System and method for predicting residual life of wheel set of high-speed motor train unit

Abstract

The invention discloses a high-speed motor train unit wheel set residual life prediction system and method, the system comprises a database, a data access layer, a business logic layer and a visualization layer, the business logic layer is provided with a wheel set residual life prediction model, the database is communicated with the data access layer, and the visualization layer is provided with a wheel set residual life prediction model. The service logic layer is communicated with the data access layer, the service logic layer is communicated with the visualization layer, and the visualization layer is used for realizing setting of life prediction parameters and visual query of information. The business logic layer obtains data in the database through the data access layer and predicts the wheel set residual life through the wheel set residual life prediction model, and the database is used for storing the data; the motor train unit wheel set residual life prediction method can relatively accurately predict the motor train unit wheel set residual life, provides production guidance for motor train unit wheel set centralized maintenance plan arrangement, and saves motor train unit operation cost.

Description

Technical Field

[0001] The present invention relates to the technical field of EMU equipment maintenance, and in particular to a system and method for predicting the remaining life of a high-speed EMU wheel set. Background Art

[0002] Wheel sets are the main components of EMUs. While they have the functions of carrying, traction, braking and guiding, they are also important consumable parts in EMUs. With the development of domestic high-speed EMUs, the number of high-speed EMUs is increasing, and the number of EMU wheel sets is also increasing. As one of the three major wear parts, wheelsets need to be replaced in time when the size parameters of EMU wheelsets are worn or the time limit for turning and repair is about to be reached.

[0003] The number of EMU wheel sets overhauled and replaced is large, and the EMU wheel set maintenance cycle is long. In actual engineering applications, a replacement plan needs to be made in advance before the EMU wheelset reaches the limit. However, the existing technology cannot predict the remaining life of the EMU wheelset, so it is impossible to effectively prepare and match the wheelset in advance before the limit, which greatly increases the EMU operating cost.

[0004] In view of the above-mentioned defects, the inventors of the present invention finally obtained the present invention after a long period of research and practice. Summary of the invention

[0005] In order to solve the above-mentioned technical defects, the technical solution adopted by the present invention is to provide 1. A prediction system for the remaining life of high-speed EMU wheelsets, including a database, a data access layer, a business logic layer and a visualization layer, the business logic layer is provided with a wheelset remaining life prediction model, the database is connected to the data access layer, the business logic layer is connected to the data access layer, the business logic layer is connected to the visualization layer, the visualization layer is used to realize the setting of life prediction parameters and the visual query of information, the business logic layer obtains the data in the database through the data access layer and predicts the remaining life of the wheelset through the wheelset remaining life prediction model, and the database is used for storing data.

[0006] Preferably, the database includes a first database and a second database, the first database is used to store turning and repair detection data, life prediction parameters, marshaling files and life prediction data, and the second database is used to store mileage data.

[0007] Preferably, the visualization layer includes a life prediction parameter module, a marshaling archive module, a turning and repair record module and a life prediction module, the life prediction parameter module is used to set and query the life prediction parameters, the marshaling archive module is used to maintain and query the marshaling archives, the turning and repair record module is used to query the turning and repair detection data, and the life prediction module is used to query the life prediction results.

[0008] Preferably, a method for predicting the remaining life of a high-speed train wheelset, using the system for predicting the remaining life of a high-speed train wheelset, comprises the steps of:

[0009] S1, establishing a wheelset remaining life prediction model;

[0010] S2, by collecting real-time data in real time, applying the trained mathematical model of the remaining life of the wheelset to the real-time data to obtain a life prediction result.

[0011] Preferably, in step S1, the wheelset wear law model and wheelset turning and repair law model of each vehicle model are constructed by collecting original data and processing data, and characteristic values ​​are extracted based on data correlation feature construction and combined with the wheelset wear law model and the wheelset turning and repair law model for training to build the mathematical model of the remaining life of the wheelset.

[0012] Preferably, the original data includes basic information of the train formation, operating route information, mileage, mileage between advanced repairs, historical turning and repair data of the turning machine, wear data of wheelsets of different vehicle models and application standards.

[0013] Preferably, the wheelset wear law model and the wheelset turning law model are constructed by calculating the periodic wear law of wheelset parameters of different models on different routes, the periodic turning law of wheelset parameters of different models, the weighted average daily mileage of the vehicle group, and the relationship between wheel rim compensation and wheel diameter turning amount.

[0014] Preferably, the characteristic values ​​include car type, wheel diameter, wheel rim parameters, wear and turning parameters, advanced repair interval mileage, average daily deformation kilometers, wheelset turning cycle, minimum operating limits of wheelsets of different models and minimum advanced repair factory parameters.

[0015] Preferably, in step S2, real-time data is collected in real time, and the trained mathematical model of the remaining life of the wheelset is applied to the real-time data to achieve wheel health status assessment and remaining life prediction.

[0016] Preferably, the real-time data includes the mileage of high-speed EMUs, the average daily mileage of EMUs, wheelset size data, and wheelset wear and turning data of EMUs on different operating routes.

[0017] Preferably, the data processing process of the wheelset remaining life prediction model is as follows:

[0018]

[0019]

[0020]

[0021] G4=MOD(G1,(t4+t5));

[0022]

[0023]

[0024]

[0025] G8=G5―G6;

[0026]

[0027]

[0028]

[0029] Z2 = MIN (Z1 in the same compartment as 1-8);

[0030]

[0031] Z4 = MIN (Z3 in the same compartment 1-8);

[0032] Z5=IF(OR(Z4―t1<0,Z2<G5),“yes”,“no”);

[0033] Z6 = IF (AND (Z3-t8) < 0, Z5 = "No"), "Yes", "No");

[0034] Among them, t1 is the lower limit of wheel diameter; t2 is the lower limit of wheel rim thickness; t3 is the upper limit of boring and repair cycle; t4 is the wear amount during boring and repair cycle; t5 is the boring and repair amount during boring and repair cycle; t6 is the average daily mileage; t7 is the interval period of advanced repair; t8 is the lower limit of wheel diameter for advanced repair; J1 is the mileage after the last boring and repair; J2 is the boring and repair label; J3 is the mileage after the last advanced repair; G is the current wheel diameter value; G1 is the remaining wheel diameter value after the next boring and repair; G2 is the time until the next boring and repair; G3 is the integer of the remaining wheel diameter value; G4 is the remainder of the remaining wheel diameter value; G5 is the advanced repair date; G6 is the next boring and repair date; G7 is the wheel diameter reduction value after the next boring and repair; G8 is the time until the advanced repair; G9 is the integer of the remaining date; G10 is the remainder of the remaining date; Z1 is the wheel diameter limit date; Z2 is the wheel diameter limit date; Z3 is the wheel diameter value before entering the factory for advanced repair; Z4 is the wheel diameter value before entering the factory for advanced repair; Z5 is whether to replace during use; Z6 is whether to replace during advanced repair.

[0035] Preferably, the process of determining the mileage J1 after the last turning and repair is as follows:

[0036] Create axle labels and vehicle group labels;

[0037] The process of establishing the axle label is as follows:

[0038] A database is created for each axle of the train set, and each axle is labeled with a time stamp indicating whether it is full-line turning or temporary turning;

[0039] The process of judging full-line turning and temporary turning is as follows: for each axle, when the difference between the wheel diameter before turning and the wheel diameter after turning is greater than 0.5mm, it is judged as effective turning. After it is determined to be effective turning, the category in the "turning category" of the vehicle group where the axle belongs is used to judge whether it is temporary turning or full-line turning. All except those marked as full-line turning are temporary turning.

[0040] When querying the database of the same axle, the data of "km traveled since last turning and repair" is divided into the following four situations:

[0041] First, the whole train is turned and then the temporary turning is performed. The running distance of the train set corresponding to the temporary turning date is subtracted from the running distance of the train set on the corresponding date of the whole train turning, which is the "running kilometers after the last turning";

[0042] The second is to first turn the whole train and then turn the whole train, and directly get the mileage as 0;

[0043] The third method is to first do the temporary turning and then the full turning, and the mileage is directly obtained as 0;

[0044] Fourth, whether the current temporary turning is still temporary turning, calculate the mileage between the last temporary turning and the last full turning;

[0045] The process of establishing the vehicle group label is as follows:

[0046] A database is created for the train set itself; the train set is labeled as full train lapping or "temporary lapping", and the latest label with time is stored. When the train set lapping category is full train lapping, the "mileage after last lapping" data with "axle temporary lapping label" is refreshed, and the "mileage after last lapping" is directly recorded as 0.

[0047] Preferably, the repair label J2 with date, the unit is date, is used to calculate the interval date and mileage.

[0048] Preferably, the running time J3 after the last advanced repair is the running time of the current train set minus the running time of the train set at the advanced stage.

[0049] Compared with the prior art, the beneficial effect of the present invention is that the present invention can relatively accurately predict the remaining life of the EMU wheelset, provide production guidance for the centralized maintenance plan arrangement of the EMU wheelset, and save the EMU operating cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a structural schematic diagram of the system for predicting the remaining life of a high-speed train wheel set;

[0051] Figure 2 A schematic flow chart of the method for predicting the remaining life of a high-speed train set wheelset;

[0052] Figure 3 This is a schematic diagram of data processing in step S2. DETAILED DESCRIPTION

[0053] The above and other technical features and advantages of the present invention are described in more detail below in conjunction with the accompanying drawings.

[0054] Embodiment 1

[0055] like Figure 1 As shown, Figure 1 The figure is a schematic diagram of the structure of the system for predicting the remaining life of high-speed EMU wheelsets.

[0056] The prediction system for the remaining life of high-speed EMU wheelsets described in the present invention comprises a database, a data access layer, a business logic layer and a visualization layer, the business logic layer is provided with a wheelset remaining life prediction model, the database is connected to the data access layer, the business logic layer is connected to the data access layer, the business logic layer is connected to the visualization layer, the visualization layer is used to realize the setting of life prediction parameters and the visualization query of information, the business logic layer obtains the data in the database through the data access layer and predicts the remaining life of the wheelset through the wheelset remaining life prediction model, and the database is used to store data.

[0057] The database includes a first database and a second database, the first database is used to store turning and repair detection data, life prediction parameters, marshaling files and life prediction data, and the second database is used to store mileage data.

[0058] The visualization layer includes a life prediction parameter module, a marshaling archive module, a turning and repair record module and a life prediction module. The life prediction parameter module is used to set and query the life prediction parameters, the marshaling archive module is used to maintain and query the marshaling archives, the turning and repair record module is used to query the turning and repair detection data, and the life prediction module is used to query the life prediction results.

[0059] like Figure 2 As shown, Figure 2 The figure is a flow chart of the method for predicting the remaining life of a high-speed train wheel set. The method for predicting the remaining life of a high-speed train wheel set of the present invention comprises the following steps:

[0060] S1, establishing a wheelset remaining life prediction model;

[0061] S2, by collecting real-time data in real time, applying the trained mathematical model of the remaining life of the wheelset to the real-time data to obtain a life prediction result.

[0062] Preferably, in step S1, the wheelset wear law model and wheelset turning and repair law model of each vehicle model are constructed by collecting original data and processing data, and characteristic values ​​are extracted based on data correlation feature construction and combined with the wheelset wear law model and the wheelset turning and repair law model for training to build the mathematical model of the remaining life of the wheelset.

[0063] The original data includes basic information of the train formation, operating route information, mileage, mileage between advanced repairs, historical turning and repair data of the turning machine, wear data of wheelsets of different models and application standards.

[0064] The wheelset wear law model and the wheelset turning law model are constructed by using the periodic wear law of wheelset parameters of different models, the periodic turning law of wheelset parameters of different models, the weighted average calculation of the average daily mileage of the vehicle group, and the relationship between wheel rim compensation and wheel diameter turning amount.

[0065] The characteristic values ​​include carriage type, wheel diameter, wheel rim parameters, wear and turning parameters, mileage between advanced repairs, average daily deformation kilometers, wheelset turning cycle, minimum operating limits of wheelsets of different models and minimum parameters for advanced repairs before leaving the factory.

[0066] like Figure 3 As shown, Figure 3 The data processing diagram of step S2 is shown in FIG. 1 . Preferably, in step S2, by collecting real-time data in real time, the trained mathematical model of the remaining life of the wheelset is applied to the real-time data to achieve wheel health status assessment and remaining life prediction. The advantage of the mathematical model of the remaining life of the wheelset is that it takes into account the actual periodic wear of the wheelset and the operating rules of the turning and repairing group, avoids interference caused by simulation calculations, and improves the accuracy of the prediction results.

[0067] Preferably, the real-time data includes the mileage of high-speed EMUs, the average daily mileage of EMUs, wheelset size data, and wheelset wear and turning data of EMUs on different operating routes.

[0068] exist Figure 3 The inspections include LY inspection (dynamic inspection of wheelsets), turning and repair inspection, LU inspection (rim and spoke flaw detection inspection), and LHZ inspection (non-destructive inspection). The wheelset size data and wheelset wear and turning and repair data of EMUs with different operating routes can be obtained through the inspections.

[0069] The present invention can relatively accurately predict the remaining service life of EMU wheelsets, provide production guidance for the planning of centralized maintenance of EMU wheelsets, and save EMU operating costs.

[0070] Embodiment 2

[0071] In this embodiment, the CRH3 and CR400AF EMU wheelsets are taken as the research objects, and the measured data of wheel turning machine is used as the driving force. The wear law model and the turning and repair law model are established based on mathematical statistics; the wheelset remaining life prediction model is built according to the characteristic parameters such as the turning and repair cycle of each model of EMU, the average daily mileage, the mileage between advanced repairs of the train set, the wheelset cycle wear and turning amount, and the EMU wheelset technical parameters. The trained model is integrated with the wheel turning and flaw detection module in real time, and the prediction system of the present invention is constructed through data interaction through the Web Service interface, with Java and Python as the back-end development languages, and using MySQL and mongoDB databases.

[0072] Specifically, driven by the wheel boring machine measurement data before and after wheelset boring and repair in the wheel boring and flaw detection module, the wheel diameter value and rim thickness historical parameters before and after boring and repair in the boring and repair cycle are deeply mined, and the wear data of wheel diameter value and rim thickness and boring and repair data are cleaned and screened based on mathematical statistics to establish the wheel set size parameter wear and boring and repair models for CR400AF and CRH3 EMUs respectively.

[0073] According to the analysis of the periodic wear and boring rules of CRH3 and CR400AF EMU wheelsets, it can be seen that the average periodic wear of the CRH3 EMU is about 1.84mm, the average boring and boring of the middle car is about 2.23mm, and the average boring and boring of the lead car is about 6.92mm; the average periodic wear of the CR400AF EMU is about 1.48mm. Because the wheel flange wear of the CR400AF EMU wheelset cannot be ignored during operation, the wheel flange thickness parameter has a minimum threshold of 26mm in use. Increasing the wheel flange thickness parameter during periodic boring and boring requires more wheel diameter.

[0074] Through the analysis of historical data of periodic wear of wheel flange thickness, the periodic wear of wheel flange thickness is 0.42mm; while the wheel flange thickness increases by about 0.53mm, consuming about 6.5mm of wheel diameter; the wheel flange thickness increases by about 0.96mm, consuming about 7.8mm of wheel diameter; according to a certain attachment, the maximum periodic wear of wheel flange thickness is close to 1mm. In order to prevent the wheel flange thickness from reaching the limit during use and causing emergency turning, the minimum wheel flange thickness threshold is set to 27mm after periodic turning. When the wheel flange thickness is less than 27mm before periodic turning, it is necessary to increase the wheel flange thickness parameter by turning more wheel diameter values; the prototype wheel flange thickness of CR400AF EMU wheelset is 32.9mm, and not all wheelset wheel flange thickness values ​​are increased to above the periodic wear of 0.42mm during periodic turning. Therefore, the average value of wheel diameter periodic turning amount of 4.29mm is selected when predicting the life of CR400AF EMU wheelset. According to the line and operation of a certain assigned station, the mileage of CRH3 and CR400AF assigned trains throughout the year was counted, and the weighted average was used to obtain an average daily mileage of 1,700 kilometers; according to the "Notice of the Ministry of Mechanical and Vehicle Engineering of the National Railway Group on Issuing the Summary and Review Opinions on the Extension of the Verification of the Advanced Maintenance and Overhaul Cycle of the Harmony EMU", the mileage between advanced maintenance intervals of EMUs is required to be 1.45 million kilometers. Due to the extension of the verification, some trains may have an advanced maintenance interval of 1.65 million kilometers. The advanced maintenance mileage intervals of different trains are determined according to actual conditions. According to the mileage between advanced maintenance intervals of trains of different models, the average daily mileage of trains of different assigned stations, the actual turning and overhaul cycle of trains, the wear and overhaul amount of different models, the wheelset remaining life prediction model is established and the wheel life is judged. The specific process is as follows:

[0075] The vehicle model variables are shown in the following table:

[0076]

[0077] The basic data are the mileage after the last lathing repair J1, the lathing repair label J2 and the mileage after the last advanced repair J3. The mileage after the last lathing repair J1, the lathing repair label J2, the mileage after the last advanced repair J3 and the current wheel diameter value G are obtained by processing the real-time data and the database.

[0078] The process of determining the mileage after the last turning and repair is as follows:

[0079] Create axle labels and vehicle group labels;

[0080] The specific process of the axle label is as follows:

[0081] A database is created for each axle of the train set. Each axle is labeled with a time stamp indicating whether it is full-line turning or temporary turning.

[0082] The process of judging full-row turning and temporary turning is as follows: for each axle, when the difference between the wheel diameter before turning and the wheel diameter after turning is greater than 0.5mm, it is judged as a valid turning repair. After being determined as a valid turning repair, the category in the "Turning Category" of the vehicle group where the axle is located is used to judge whether it is a temporary turning or a full-row turning label (all those except those marked as full-row turning are temporary turnings).

[0083] When querying the database of the same axle, the data of "km traveled since last turning and repair" is divided into the following four situations:

[0084] First, the whole train is turned and then the temporary turning is performed. At this time, the running distance of the train set corresponding to the temporary turning date is calculated minus the running distance of the train set on the corresponding date of the full train turning, which is the "running kilometers after the last turning";

[0085] The second is to first turn the whole train and then turn the whole train, and directly get the mileage as 0;

[0086] The third method is to first do the temporary turning and then the full turning, and the mileage is directly obtained as 0;

[0087] Fourth, whether it is before the current temporary turning or the temporary turning, the mileage between the last temporary turning and the previous full turning should be calculated.

[0088] The specific process of the vehicle group label is as follows:

[0089] A database is created for the train set itself. The train set is labeled as full train lapping or "temporary lapping", and the latest label with time is stored. When the lapping category of the train set is full train lapping, the "mileage after last lapping" data with the "axle temporary lapping label" is refreshed, and the "mileage after last lapping" is directly recorded as 0.

[0090] The repair label with the date is in days, which is used to calculate the interval date and mileage.

[0091] The running time after the last advanced repair is the running time of the current train set (during temporary turning or full train turning) minus the running time of the train set during the advanced repair.

[0092] The data processing process is:

[0093] Current wheel diameter remaining / mm = current wheel diameter value / mm - minimum operating limit wheel diameter value / mm (t1)

[0094] Current rim thickness remaining / mm = Current rim thickness / mm - Minimum operating limit rim thickness / mm (t2)

[0095] Mileage after last boring and repair / km = Mileage when wheel diameter changes / km – Mileage when whole train is boring and repair / km

[0096]

[0097]

[0098] The remainder of the remaining wheel diameter = MOD (the remaining wheel diameter after the next turning / mm, (t4+t5))

[0099]

[0100] Wheel diameter expiration date Each carriage wheel diameter takes the minimum value / mm = MIN (wheel diameter expiration date for digits 1-8 of the same carriage)

[0101]

[0102]

[0103]

[0104] Days to advanced maintenance / days = advanced maintenance date / year.month.day - next turning maintenance date / year.month.day

[0105]

[0106]

[0107]

[0108] The minimum wheel diameter of each carriage before entering the advanced repair factory / mm = MIN (the wheel diameter of the same carriage 1-8 before entering the advanced repair factory / mm)

[0109] Whether to change the application

[0110] =IF(OR(minimum value of each car before advanced repair / mm-minimum operating limit wheel diameter / mm(t1), minimum value of each car on wheel diameter expiration date / year.month.day<advanced repair date / year.month.day), "yes", "no")

[0111] Advanced repair whether to replace

[0112] =IF(AND(front wheel diameter value of advanced repair factory / mm-868mm(t8)<0, whether to replace during use=“No”), “Yes”, “No”).

[0114] Table 1 Technical standards for wheel pair data after reconditioning of CRH3 and CR400AF EMUs

[0115]

[0116]

[0117] Table 2 Variable values ​​in the mathematical model for predicting the remaining life of wheelsets of different vehicle models

[0118]

[0119]

[0120] Table 3 Comparison between the predicted time of the remaining life of the wheelset by the mathematical model and the actual replacement time in operation

[0121]

[0122]

[0123] The remaining life of the wheelset is compared with the actual replacement time of the wheelset in use by using the wheelset remaining life prediction model, as shown in Table 3. As can be seen from Table 3, the predicted wheelset life limit time by using the wheelset remaining life prediction model is roughly similar to the actual replacement time of the wheelset in use, which can relatively accurately predict the remaining life of the EMU wheelset, provide production guidance for the centralized maintenance plan arrangement of the EMU wheelset, and save the EMU operating cost; although not completely consistent, the main reasons are:

[0124] The wear of wheel diameter and wheel rim thickness parameters of each wheelset is different, which is related to complex factors such as dynamic performance, road conditions, wheel size parameters after turning, weather, etc. The wear data in the mathematical model is the average value taken based on a large amount of historical wear data. The prediction result of the remaining life of the wheelset is related to the turning strategy. In actual turning, it is necessary to consider factors such as the wheel diameter difference constraint of the entire wheel, polygon data, equivalent taper data, restoration of the minimum wheel rim thickness threshold, and turning machine error value. The wheel diameter turning amount in the prediction model is based on the average value of historical data, so there will be a small range of error between the calculation result and the actual wheelset limit time. The actual replacement time of the wheelset will be earlier or closer to the limit time predicted in the model. The reason is to ensure the minimum wheelset data application limit requirements, combined with the actual use of the vehicle group, centralized wheel replacement through manual intervention.

[0125] In this embodiment, the wheelset remaining life prediction model is completed in Python. The back-end Java program assembles various data and then calls the model algorithm to generate prediction results, which are displayed on the front-end JSP page. In the development of the wheelset remaining life prediction model, the data access layer uses two databases, MySQL and mongoDB. Basic archive information and test data are stored in the MySQL database, and mileage information is stored in the mongoDB database through the EMIS system; the development languages ​​used in the business logic layer are mainly Java and Python. Java encapsulates the data required for life prediction, mainly life prediction parameters, archive information, turning and repair test data, etc. After the data is encapsulated, the Python program is called for calculation, and the calculation results are stored in the MySQL database; the languages ​​used in the visualization layer are js, jQuery, HTML5, etc., which are used to query and display life prediction data on the page.

[0126] By comparing the remaining life prediction results of the wheelset in the system of the present invention, the prediction results in the mathematical model and the actual replacement time of the wheelset, it can be seen from the comparison results that the remaining life prediction results of the wheelset are highly consistent with the actual time limit on site, as shown in Table 4. The remaining life prediction of the wheelset in the present invention has the function of guiding the arrangement of wheel replacement plans on site.

[0127] Table 4 Comparison between the time when the wheelset reaches the limit predicted by the wheelset remaining life module developed in the system and the actual replacement time in operation

[0128]

[0129] The above description is only a preferred embodiment of the present invention, which is only illustrative and not restrictive of the present invention. Those skilled in the art understand that many changes, modifications, and even equivalences may be made to the present invention within the spirit and scope defined by the claims of the present invention, but all of them will fall within the scope of protection of the present invention.

Claims

1. A system for predicting the remaining life of high-speed train wheelsets, characterized in that: It includes a database, a data access layer, a business logic layer and a visualization layer. The business logic layer is provided with a wheelset remaining life prediction model. The database is connected to the data access layer, the business logic layer is connected to the data access layer, the business logic layer is connected to the visualization layer, the visualization layer is used to realize the setting of life prediction parameters and the visualization query of information, the business logic layer obtains the data in the database through the data access layer and predicts the wheelset remaining life through the wheelset remaining life prediction model, and the database is used for data storage.

2. The system for predicting the remaining life of high-speed train wheelsets according to claim 1, characterized in that: The database includes a first database and a second database, the first database is used to store turning and repair detection data, life prediction parameters, marshaling files and life prediction data, and the second database is used to store mileage data.

3. The system for predicting the remaining life of high-speed train wheelsets as claimed in claim 2, characterized in that: The visualization layer includes a life prediction parameter module, a marshaling archive module, a turning and repair record module and a life prediction module. The life prediction parameter module is used to set and query the life prediction parameters, the marshaling archive module is used to maintain and query the marshaling archives, the turning and repair record module is used to query the turning and repair detection data, and the life prediction module is used to query the life prediction results.

4. A method for predicting the remaining life of a high-speed train wheel set, characterized in that: The system for predicting the remaining life of a high-speed train wheelset as claimed in claim 3 comprises the following steps: S1, establishing a wheelset remaining life prediction model; S2, by collecting real-time data in real time, applying the trained mathematical model of the remaining life of the wheelset to the real-time data to obtain a life prediction result.

5. The method for predicting the remaining life of a high-speed train wheelset according to claim 4, characterized in that: In the step S1, the wheelset wear law model and the wheelset turning and repair law model of each vehicle model are constructed by collecting original data and processing data, and the characteristic values ​​are extracted based on the data correlation feature structure and combined with the wheelset wear law model and the wheelset turning and repair law model for training to build the mathematical model of the remaining life of the wheelset.

6. The method for predicting the remaining life of a high-speed train wheelset according to claim 5, characterized in that: The original data includes basic information of the train formation, operating route information, mileage, mileage between advanced repairs, historical turning and repair data of the turning machine, wear data of wheelsets of different models and application standards.

7. The method for predicting the remaining life of a high-speed train wheelset according to claim 5, characterized in that: The wheelset wear law model and the wheelset turning law model are constructed by calculating the periodic wear law of wheelset parameters of different models on different routes, the periodic turning law of wheelset parameters of different models, the weighted average value to calculate the average daily mileage of the vehicle group, and the relationship between wheel rim compensation and wheel diameter turning amount.

8. The method for predicting the remaining life of a high-speed train wheelset according to claim 5, characterized in that: The characteristic values ​​include carriage type, wheel diameter, wheel rim parameters, wear and turning parameters, mileage between advanced repairs, average daily deformation kilometers, wheelset turning cycle, minimum operating limits of wheelsets of different models and minimum parameters for advanced repairs before leaving the factory.

9. The method for predicting the remaining life of a high-speed train wheelset according to claim 4, characterized in that: In the step S2, real-time data is collected in real time, and the trained mathematical model of the remaining life of the wheelset is applied to the real-time data to achieve wheel health status assessment and remaining life prediction.

10. The method for predicting the remaining life of a high-speed train wheelset according to claim 9, characterized in that: The real-time data includes the mileage of high-speed EMUs, the average daily mileage of EMUs, wheelset size data, and wheelset wear and turning data of EMUs on different operating routes.

Images