiers, decision tree classifiers, and extreme gradient boosting tree classifiers.

[0025] In some implementations, the step of collecting a training sample dataset for training the braking performance evaluation model includes:

[0026] Collect multiple operating condition data to characterize the working conditions of the rail vehicle and multiple braking data to characterize the performance of the rail vehicle's braking system during braking; and

[0027] Correlation analysis is performed on the multiple operating condition data and the multiple braking data to obtain the multiple characteristic parameters.

[0028] In some implementations, the step of performing correlation analysis on the plurality of operating condition data and the plurality of braking data to obtain the plurality of characteristic parameters includes:

[0029] A first correlation analysis was performed on the multiple operating condition data and the multiple braking data to obtain multiple parameters related to braking performance evaluation as feature parameters; and

[0030] A second correlation analysis is performed on the feature parameters to reduce their dimensionality.

[0031] The present invention also provides a rail vehicle braking performance evaluation system, comprising:

[0032] The data acquisition module is used to collect multiple operating condition data characterizing the working conditions of the rail vehicle and multiple braking data characterizing the performance of the rail vehicle's braking system during braking; and

[0033] The braking performance evaluation module pre-stores a braking performance evaluation model constructed according to the above-mentioned method for constructing a rail vehicle braking performance evaluation model. After receiving the operating condition data and braking data collected by the data acquisition module, the braking performance evaluation module calls the braking performance evaluation model to process the received data in order to evaluate whether the braking performance of the rail vehicle is abnormal.

[0034] The present invention also provides an electronic device, comprising:

[0035] processor;

[0036] A memory storing a computer program configured to be executed by the processor, which, when executing the computer program, implements the method for constructing a rail vehicle braking performance evaluation model as described above.

[0037] The present invention also provides a readable storage medium storing a computer program, which, when executed by a processor, implements the method for constructing the above-described rail vehicle braking performance evaluation model.

[0038] By adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art.

[0039] 1. This invention evaluates and judges the braking capacity of a rail vehicle by conducting characteristic analysis of the rail vehicle braking system and constructing a rail vehicle braking performance evaluation model, thereby meeting the daily operation and maintenance needs of the vehicle and improving the normal operation rate of the braking system.

[0040] 2. This invention performs statistical analysis on the deceleration in each deceleration process subset and labels the deceleration process subset of the rail vehicle during braking based on the results of the statistical analysis, thereby achieving the purpose of automatically generating training sample labels and improving data processing efficiency and accuracy.

[0041] 3. This invention constructs a braking performance evaluation model based on a machine learning network architecture of multi-model fusion voting, thereby improving the adaptability and robustness of the braking performance evaluation model. Attached Figure Description

[0042] The accompanying drawings, as part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments and descriptions of the invention are used to explain the invention, but do not constitute an undue limitation of the invention. Obviously, the drawings described below are merely some embodiments, and those skilled in the art can obtain other drawings based on these drawings without creative effort. In the drawings:

[0043] Figure 1 This is a flowchart illustrating a method for constructing a braking performance evaluation model for rail vehicles according to an exemplary embodiment of the present invention.

[0044] Figure 2 This is a schematic flowchart of S110 according to an exemplary embodiment of the present invention;

[0045] Figure 3 This is a flowchart illustrating step S120 according to an exemplary embodiment of the present invention;

[0046] Figure 4A and Figure 4B This is a schematic diagram of the data evolution process during the tagging process according to an exemplary embodiment of the present invention;

[0047] Figure 5 This is a flowchart illustrating step S123 according to an exemplary embodiment of the present invention;

[0048] Figure 6 This is a flowchart illustrating step S130 according to an exemplary embodiment of the present invention;

[0049] Figure 7 This is a schematic diagram of a rail vehicle braking performance evaluation system provided according to an exemplary embodiment of the present invention;

[0050] Figure 8 This is a schematic diagram of the structure of an electronic device provided according to an exemplary embodiment of the present invention.

[0051] In the diagram: 200, Rail vehicle braking performance evaluation system; 210, Data acquisition module; 220, Braking performance evaluation module; 230, Data management module;

[0052] 300. Electronic device; 301. Processor; 302. Memory; 303. Bus; 304. Communication interface.

[0053] It should be noted that these accompanying drawings and textual descriptions are not intended to limit the scope of the invention in any way, but rather to illustrate the concept of the invention to those skilled in the art by referring to specific embodiments. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments will be clearly and completely described below with reference to the accompanying drawings. The following embodiments are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0055] In the description of this invention, it should be noted that the terms "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0056] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0057] Figure 1 The flowchart illustrates a method 100 for constructing a rail vehicle braking performance evaluation model according to an exemplary embodiment of the present invention.

[0058] like Figure 1 As shown, the method 100 for constructing a rail vehicle braking performance evaluation model includes the following steps:

[0059] S110. Collect a training sample dataset for training the braking performance evaluation model. The training sample dataset includes multiple deceleration process sub-datasets, and each deceleration process sub-dataset includes multiple feature parameters of the rail vehicle during the braking process.

[0060] S120. Label multiple deceleration process subsets in the training sample dataset using the deceleration of each deceleration process subset, wherein the labels include normal braking and abnormal braking; and

[0061] S130. Train a braking performance evaluation model based on the labeled training sample dataset. The trained braking performance evaluation model is used to evaluate whether the braking performance of rail vehicles is abnormal.

[0062] It should be understood that the steps shown in the method 100 for constructing the rail vehicle braking performance evaluation model are not exclusive. The method 100 may also include additional steps not shown and / or the steps shown may be omitted. The scope of the invention is not limited in this respect. Referring below... Figures 1 to 3 Describe steps S110 to S130 in detail.

[0063] S110

[0064] Training sample datasets are an important part of machine learning, as they help machine learning systems build efficient models.

[0065] In step S110, a training sample dataset for training the braking performance evaluation model is collected. The training sample dataset includes multiple deceleration process sub-datasets, and each deceleration process sub-dataset includes multiple feature parameters of the rail vehicle during the braking process.

[0066] In some implementations, such as Figure 2 As shown, step S110 includes:

[0067] S111. Collect multiple operating condition data to characterize the working conditions of the rail vehicle and multiple braking data to characterize the performance of the rail vehicle's braking system during braking; and

[0068] S112. Perform correlation analysis on the multiple operating condition data and the multiple braking data to obtain the multiple characteristic parameters.

[0069] Specifically, in step S111, multiple operating condition data characterizing the working conditions of the rail vehicle and multiple braking data characterizing the performance of the rail vehicle's braking system during braking can be collected. In step S112, a first correlation analysis is performed on the multiple operating condition data and the multiple braking data to obtain multiple parameters related to braking performance evaluation as feature parameters. These include multiple operating condition data characterizing the working conditions of the rail vehicle, such as the type of rail vehicle, train number, car number, and formation number, as well as multiple braking data characterizing the performance of the rail vehicle's braking system, such as time, train speed, bogie air spring pressure, actual traction / electric braking force percentage, TBM output driver controller braking level, emergency braking UB, GW transfer to SBM electric braking setting, MR pressure, actual value of air braking force obtained by the train output by the TBM, maximum air braking force obtainable by the train, actual mass of the train, the train is currently sanding, percentage of traction / electric braking force applied, brake cylinder pressure, emergency braking EB, and GW transfer to TBM for main air compressor start-up of the traction unit.

[0070] Then, a second correlation analysis is performed on the feature parameters to reduce their dimensionality. For example, similar parameters with high correlation are averaged. For instance, the MR pressures of different carriages are highly correlated, so the average MR pressure across multiple carriages is taken as the MR pressure in the training sample dataset.

[0071] The first correlation analysis of the multiple operating condition data and the multiple braking data, and the second correlation analysis of the characteristic parameters, can both be performed using a correlation coefficient matrix to calculate the autocorrelation between the parameters. For example, a correlation coefficient of 0 indicates that the two parameters are not correlated; a correlation coefficient greater than 0.1 is considered to indicate a high correlation.

[0072] It should be noted that rail vehicles have multiple braking levels. When a braking level changes from 0 to another value, data sampling can begin after a certain delay. This is because there is a delay between the change in braking level and its actual effect on braking. The data collection continues until the speed reaches 0, or the braking level remains 0 for more than 3 consecutive seconds. This period is considered part of a deceleration process. If, during this process, the braking level is 0 but the duration does not exceed 3 seconds, it is not considered a deceleration process. All the aforementioned characteristic parameters need to be acquired during each deceleration process.

[0073] S120

[0074] In step S120, each deceleration process subset in the multiple deceleration process subsets is labeled.

[0075] It should be noted that the braking-related data collected in step S110 is raw data and is not directly related to the braking capability status. Therefore, in this invention, the braking performance evaluation parameter—train braking deceleration—is first constructed, and then the deceleration of each deceleration process subset is used to label multiple deceleration process subsets in the training sample dataset.

[0076] In some implementations, such as Figure 3 As shown, step S120 includes:

[0077] S121. Divide the multiple deceleration process subsets into several level data groups according to the braking level of the rail vehicle, and calculate the average speed in each deceleration process subset in each level data group.

[0078] S122. Based on the preset speed interval and the calculated average speed, divide the multiple deceleration process subsets in each level data group into several working condition data groups, and calculate the deceleration in each deceleration process subset in each working condition data group within each level data group; and

[0079] S123. Perform statistical analysis on the calculated deceleration, and label each deceleration process subset in the multiple deceleration process subsets according to the results of the statistical analysis.

[0080] Figure 4A and Figure 4B This illustrates the data evolution process during the labeling process. The following section combines... Figure 4A and Figure 4B The steps S121 to S123 described above are explained in detail.

[0081] like Figure 4A As shown, the training sample dataset includes n deceleration process subsets, each containing multiple feature parameters of the rail vehicle during braking. Assuming the rail vehicle has 7 braking levels, in step S121, the multiple deceleration process subsets are divided into 7-level data groups (B1 to B7) according to the rail vehicle's braking level. Each level data group includes at least one deceleration process subset. It is understood that the number of deceleration process subsets included in each of the B1 to B7 level data groups is determined based on actual conditions.

[0082] Furthermore, the average speed in each deceleration process subset is calculated within the seven data groups from B1 to B7. As an example, the initial speed Vinitial of the vehicle braking period and the final speed Vfinal of the data slice can be selected to calculate the average speed, for example, Vaverage = (Vinitial + Vfinal) / 2. It should be noted that the step of calculating the average speed can be performed before or after the step of dividing the data into data groups; this invention does not impose any restrictions on this.

[0083] In step S122, the multiple deceleration process subsets in each level data group are divided into V1 working condition data group to Vn working condition data group according to the preset speed interval (e.g., 5 km / h) and the calculated average speed.

[0084] Furthermore, the deceleration in each deceleration process subset within each working condition data group of each level data group is calculated. As an example, the initial speed Vinitial of the vehicle during braking, the final speed Vfinal of the data slice, and the time t taken can be selected to calculate the deceleration Va, for example, Va = (Vfinal - Vinitial) / t. It should be noted that the step of calculating the deceleration can be performed before or after the step of dividing the working condition data groups; this invention does not impose any limitation on this.

[0085] As an example, assume that the B1 level data group includes 25 deceleration process subsets. In step S121, the average speeds of these 25 deceleration process subsets are calculated to be 3 km / h, 4 km / h, 5 km / h, 6 km / h, 7 km / h, 8 km / h, 9 km / h, 10 km / h, 11 km / h, 12 km / h, 13 km / h, 14 km / h, 15 km / h, 15 km / h, 17 km / h, 18 km / h, 19 km / h, 20 km / h, 21 km / h, 22 km / h, 23 km / h, 24 km / h, 25 km / h, 26 km / h, and 27 km / h. In step S122, the above 25 deceleration process subsets are divided into V1 working condition data group to V5 working condition data group according to a speed interval of 5 km / h. The V1 test data set includes deceleration process subsets with average speeds of 3 km / h, 4 km / h, 5 km / h, 6 km / h, and 7 km / h; the V2 test data set includes deceleration process subsets with average speeds of 8 km / h, 9 km / h, 10 km / h, 11 km / h, and 12 km / h; and so on. The V5 test data set includes deceleration process subsets with average speeds of 23 km / h, 24 km / h, 25 km / h, 26 km / h, and 27 km / h. Then, calculate the deceleration Va1...Va5 for the deceleration process subsets with average speeds of 3km / h, 4km / h, 5km / h, 6km / h, and 7km / h in the V1 working condition data group, and the deceleration Va1...Va5 for the deceleration process subsets with average speeds of 8km / h, 9km / h, 10km / h, 11km / h, and 12km / h in the V2 working condition data group, and so on. Calculate the deceleration Va1...Va5 for the deceleration process subsets with average speeds of 23km / h, 24km / h, 25km / h, 26km / h, and 27km / h in the V5 working condition data group.

[0086] Multiple deceleration process subsets in the B2 to B7 level bit data groups can also be divided into several working condition data groups according to the above process, and the deceleration of each deceleration process subset can be calculated. This will not be elaborated further here. It is understood that the number of working condition data groups in the B2 to B7 level bit data groups can be determined according to actual conditions. Furthermore, the above examples are merely illustrative of the technical solutions of this invention and do not constitute a limitation on the scope of protection of this invention.

[0087] In step S123, the deceleration of the deceleration process subset calculated in each set of working condition data is statistically analyzed, and then each deceleration process subset is labeled as normal braking or abnormal braking based on the results of the statistical analysis.

[0088] In some implementations, such as Figure 5 As shown, step S123 includes:

[0089] S1231. Based on the calculated deceleration, draw the deceleration box plot of different working condition data groups in each level data group;

[0090] S1232. According to the preset deceleration scoring system, and in conjunction with the drawn deceleration box plot, calculate the average score of deceleration for different working condition data groups in each level data group.

[0091] S1233. Draw an average box plot based on the average deceleration of different working condition data groups in each level data group; and

[0092] S1234. Label the multiple deceleration process subsets included in the training sample dataset by combining the drawn average bin map.

[0093] Continue to refer to Figure 4A As shown, in step S1231, deceleration box plots are drawn for different working condition data groups within each working condition data group from B1 to B7, based on the deceleration of multiple deceleration process subsets in each working condition data group. Specifically, after calculating the deceleration of each deceleration process subset in each working condition data group, the upper edge, lower edge, and upper and lower quartiles of this data group are identified; then, the two quartiles are connected to draw the box; finally, the upper and lower edges are connected to the box to draw the deceleration box plot for that working condition data group. This process is repeated to draw deceleration box plots for all working condition data groups from B1 to B7.

[0094] In step S1232, according to the preset deceleration scoring system, the average score of deceleration for different working condition data groups in each level data group is calculated based on the drawn deceleration box plot. As an example, different scores are assigned based on the upper and lower boundaries and the upper and lower quartiles of the deceleration box plot: 0 points for deceleration less than or equal to the lower boundary, 1 point for deceleration greater than the lower boundary but less than or equal to the lower quartile, 2 points for deceleration greater than the lower quartile but less than or equal to the upper quartile, 3 points for deceleration greater than the upper quartile but less than or equal to the upper boundary, and 4 points for deceleration greater than the upper boundary. It should be noted that deceleration less than the lower boundary or greater than the upper boundary of the corresponding box plot are outliers, and corresponding scores need to be assigned to these outliers as well.

[0095] Furthermore, the average score of deceleration in each working condition data group within each level data group is calculated based on the drawn deceleration box plot, and this score is used as the final score for that working condition data group.

[0096] Reference Figure 4BAs shown, in step S1233, the final scores of all working condition data groups are summarized, and an average box girder diagram is drawn based on this summarized data. The process of drawing the average box girder diagram can refer to the process of drawing the deceleration box girder diagram in step S1231, and will not be described again here.

[0097] In step S1234, labels are assigned to multiple deceleration process subsets included in the training sample dataset based on the final scores of all working condition data groups and the drawn average binning plot. For example, when the final scores of some working condition data groups are less than the lower boundary of the average binning plot, they are considered to have weak braking performance; the rest are considered to have normal braking performance. Therefore, the deceleration process subsets in working condition data groups whose final scores are less than the lower boundary of the average binning plot are labeled as having abnormal braking, while the deceleration process subsets in the remaining working condition data groups are labeled as having normal braking. This completes the labeling of all deceleration process subsets, forming the sample labels for the braking performance evaluation model.

[0098] S130

[0099] In step S130, a braking performance evaluation model is trained based on the labeled training sample dataset.

[0100] In some implementations, such as Figure 6 As shown, step S130 includes:

[0101] S131. Based on the labeled training sample dataset, different types of classifiers are trained separately to obtain multiple classifiers whose training effect meets the preset requirements.

[0102] S132. Integrate the multiple classifiers based on a voting method to obtain an initial braking performance evaluation model; and

[0103] S133. Train the initial braking performance evaluation model based on the labeled training sample data to obtain the braking performance evaluation model.

[0104] Specifically, in step S131, based on the labeled training sample dataset, different classifiers are used for training (including: logistic regression classifier, support vector machine classifier, K-nearest neighbor classifier, Bagging classifier, random forest classifier, decision tree classifier, and extreme gradient boosting tree classifier). The three best-performing classifiers are selected based on metrics such as accuracy, recall, and algorithm efficiency. For example, higher accuracy and recall, and shorter running time, indicate a better algorithm.

[0105] In step S132, the three selected superior classifier models are integrated using a voting method to obtain an initial braking performance evaluation model. In step S133, based on the labeled training sample dataset, the integrated initial braking performance evaluation model is retrained to obtain the final braking performance evaluation model. The final braking performance evaluation model can assess whether the overall braking performance of the rail train is abnormal.

[0106] Figure 7 A schematic diagram of a rail vehicle braking performance evaluation system provided according to an exemplary embodiment of the present invention is shown.

[0107] like Figure 7 As shown, the rail vehicle braking performance evaluation system 200 includes a data acquisition module 210 and a braking performance evaluation module 220. The data acquisition module 210 collects multiple operating condition data characterizing the rail vehicle's operating conditions and multiple braking data characterizing the performance of the rail vehicle's braking system during braking. The braking performance evaluation module 220 pre-stores a braking performance evaluation model constructed according to the rail vehicle braking performance evaluation model construction method 100. After receiving the operating condition data and braking data collected by the data acquisition module, the braking performance evaluation module calls the braking performance evaluation model to process the received data to evaluate whether the rail vehicle's braking performance is abnormal.

[0108] In some implementations, the rail vehicle braking performance evaluation system 200 further includes a data management module 230, which stores the raw vehicle and braking-related data and braking performance evaluation results collected by the management system, and pushes the results data judged as abnormal to the vehicle network system.

[0109] The present invention also provides an electronic device, such as... Figure 8 As shown, the construction method 100 for running the above-mentioned rail vehicle braking performance evaluation model includes electronic devices such as the rail vehicle's vehicle-mounted system or a cloud connected to the vehicle-mounted system.

[0110] See Figure 8 The diagram shows the structure of an electronic device 300, which includes a processor 301 and a memory 302 communicatively connected to the processor 301. The memory 302 is used to store one or more computer programs, which are executed by the processor 301 to implement the method 100 described above.

[0111] Figure 8 The electronic device 300 shown also includes a bus 303 and a communication interface 304. The processor 301, the communication interface 304 and the memory 302 are connected through the bus 303.

[0112] The memory 302 may include high-speed random access memory (RAM), or it may also include non-volatile memory 302, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 304 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc. The bus 303 may be an ISA bus, PCI bus, or EISA bus, etc. The bus 303 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 8 The symbol is represented by only one double-headed arrow, but this does not mean that there is only one bus 303 or one type of bus 303.

[0113] Processor 301 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 301 or by instructions in software form. Processor 301 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor can be a microprocessor, or processor 301 can be any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a readily available storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 302. The processor 301 reads the information from memory 302 and, in conjunction with its hardware, completes the steps of the method described in the foregoing embodiments.

[0114] An exemplary embodiment of the present invention also provides a computer-readable storage medium storing a computer program. When the computer program is called and executed by the processor 301, the computer-executable instructions cause the processor 301 to implement the above-described method 100 for constructing the rail vehicle braking performance evaluation model. For specific implementation details, please refer to the method embodiment, which will not be repeated here.

[0115] The computer program product of the method 100 for constructing the braking performance evaluation model of rail vehicles and the electronic device 300 provided in the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods in the preceding method embodiments. For specific implementation, please refer to the method embodiments, which will not be repeated here.

[0116] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described device and / or electronic equipment can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0117] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-described technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for constructing a braking performance evaluation model for rail vehicles, characterized in that, include: A training sample dataset is collected for training the braking performance evaluation model. The training sample dataset includes multiple deceleration process sub-datasets, and each deceleration process sub-dataset includes multiple feature parameters of the rail vehicle during the braking process. Label multiple deceleration process subsets in the training sample dataset using the deceleration of each deceleration process subset, including: The multiple deceleration process subsets are divided into several level data groups according to the braking level of the rail vehicle, and the average speed in each deceleration process subset in each level data group is calculated. Based on the preset speed intervals and the calculated average speed, the multiple deceleration process subsets in each stage data group are divided into several working condition data groups, and the deceleration in each deceleration process subset within each working condition data group in each stage data group is calculated; and Statistical analysis is performed on the calculated decelerations, and each deceleration process subset is labeled based on the results of the statistical analysis, including: Based on the calculated deceleration, draw the deceleration box plot for different working condition data groups in each level data group; According to the preset deceleration scoring system, the average score of deceleration in different working condition data groups in each level data group is calculated based on the drawn deceleration box plot. A box plot of the average deceleration is drawn based on the average deceleration of different working condition data groups in each level data group; and The average binning plots are used to label multiple deceleration process subsets within the training sample dataset, with labels including normal braking and abnormal braking; and A braking performance evaluation model is trained based on a labeled training sample dataset. The trained braking performance evaluation model is used to evaluate whether the braking performance of rail vehicles is abnormal.

2. The method for constructing a rail vehicle braking performance evaluation model according to claim 1, characterized in that, The steps for training a braking performance evaluation model based on a labeled training sample dataset include: Based on the labeled training sample dataset, different types of classifiers are trained separately to obtain multiple classifiers whose training performance meets the preset requirements. The multiple classifiers are integrated using a voting method to obtain an initial braking performance evaluation model; and The initial braking performance evaluation model is trained based on the labeled training sample data to obtain the braking performance evaluation model.

3. The method for constructing a rail vehicle braking performance evaluation model according to claim 2, characterized in that, The different types of classifiers include logistic regression classifiers, support vector machine classifiers, K-nearest neighbor classifiers, Bagging classifiers, random forest classifiers, decision tree classifiers, and extreme gradient boosting tree classifiers.

4. The method for constructing a rail vehicle braking performance evaluation model according to any one of claims 1 to 3, characterized in that, The steps for collecting the training sample dataset for training the braking performance evaluation model include: Collect multiple operating condition data to characterize the working conditions of the rail vehicle and multiple braking data to characterize the performance of the rail vehicle's braking system during braking; and Correlation analysis is performed on the multiple operating condition data and the multiple braking data to obtain the multiple characteristic parameters.

5. The method for constructing a rail vehicle braking performance evaluation model according to claim 4, characterized in that, The steps of performing correlation analysis on the multiple operating condition data and the multiple braking data to obtain the multiple characteristic parameters include: A first correlation analysis was performed on the multiple operating condition data and the multiple braking data to obtain multiple parameters related to braking performance evaluation as feature parameters; and A second correlation analysis is performed on the feature parameters to reduce their dimensionality.

6. A system for evaluating the braking performance of rail vehicles, characterized in that, include: The data acquisition module is used to collect multiple operating condition data that characterize the working conditions of the rail vehicle and multiple braking data that characterize the performance of the rail vehicle's braking system during the braking process. as well as A braking performance evaluation module is provided, wherein a braking performance evaluation model constructed according to the method for constructing a rail vehicle braking performance evaluation model according to any one of claims 1 to 5 is pre-stored in the braking performance evaluation module. After receiving the operating condition data and braking data collected by the data acquisition module, the braking performance evaluation module calls the braking performance evaluation model to process the received data in order to evaluate whether the braking performance of the rail vehicle is abnormal.

7. An electronic device, characterized in that, include: processor; A memory storing a computer program configured to be executed by the processor, which, when executing the computer program, implements the method for constructing a rail vehicle braking performance evaluation model as described in any one of claims 1 to 5.

8. A readable storage medium, characterized in that, A computer program is stored on a readable storage medium, which, when executed by a processor, implements the method for constructing a rail vehicle braking performance evaluation model according to any one of claims 1 to 5.

Images