Prediction method, system, equipment and medium for brake main air path pressure leakage

Abstract

The present invention discloses a prediction method, system, equipment and medium for the leakage of the braking main air path pressure. The prediction method comprises: obtaining the running status data of the target subway train in the target time period; filtering out the parking status data corresponding to the longest parking time period of the target subway train from the running status data according to the parking confirmation parameters; calculating and obtaining the effective data of the main air path pressure in the longest parking time period based on the parking status data; inputting the effective data of the main air path pressure into the leakage model, and outputting the leakage status of the braking main air path pressure of the target subway train in the target time period. The present invention calculates and obtains the effective data of the main air path pressure under the condition of limited sensor accuracy; improves the reliability and effectiveness of the pressure leakage prediction data, and enhances the efficiency and accuracy of the main air path pressure prediction of the subway train; timely guides the maintenance work of the main air path of the subway train, and reduces the cost of repair and maintenance.

Description

Technical Field

[0001] The present invention relates to the technical field of train brake main air path pressure detection technology, and in particular to a method, system, equipment and medium for predicting brake main air path pressure leakage. Background Art

[0002] In winter, due to thermal expansion and contraction, the sealing of pipe joints is poor, resulting in more pipeline leakage faults. In the traditional vehicle maintenance process, the detection method for airtightness leak points mainly relies on listening, manual testing, soap water testing, etc. However, this detection method is inefficient and small leaks are difficult to detect.

[0003] With the continuous deepening of intelligent subway operation and maintenance, and the rapid development of technologies such as sensors, embedded systems, network communications, and big data, subway trains generally use on-board data transmission devices to send faults and key driving parameters that occur during vehicle operation to ground data centers, thereby monitoring the status of the train in real time. However, due to limitations such as communication capacity and high sensor costs, the types of gas path pressure signals transmitted by vehicles are relatively small, the accuracy of pressure data is not high enough, and the accuracy of leakage prediction is low; additional sensors need to be added, which is costly and difficult to promote. Summary of the invention

[0004] The technical problem to be solved by the present invention is to overcome the defects of low efficiency and poor accuracy in the prediction of brake main air path pressure in the prior art, and to provide a method, system, equipment and medium for predicting brake main air path pressure leakage.

[0005] The present invention solves the above technical problems through the following technical solutions:

[0006] In a first aspect, the present invention provides a method for predicting brake main air path pressure leakage, the prediction method comprising:

[0007] Acquiring the running status data of the target subway train in the target time period; the running status data includes parking confirmation parameters;

[0008] According to the parking confirmation parameter, the parking status data corresponding to the longest parking time period of the target subway train is selected from the running status data;

[0009] Calculate and obtain effective data of the main air path pressure during the longest parking period based on the parking state data;

[0010] The effective data of the main air path pressure is input into a leakage model, and the leakage state of the braking main air path pressure of the target subway train in the target time period is output; the leakage model is a mathematical model constructed by a two-dimensional threshold array and a two-dimensional attribute array, each threshold in the two-dimensional threshold array represents the holding time threshold of the main air path pressure at a set temperature and a set pressure, the attribute value of the two-dimensional attribute array includes true value and false value, and the dimension of the two-dimensional threshold array includes temperature and pressure.

[0011] Preferably, the parking confirmation parameters include train speed, driver's cab activation coefficient and air compressor working coefficient;

[0012] The step of selecting the parking status data corresponding to the longest parking time period of the target subway train from the running status data according to the parking confirmation parameter comprises:

[0013] According to the train speed, the driver's cab activation coefficient and the air compressor working coefficient, intercepting at least one alternative parking time period from the target time period;

[0014] or,

[0015] The operating status data also includes time parameters and reference parameters, the reference parameters include the brake main air path pressure; the parking confirmation parameters include train speed, driver's cab activation coefficient and air compressor working coefficient;

[0016] According to the train speed, the driver's cab activation coefficient and the air compressor working coefficient, intercepting at least one preliminary parking time period from the target time period;

[0017] Comparing every two adjacent air pressure change values ​​in the brake main air path pressure with a first threshold value, and determining at least one alternative parking time period from the primary parking time period according to the comparison result;

[0018] Selecting the longest time period from the candidate parking time periods as the longest parking time period;

[0019] The parking state data corresponding to the longest parking time period is intercepted from the running state data.

[0020] Preferably, the step of calculating and obtaining effective data of the main air path pressure during the longest parking period based on the parking state data comprises:

[0021] Extracting a plurality of different air pressure values ​​from the parking state data, and deleting the maximum and minimum air pressure values ​​to generate an initial air pressure array;

[0022] Eliminate the air pressure values ​​corresponding to the sampling time intervals greater than the second threshold from the initial air pressure array, and generate a pressure holding set and timeout judgment information;

[0023] generating a candidate pressure maintenance sample set and pressure maintenance preset judgment information according to the pressure maintenance set and the parking state data;

[0024] Comparing the number of samples in the candidate pressure maintenance sample set with a preset number;

[0025] If the number of samples is not greater than the preset number, the candidate pressure maintenance sample set is directly used as the effective pressure maintenance sample set, and the candidate pressure maintenance sample judgment information and the effective data of the main air path pressure are generated;

[0026] If the number of samples is greater than the preset number, the samples in the candidate pressure maintenance set are sorted in descending order of pressure maintenance duration, and a valid pressure maintenance sample set that meets the preset number is obtained by screening, and the candidate pressure maintenance sample judgment information and the main air path pressure valid data are generated;

[0027] The effective pressure holding sample set includes at least one of a pressure holding value, a holding temperature value, a pressure holding time and a starting time point.

[0028] Preferably, the step of generating a candidate pressure maintenance sample set according to the pressure maintenance set and the parking state data comprises:

[0029] selecting at least one target pressure maintenance value from the set of pressure maintenance values;

[0030] After comparing the difference between the air pressure value in the parking state data and the target pressure holding value with a third threshold, a plurality of candidate target pressure holding periods are intercepted from the longest parking period according to the comparison result;

[0031] Selecting a longest target pressure holding period from the candidate target pressure periods, and determining whether the longest target pressure holding period meets a preset condition;

[0032] If satisfied, generating the candidate pressure maintenance sample set according to the target pressure maintenance state data corresponding to the longest target pressure maintenance period intercepted from the parking state data;

[0033] If not, the step of selecting at least one target pressure holding value from the pressure holding value set is performed again until all pressure holding values ​​in the pressure holding value set are traversed.

[0034] Preferably, the prediction method obtains the leakage model by training through the following steps, including:

[0035] Collecting pressure maintenance data of multiple normal main air path pressures of subway trains in a historical period;

[0036] The pressure maintenance data is input into the initial model for training to obtain the leakage model.

[0037] Preferably, the pressure maintenance data includes standard parameters, and the step of inputting the pressure maintenance data into an initial model for training to obtain the leakage model includes:

[0038] Select the pressure maintenance data whose standard parameters are within the set range as the effective pressure maintenance data;

[0039] Based on the pressure holding time in the effective pressure holding data, the pressure holding time threshold of the two-dimensional threshold array in the initial model is corrected to generate the leakage model.

[0040] Preferably, after obtaining the running status data of the target subway train in the target time period, the prediction method further comprises:

[0041] Preprocessing the running status data; the preprocessing includes performing integrity detection and abnormal data correction on the running status data, and generating preprocessing judgment information;

[0042] Among them, the main air path pressure valid data includes the effective pressure maintenance sample set, the pre-processing judgment information, the timeout judgment information, the pressure maintenance preset judgment information and the alternative pressure maintenance sample judgment information.

[0043] In a second aspect, the present invention provides a prediction system for brake main air path pressure leakage, the prediction system comprising:

[0044] An acquisition module, used to acquire the running status data of the target subway train in the target period; the running status data includes parking confirmation parameters;

[0045] A screening module, configured to screen out parking status data corresponding to the longest parking time period of the target subway train from the running status data according to the parking confirmation parameter;

[0046] A calculation module, configured to calculate and obtain effective data of the main air path pressure during the longest parking period based on the parking state data;

[0047] A prediction module is used to input the effective data of the main air path pressure into a leakage model, and output the leakage state of the braking main air path pressure of the target subway train in a target period; the leakage model is a mathematical model constructed by a two-dimensional threshold array and a two-dimensional attribute array, each threshold in the two-dimensional threshold array represents the holding time threshold of the main air path pressure at a set temperature and a set pressure, the attribute value of the two-dimensional attribute array includes true value and false value, and the dimension of the two-dimensional threshold array includes temperature and pressure.

[0048] Preferably, the parking confirmation parameters include train speed, cab activation coefficient and air compressor working coefficient; the screening module includes:

[0049] A first interception unit is used to intercept at least one alternative parking time period from the target time period according to the train speed, the driver's cab activation coefficient and the air compressor working coefficient;

[0050] or,

[0051] The operating status data also includes time parameters and reference parameters, the reference parameters include the brake main air path pressure, and the parking confirmation parameters include train speed, driver's cab activation coefficient and air compressor working coefficient; the screening module includes:

[0052] A second interception unit is used to intercept at least one preliminary parking time period from the target time period according to the train speed, the driver's cab activation coefficient and the air compressor working coefficient;

[0053] a first comparison unit, configured to compare every two adjacent pressure change values ​​of the brake main air path pressure with a first threshold value, and determine at least one candidate parking time period from the primary parking time period according to the comparison result;

[0054] A first screening unit, configured to screen out a longest period from the candidate parking time periods as the longest parking time period;

[0055] The third intercepting unit is used to intercept the parking state data corresponding to the longest parking time period from the running state data.

[0056] Preferably, the calculation module comprises:

[0057] An extraction unit, used to extract a plurality of different air pressure values ​​from the parking state data, and delete the maximum and minimum air pressure values ​​to generate an initial air pressure array;

[0058] A removal unit, used to remove the air pressure values ​​corresponding to the sampling time intervals greater than the second threshold from the initial air pressure array, and generate a pressure holding set and timeout judgment information;

[0059] A generating unit, configured to generate a candidate pressure maintenance sample set and pressure maintenance preset judgment information according to the pressure maintenance set and the parking state data;

[0060] A second comparison unit, used for comparing the number of samples in the candidate pressure maintenance sample set with a preset number;

[0061] A first processing unit, configured to directly use the candidate pressure maintenance sample set as a valid pressure maintenance sample set if the number of samples is not greater than the preset number, and generate candidate pressure maintenance sample judgment information and the main air path pressure valid data;

[0062] A second processing unit is configured to, if the number of samples is greater than the preset number, sort the samples in the candidate pressure maintenance set in descending order of pressure maintenance duration, screen and obtain a valid pressure maintenance sample set that meets the preset number, and generate the candidate pressure maintenance sample judgment information and the main air path pressure valid data;

[0063] The effective pressure holding sample set includes at least one of a pressure holding value, a holding temperature value, a pressure holding time and a starting time point.

[0064] Preferably, the generating unit is specifically used for:

[0065] selecting at least one target pressure maintenance value from the set of pressure maintenance values;

[0066] After comparing the difference between the air pressure value in the parking state data and the target pressure holding value with a third threshold, a plurality of candidate target pressure holding periods are intercepted from the longest parking period according to the comparison result;

[0067] Selecting a longest target pressure holding period from the candidate target pressure periods, and determining whether the longest target pressure holding period meets a preset condition;

[0068] If satisfied, generating the candidate pressure maintenance sample set according to the target pressure maintenance state data corresponding to the longest target pressure maintenance period intercepted from the parking state data;

[0069] If not, the step of selecting at least one target pressure holding value from the pressure holding value set is performed again until all pressure holding values ​​in the pressure holding value set are traversed.

[0070] Preferably, the prediction system obtains the leakage model by training the following modules, including:

[0071] A collection module, used for collecting pressure maintenance data of a plurality of normal main air path pressures of a subway train in a historical period;

[0072] The training module is used to input the pressure maintenance data into the initial model for training to obtain the leakage model.

[0073] Preferably, the pressure maintenance data includes standard parameters, and the training module includes:

[0074] A selection unit, used for selecting the pressure maintenance data whose standard parameters are within a set range as effective pressure maintenance data;

[0075] A correction unit is used to correct the pressure holding time threshold of the two-dimensional threshold array in the initial model based on the pressure holding time in the effective pressure holding data to generate the leakage model.

[0076] The prediction system also includes:

[0077] A preprocessing module, used for preprocessing the running status data; the preprocessing includes performing integrity detection and abnormal data correction on the running status data, and generating preprocessing judgment information;

[0078] Among them, the main air path pressure valid data includes the effective pressure maintenance sample set, the pre-processing judgment information, the timeout judgment information, the pressure maintenance preset judgment information and the alternative pressure maintenance sample judgment information.

[0079] In a third aspect, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method for predicting brake main air path pressure leakage as described in the first aspect when executing the computer program.

[0080] In a fourth aspect, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for predicting the brake main air path pressure leakage described in the first aspect is implemented.

[0081] The positive and progressive effects of the present invention are: providing a prediction method, system, equipment and medium for the pressure leakage of the braking main air path, the prediction method obtains a leakage model by training based on the relevant parameters of the braking main air path of the subway train, calculates the effective data of the main air path pressure based on the running status data of the subway train, and inputs the effective data of the main air path pressure into the leakage model to predict the air pressure leakage of the train. Under the condition of limited sensor accuracy, the present invention calculates the effective data of the main air path pressure; improves the reliability and effectiveness of the pressure leakage prediction data, enhances the efficiency and accuracy of the main air path pressure prediction of the subway train; timely guides the maintenance work of the main air path of the subway train, and reduces the cost of repair and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 This is a flow chart of a method for predicting brake main air path pressure leakage according to embodiment 1 of the present invention.

[0083] Figure 2 This is a flow chart of step S3 of the method for predicting brake main air path pressure leakage in embodiment 1 of the present invention.

[0084] Figure 3 This is a flow chart of step S4 of the method for predicting brake main air path pressure leakage in embodiment 1 of the present invention.

[0085] Figure 4 This is a flow chart of step S43 of the method for predicting brake main air path pressure leakage in embodiment 1 of the present invention.

[0086] Figure 5 This is a module schematic diagram of a system for predicting brake main air path pressure leakage according to Embodiment 2 of the present invention.

[0087] Figure 6 This is a schematic diagram of the results of the initial training of the leakage model of the brake main air path pressure leakage prediction system of Example 2 of the present invention.

[0088] Figure 7 This is a schematic diagram of the results of completing the leakage model training of the brake main air path pressure leakage prediction system of Example 2 of the present invention.

[0089] Figure 8 This is a schematic diagram of the hardware structure of an electronic device according to Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0090] The present invention is further described below by way of examples, but the present invention is not limited to the scope of the examples.

[0091] Example 1

[0092] like Figure 1As shown, this embodiment provides a method for predicting brake main air path pressure leakage, the prediction method comprising the following steps:

[0093] S1. Obtaining the running status data of the target subway train in the target time period; the running status data includes parking confirmation parameters.

[0094] S2. Preprocess the running status data; the preprocessing includes integrity detection and abnormal data correction of the running status data; and generate preprocessing judgment information.

[0095] S3. Filter out parking status data corresponding to the longest parking time period of the target subway train from the running status data according to the parking confirmation parameters.

[0096] S4. Calculate and obtain effective data of the main air path pressure during the longest parking period based on the parking status data.

[0097] S5. Input the effective data of the main air path pressure into the leakage model, and output the leakage state of the braking main air path pressure of the target subway train in the target period. The leakage model is a mathematical model constructed by a two-dimensional threshold array and a two-dimensional attribute array. The dimensions of the two-dimensional threshold array include temperature and pressure. Each threshold in the two-dimensional threshold array represents the holding time threshold of the main air path pressure at a set temperature and a set pressure. The attribute value of the two-dimensional attribute array includes true value and false value.

[0098] In step S1, the driving data related to the train operation collected by sensors and other electronic devices on the target subway train is received at regular intervals. The operating status data related to the main air path pressure in the target period (from midnight on December 1, 2021 to 24:00 on December 2, 2021) is obtained from the driving data. The operating status data may include the current time, the main air path pressure value, the outside environment temperature value, and the parking confirmation parameters.

[0099] In step S2, first check the integrity of the operating status data. If the data is complete, further check whether the data format meets the preset requirements (for example, pkl or csv file). If it meets the data format requirements, finally determine whether the data volume meets the minimum data volume limit. If any of the above check results fails, a prompt message indicating that the data acquisition failed can be returned. Identify abnormal data points corresponding to the main air path pressure in the operating status data, such as single mutation data points and data points that exceed the preset range (for example, the preset range is 7.5bar~9.0bar). Correct the abnormal data points by replacing the nearest value, correcting the average value, correcting the default value, etc. If the correction fails, a prompt message indicating that the correction failed can be returned.

[0100] In step S3, multiple parking time periods are selected from the operation status data according to the parking confirmation parameter, and the longest parking time period is selected from the multiple parking time periods. The parking status data corresponding to the longest parking time period is further selected from the operation status data.

[0101] It can be understood that the parking confirmation parameter can be used as the first screening condition. In order to ensure the smoothness of the pressure change, the main air path pressure change interval less than or equal to 0.1 bar can also be used as the second screening condition, thereby determining the longest parking time period of the target subway train in the target period.

[0102] In step S4, based on the pressure values ​​corresponding to the non-repetitive main air path pressures appearing in the parking status data, the maximum pressure value and the minimum pressure value are removed, and finally invalid data such as the main air path pressure corresponding to the abnormal sampling moment are screened out to generate a pressure holding value set.

[0103] Each time, a target pressure holding value (for example, 8.5 bar) is selected from the pressure holding value set, and the parking state data is traversed to select at least one alternative pressure holding value greater than or equal to 8.5 bar and less than 8.6 bar. Further determine whether the sampling periods corresponding to the target pressure holding value and the alternative pressure holding value are reasonable. If reasonable, select the longest sampling period as the pressure holding time corresponding to the main air path pressure holding value of 8.5 bar for the target subway train within the target period. Similarly, select the next target pressure holding value (for example, 8.6 bar) from the pressure holding set, and determine the pressure holding time corresponding to the main air path pressure holding value of 8.6 bar for the target subway train within the target period in the above manner, until all pressure holding values ​​in the pressure holding set are traversed, and generate valid data on the main air path pressure.

[0104] In step S5, the effective data of the main air path pressure is input into the pre-trained leakage model, and the leakage state result of the pressure of the target subway train in the target period is output. It can be understood that the data state result of the effective data of the main air path pressure can also be output. For example, the data state result can include data integrity analysis, preprocessing analysis, parking time existence analysis, timeout point existence analysis and pressure holding time length analysis.

[0105] In one feasible solution, Figure 2 As shown, the parking confirmation parameters include train speed, cab activation coefficient and air compressor working coefficient; step S3 specifically includes:

[0106] S31. According to the train speed, the driver's cab activation coefficient and the air compressor working coefficient, at least one alternative parking time period is intercepted from the target time period.

[0107] S34: Filter out the longest period from the candidate parking time periods as the longest parking time period.

[0108] S35. Extract the parking status data corresponding to the longest parking time period from the running status data.

[0109] In step S31, the time periods in the running status data in which the train speed is zero, the activation coefficients of the front and rear driver cabs are both zero, and the air compressor working coefficient is zero are used as candidate parking time periods for the target subway train.

[0110] In another feasible solution, Figure 2 As shown, the running status data also includes time parameters and reference parameters, the reference parameters include the brake main air path pressure; the parking confirmation parameters include the train speed, the driver's cab activation coefficient and the air compressor working coefficient; step S3 specifically includes:

[0111] S32. According to the train speed, the driver's cab activation coefficient and the air compressor working coefficient, at least one preliminary parking time period is intercepted from the target time period.

[0112] S33: Compare every two adjacent air pressure change values ​​in the brake main air path pressure with the first threshold value, and determine at least one alternative parking time period from the initially selected parking time periods according to the comparison result.

[0113] S34: Filter out the longest period from the candidate parking time periods as the longest parking time period.

[0114] S35. Extract the parking status data corresponding to the longest parking time period from the running status data.

[0115] In step S32, the time periods in the running status data in which the train speed is zero, the activation coefficients of the front and rear driver cabs are both zero, and the air compressor working coefficient is zero are used as the primary parking time periods of the target subway train.

[0116] In step S33-step S35, the brake main air path pressures corresponding to the multiple pre-selected parking time periods are processed by preset data. Specifically, after removing the invalid pre-selected parking time periods whose adjacent air pressure change difference in the brake main air path pressure is greater than the first threshold (0.1 bar), the remaining pre-selected parking time periods are used as the alternative parking time periods, and the longest parking time period is selected from the alternative parking time periods, and the parking status data corresponding to the longest parking time period is further obtained. If all the pre-selected parking time periods do not meet the conditions, a prompt message indicating that the parking status data acquisition failed can be returned.

[0117] In one feasible solution, Figure 3 As shown, step S4 specifically includes:

[0118] S41, extracting multiple different air pressure values ​​from the parking state data, and deleting the maximum and minimum air pressure values ​​to generate an initial air pressure array.

[0119] S42: Eliminate the air pressure values ​​corresponding to the sampling time intervals greater than the second threshold from the initial air pressure array, and generate a pressure maintenance set and timeout judgment information.

[0120] S43: Generate an alternative pressure maintenance sample set and pressure maintenance preset judgment information according to the pressure maintenance set and the parking state data.

[0121] S44, comparing the number of samples in the candidate pressure holding sample set with the preset number. If the number of samples is not greater than the preset number, executing step S45, if the number of samples is greater than the preset number, executing step S46.

[0122] S45. Directly use the alternative pressure maintenance sample set as the effective pressure maintenance sample set, and generate alternative pressure maintenance sample judgment information and main air path pressure effective data.

[0123] S46. After sorting the samples in the candidate pressure maintenance set in descending order of pressure maintenance duration, a valid pressure maintenance sample set that meets a preset number is obtained by screening, and candidate pressure maintenance sample judgment information and main air path pressure valid data are generated.

[0124] The effective pressure holding sample set includes at least one of a pressure holding value, a holding temperature value, a pressure holding time and a starting time point.

[0125] In step S41, all non-repeating main air path pressure values ​​in the parking state data are counted. The non-repeating air pressure values ​​are [8.6 bar, 8.7 bar, 8.8 bar, 8.9 bar]. The maximum value of 8.9 bar and the minimum value of 8.6 bar are removed from the non-repeating air pressure values, and the initial air pressure array is generated as [8.7 bar, 8.8 bar].

[0126] In step S42, after removing two pressure values ​​whose sampling time interval is greater than a second threshold value (e.g., 10 minutes) from the initial pressure array, a pressure holding set of [8.7 bar, 8.8 bar] and timeout information is generated. The timeout information may include whether there is an abnormal sampling time interval data point in the initial pressure array, that is, a timeout data pressure point.

[0127] In step S43-step S46, the number of samples in the candidate pressure maintenance set sample is compared with the preset number N, and the effective pressure maintenance sample set is determined according to the comparison result, and then the effective pressure maintenance sample set is further used to generate the main air path pressure valid data and the candidate pressure maintenance sample judgment information. The candidate pressure maintenance sample judgment information includes information on whether the number of samples is too many, too few, or reasonable.

[0128] It should be noted that the effective data of the main air path pressure also includes pre-processing judgment information, the timeout judgment information, pressure maintenance preset judgment information and alternative pressure maintenance sample judgment information.

[0129] In one feasible solution, Figure 4 As shown, step S43 specifically includes:

[0130] S431. Select at least one target pressure maintenance value from the pressure maintenance value set.

[0131] S432: After comparing the difference between the air pressure value and the target pressure holding value in the parking state data with the third threshold, a plurality of candidate target pressure holding periods are intercepted from the longest parking period according to the comparison result.

[0132] S433, select the longest target pressure holding period from the candidate target pressure periods, and determine whether the longest target pressure holding period meets the preset conditions. If so, execute step S434, if not, execute step S431 until all pressure holding values ​​in the pressure holding value set are traversed.

[0133] S434. Generate a candidate pressure maintenance sample set according to the target pressure maintenance state data corresponding to the longest target pressure maintenance period intercepted from the parking state data.

[0134] In step S431, one target pressure holding value (e.g., 8.7 bar) can be selected from the pressure holding set at a time, and in step S432, multiple alternative target pressure holding values ​​greater than or equal to 8.7 bar and not exceeding one sensitivity unit (the third threshold, e.g., 0.1 bar) are screened out from the parking status data.

[0135] In step S433, after the longest target pressure holding period is selected from each candidate target pressure holding period, it is determined whether the sampling time corresponding to the longest target pressure holding period starts from the start parking time of the target subway train or ends at the end parking time of the target subway train. If so, the longest target pressure holding period is eliminated; if not, it is further determined whether the length of the pressure holding period corresponding to the longest target pressure holding period is less than 1 minute; if so, the longest target pressure holding period is eliminated; if not, the longest target pressure holding period is retained.

[0136] In an practicable solution, the prediction method trains a leakage model through the following steps, including:

[0137] The pressure maintenance data of multiple normal main air path pressures of subway trains in a historical period are collected.

[0138] The pressure maintenance data is input into the initial model for training to obtain the leakage model.

[0139] Specifically, the pressure maintenance data of the main air path pressure of a subway train in a healthy state (no leakage or a small amount of reasonable leakage) from December 21, 2018 to February 28, 2019 and from June 29, 2019 to July 9, 2019 are collected, and the pressure maintenance data of the two periods are input into the initial model for training in units of days. It can be understood that the initial state threshold arrays of the initial model are all set to 0, and the attribute arrays are all set to false.

[0140] The pressure maintenance data includes standard parameters, and the steps of inputting the pressure maintenance data into the initial model for training to obtain the leakage model include:

[0141] Select the pressure holding data where the standard parameters are within the set range as the effective pressure holding data;

[0142] Based on the pressure holding time in the effective pressure holding data, the pressure holding time threshold of the two-dimensional threshold array in the initial model is modified to generate a leakage model.

[0143] Specifically, when the standard parameters may include temperature and pressure, the temperature variation range is pre-set to -10-50°C with an interval of 1°C, and the pressure variation range is pre-set to 7.5-9bar with an interval of 0.1bar; the data in the pressure holding data that meets the above-mentioned standard parameter setting conditions at the same time is used as the effective pressure holding data.

[0144] The multiple pressure holding times under the set temperature and set pressure conditions in the effective pressure holding data are compared with the pressure holding time thresholds in the two-dimensional threshold array (array composed of corresponding temperature and pressure) in the initial model. The minimum pressure holding time under the set pressure and set temperature is updated to the pressure holding time threshold of the corresponding two-dimensional threshold array, and the attribute value of the corresponding two-dimensional attribute array is updated to true. According to the above training method, the pressure holding time under all other temperature and pressure conditions in the effective pressure holding data is compared with that in the initial model to complete the model training and generate the leakage model.

[0145] This embodiment provides a method for predicting the leakage of the braking main air path pressure, by training a leakage model based on the relevant parameters of the braking main air path of the subway train, calculating the effective data of the main air path pressure based on the running status data of the subway train, and inputting the effective data of the main air path pressure into the leakage model to predict the air pressure leakage of the train. Under the condition of limited sensor accuracy, the present invention calculates the effective data of the main air path pressure; improves the reliability and effectiveness of the pressure leakage prediction data, enhances the efficiency and accuracy of the main air path pressure prediction of the subway train; timely guides the maintenance work of the main air path of the subway train, and reduces the cost of repair and maintenance.

[0146] Example 2

[0147] like Figure 5 As shown, this embodiment provides a prediction system for brake main air path pressure leakage, and the prediction system includes: an acquisition module 210, a preprocessing module 220, a screening module 230, a calculation module 240 and a prediction module 250.

[0148] The acquisition module 210 is used to acquire the running status data of the target subway train in the target time period; the running status data includes parking confirmation parameters.

[0149] The preprocessing module 220 is used to preprocess the running status data; the preprocessing includes integrity detection and abnormal data correction of the running status data; and generates preprocessing judgment information.

[0150] The screening module 230 is used to screen out parking status data corresponding to the longest parking time period of the target subway train from the running status data according to the parking confirmation parameters.

[0151] The calculation module 240 is used to calculate and obtain effective data of the main air path pressure during the longest parking period based on the parking state data.

[0152] The prediction module 250 is used to input the effective data of the main air path pressure into the leakage model, and output the leakage state of the braking main air path pressure of the target subway train in the target period. The leakage model is a mathematical model constructed by a two-dimensional threshold array and a two-dimensional attribute array. The dimensions of the two-dimensional threshold array include temperature and pressure. Each threshold in the two-dimensional threshold array represents the holding time threshold of the main air path pressure at a set temperature and a set pressure. The attribute value of the two-dimensional attribute array includes true value and false value.

[0153] The acquisition module 210 periodically receives the driving data related to the train operation collected by the sensors and other electronic devices on the target subway train. The acquisition module 210 obtains the operation status data related to the main air path pressure in the target period (from midnight on December 1, 2021 to 24:00 on December 2, 2021) from the driving data. The operation status data may include the current time, the main air path pressure value, the outside environment temperature value and the parking confirmation parameters.

[0154] The preprocessing module 220 first checks the integrity of the operating status data. If the data is complete, it further detects whether the data format meets the preset requirements (for example, pkl or csv file). If it meets the data format requirements, it finally determines whether the data volume meets the minimum data volume limit. If any of the above check results fails, a prompt message indicating that the data acquisition failed can be returned. Identify abnormal data points corresponding to the main air path pressure in the operating status data, such as single mutation data points and data points that exceed the preset range (for example, the preset range is 7.5bar~9.0bar). The preprocessing module 220 corrects the abnormal data points by neighbor value replacement, average value correction, default value correction, etc. If the correction fails, a prompt message indicating that the correction failed can be returned.

[0155] A plurality of parking time periods are selected from the operation status data according to the parking confirmation parameter, and the longest parking time period is selected from the plurality of parking time periods. The further screening module 230 screens out the parking status data corresponding to the longest parking time period from the operation status data.

[0156] It can be understood that the parking confirmation parameter can be used as the first screening condition. In order to ensure the smoothness of the pressure change, the main air path pressure change interval less than or equal to 0.1 bar can also be used as the second screening condition, thereby determining the longest parking time period of the target subway train in the target period.

[0157] According to the pressure values ​​corresponding to the non-repetitive main air path pressures appearing in the parking status data, the maximum pressure value and the minimum pressure value are removed. Finally, the calculation module 240 filters out invalid data such as the main air path pressure corresponding to the abnormal sampling moment to generate a set of pressure holding values.

[0158] Each time, a target pressure holding value (for example, 8.5 bar) is selected from the pressure holding value set, and the parking state data is traversed to select at least one alternative pressure holding value greater than or equal to 8.5 bar and less than 8.6 bar. Further determine whether the sampling periods corresponding to the target pressure holding value and the alternative pressure holding value are reasonable. If reasonable, select the longest sampling period as the pressure holding time corresponding to the main air path pressure holding value of 8.5 bar for the target subway train within the target period. Similarly, select the next target pressure holding value (for example, 8.6 bar) from the pressure holding set, and determine the pressure holding time corresponding to the main air path pressure holding value of 8.6 bar for the target subway train within the target period in the above manner, until all pressure holding values ​​in the pressure holding set are traversed, and generate valid data on the main air path pressure.

[0159] The effective data of the main air path pressure is input into the pre-trained leakage model, and the prediction module 250 outputs the leakage state result of the pressure of the target subway train in the target period. It can be understood that the prediction module 250 can also output the data state result of the effective data of the main air path pressure. For example, the data state result may include data integrity analysis, pre-processing analysis, parking time existence analysis, timeout point existence analysis, and pressure holding time length analysis.

[0160] In an implementable solution, the parking confirmation parameters include train speed, cab activation coefficient and air compressor working coefficient; the screening module 230 includes:

[0161] The first interception unit 231 is used to intercept at least one alternative parking time period from the target time period according to the train speed, the driver's cab activation coefficient and the air compressor working coefficient.

[0162] The first screening unit 234 is configured to screen out the longest period from the candidate parking time periods as the longest parking time period.

[0163] The third intercepting unit 235 is used to intercept the parking state data corresponding to the longest parking time period from the running state data.

[0164] The first interception unit 231 uses the time periods in the running status data in which the train speed is zero, the activation coefficients of the front and rear driver cabs are both zero, and the air compressor working coefficient is zero as the candidate parking time periods for the target subway train.

[0165] In another feasible solution, Figure 5 As shown, the running status data also includes time parameters and reference parameters, the reference parameters include the brake main air path pressure; the parking confirmation parameters include the train speed, the driver's cab activation coefficient and the air compressor working coefficient; the screening module 230 includes:

[0166] The second interception unit 232 is used to intercept at least one preliminary parking time period from the target time period according to the train speed, the driver's cab activation coefficient and the air compressor working coefficient.

[0167] The first comparison unit 233 is used to compare every two adjacent pressure change values ​​in the brake main air path pressure with the first threshold value, and determine at least one candidate parking time period from the primary parking time periods according to the comparison result.

[0168] The first screening unit 234 is configured to screen out the longest period from the candidate parking time periods as the longest parking time period.

[0169] The third intercepting unit 235 is used to intercept the parking state data corresponding to the longest parking time period from the running state data.

[0170] The second interception unit 232 uses the time periods in the running status data in which the train speed is zero, the activation coefficients of the front and rear driver cabs are both zero, and the air compressor working coefficient is zero as the primary parking time period of the target subway train.

[0171] The brake main air path pressures corresponding to the multiple pre-selected parking time periods are processed by preset data. Specifically, after the first comparison unit 233 removes the invalid pre-selected parking time periods whose adjacent air pressure change differences in the brake main air path pressures are greater than the first threshold (0.1 bar), the remaining pre-selected parking time periods are used as the alternative parking time periods, and the longest parking time period is selected from the alternative parking time periods, and the parking status data corresponding to the longest parking time period is further obtained. If all the pre-selected parking time periods do not meet the conditions, a prompt message indicating that the parking status data acquisition failed can be returned.

[0172] In one practicable solution, the calculation module 240 includes:

[0173] The extraction unit 241 is used to extract multiple different air pressure values ​​in the parking state data, delete the maximum and minimum air pressure values, and generate an initial air pressure array.

[0174] The elimination unit 242 is used to eliminate the air pressure values ​​corresponding to the sampling time intervals greater than the second threshold from the initial air pressure array, and generate a pressure maintenance set and timeout judgment information.

[0175] The generating unit 243 is used to generate a candidate pressure maintenance sample set and pressure maintenance preset judgment information according to the pressure maintenance set and the parking state data.

[0176] The second comparison unit 244 is used to compare the number of samples in the candidate pressure maintenance sample set with a preset number.

[0177] The first processing unit 245 is used to directly use the alternative pressure maintenance sample set as the effective pressure maintenance sample set if the number of samples is not greater than the preset number, and generate alternative pressure maintenance sample judgment information and main air path pressure effective data.

[0178] The second processing unit 246 is used to sort the samples in the alternative pressure maintenance set in descending order of pressure maintenance time if the number of samples is greater than the preset number, screen out a valid pressure maintenance sample set that meets the preset number, and generate alternative pressure maintenance sample judgment information and main air path pressure valid data.

[0179] The effective pressure holding sample set includes at least one of a pressure holding value, a holding temperature value, a pressure holding time and a starting time point.

[0180] The extraction unit 241 counts all non-repeated pressure values ​​of the main air path pressure in the parking state data. The non-repeated pressure values ​​are [8.6 bar, 8.7 bar, 8.8 bar, 8.9 bar]. The maximum value 8.9 bar and the minimum value 8.6 bar of the non-repeated pressure values ​​are removed to generate an initial pressure array of [8.7 bar, 8.8 bar].

[0181] After removing two pressure values ​​whose sampling interval is greater than a second threshold (e.g., 10 minutes) from the initial pressure array, the elimination unit 242 generates a pressure retention set of [8.7 bar, 8.8 bar] and timeout information. The timeout information may include whether there are abnormal data points with sampling intervals in the initial pressure array, that is, timeout data pressure points.

[0182] The number of samples in the candidate pressure maintenance set samples is compared with the preset number N, and the effective pressure maintenance sample set is determined according to the comparison result. The first processing unit 245 or the second processing unit 246 further generates the effective data of the main air path pressure and the alternative pressure maintenance sample judgment information according to the effective pressure maintenance sample set. The alternative pressure maintenance sample judgment information includes information on whether the number of samples is too many, too few, or reasonable.

[0183] It should be noted that the effective data of the main air path pressure also includes pre-processing judgment information, the timeout judgment information, pressure maintenance preset judgment information and alternative pressure maintenance sample judgment information.

[0184] In an implementable solution, the generating unit 243 is specifically used for:

[0185] At least one target pressure hold value is selected from the set of pressure hold values.

[0186] After comparing the difference between the air pressure value and the target pressure holding value in the parking state data with the third threshold, a plurality of candidate target pressure holding periods are intercepted from the longest parking period according to the comparison result.

[0187] The longest target pressure holding period is selected from the candidate target pressure periods, and it is determined whether the longest target pressure holding period meets a preset condition.

[0188] If the target pressure maintaining state data corresponding to the longest target pressure maintaining period intercepted from the parking state data is satisfied, a candidate pressure maintaining sample set is generated.

[0189] If not, the step of selecting at least one target pressure holding value from the pressure holding value set is performed again until all pressure holding values ​​in the pressure holding value set are traversed.

[0190] The generation unit 243 can select one target pressure maintenance value (e.g., 8.7 bar) from the pressure maintenance set each time, and in step S432, screen out multiple alternative target pressure maintenance values ​​greater than or equal to 8.7 bar and not exceeding one sensitivity unit (the third threshold, e.g., 0.1 bar) from the parking status data.

[0191] After selecting the longest target pressure holding period from each candidate target pressure holding period, the generating unit 243 determines whether the sampling time corresponding to the longest target pressure holding period starts from the start parking time of the target subway train or ends at the end parking time of the target subway train. If so, the longest target pressure holding period is eliminated; if not, it is further determined whether the length of the pressure holding period corresponding to the longest target pressure holding period is less than 1 minute; if so, the longest target pressure holding period is eliminated; if not, the longest target pressure holding period is retained.

[0192] In an implementable solution, the prediction system obtains a leakage model by training the following modules, including:

[0193] The collection module is used to collect the pressure maintenance data of multiple normal main air path pressures of subway trains in a historical period.

[0194] The training module is used to input the pressure maintenance data into the initial model for training to obtain the leakage model.

[0195] Specifically, the acquisition module collects the pressure maintenance data of the main air path pressure of a subway train in a healthy state (no leakage or a small amount of reasonable leakage) from December 21, 2018 to February 28, 2019 and from June 29, 2019 to July 9, 2019. The pressure maintenance data of the two periods are input into the initial model for training in units of days through the training module. It can be understood that the initial state threshold arrays of the initial model are all set to 0, and the attribute arrays are all set to false.

[0196] The pressure maintenance data includes standard parameters and the training module includes:

[0197] The selection unit is used to select the pressure maintaining data whose standard parameters are within the set range as the effective pressure maintaining data.

[0198] The correction unit is used to correct the pressure holding time threshold of the two-dimensional threshold array in the initial model based on the pressure holding time in the effective pressure holding data to generate a leakage model.

[0199] Specifically, when the standard parameters may include temperature and pressure, the temperature variation range is pre-set to -10-50°C with an interval of 1°C, and the pressure variation range is set to 7.5-9bar with an interval of 0.1bar; the selection unit takes the data in the pressure holding data that meets the above-mentioned standard parameter setting conditions as the effective pressure holding data.

[0200] The multiple pressure holding times under the set temperature and set pressure conditions in the effective pressure holding data are compared with the initial model (such as Figure 6 The correction unit updates the minimum pressure holding time under the set pressure and set temperature to the pressure holding time threshold of the corresponding two-dimensional threshold array, and updates the attribute value of the corresponding two-dimensional attribute array to true. According to the above training method, the pressure holding time under all other temperature and pressure conditions in the effective pressure holding data is compared with that in the initial model to complete the model training (as shown in FIG. Figure 7 As shown), a leakage model is generated.

[0201] This embodiment provides a prediction system for brake main air path pressure leakage, which is based on the subway train brake main air path related parameters, and the leakage model is trained. The calculation module calculates the effective data of the main air path pressure based on the subway train running status data, and the prediction module inputs the effective data of the main air path pressure into the leakage model to predict the air pressure leakage of the train. Under the condition of limited sensor accuracy, the present invention calculates the effective data of the main air path pressure; improves the reliability and effectiveness of the pressure leakage prediction data, and enhances the efficiency and accuracy of the main air path pressure prediction of the subway train; timely guides the maintenance work of the main air path of the subway train, and reduces the cost of repair and maintenance.

[0202] Example 3

[0203] Figure 8 The structure diagram of an electronic device provided in this embodiment. The electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the method for predicting the pressure leakage of the brake main air path of embodiment 1 is implemented. Figure 8 The electronic device 60 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0204] The electronic device 60 may be in the form of a general-purpose computing device, for example, it may be a server device. The components of the electronic device 60 may include but are not limited to: at least one processor 61, at least one memory 62, and a bus 63 connecting different system components (including the memory 62 and the processor 61).

[0205] The bus 63 includes a data bus, an address bus, and a control bus.

[0206] The memory 62 may include a volatile memory, such as a random access memory (RAM) 621 and / or a cache memory 622 , and may further include a read-only memory (ROM) 623 .

[0207] The memory 62 may also include a program / utility 625 having a set (at least one) of program modules 624, such program modules 624 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0208] The processor 61 executes various functional applications and data processing by running the computer programs stored in the memory 62, such as the method for predicting the pressure leakage of the brake main air path of the first embodiment of the present invention.

[0209] The electronic device 60 may also communicate with one or more external devices 64 (e.g., keyboards, pointing devices, etc.). Such communication may be performed via an input / output (I / O) interface 65. Furthermore, the model-generated device 630 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 66. As shown, the network adapter 66 communicates with other modules of the model-generated device 60 via a bus 63. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the model-generated device 60, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, and data backup storage systems, etc.

[0210] It should be noted that although several units / modules or sub-units / modules of the electronic device are mentioned in the above detailed description, this division is merely exemplary and not mandatory. In fact, according to an embodiment of the present invention, the features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided into multiple units / modules to be embodied.

[0211] Example 4

[0212] This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the steps of the method for predicting the pressure leakage of the brake main air path of the embodiment 1 are implemented.

[0213] The readable storage medium may include but is not limited to: a portable disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical storage device, a magnetic storage device or any suitable combination of the above.

[0214] In a possible implementation, the present invention can also be implemented in the form of a program product, which includes a program code. When the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps of the method for predicting the brake main air path pressure leakage of Example 1.

[0215] The program code for executing the present invention may be written in any combination of one or more programming languages, and may be executed entirely on a user device, partially on a user device, as an independent software package, partially on a user device and partially on a remote device, or entirely on a remote device.

[0216] Although the specific embodiments of the present invention are described above, it should be understood by those skilled in the art that this is only for illustration and the protection scope of the present invention is defined by the appended claims. Those skilled in the art may make various changes or modifications to these embodiments without departing from the principles and essence of the present invention, but these changes and modifications all fall within the protection scope of the present invention.

Claims

1. A method for predicting brake main air path pressure leakage, characterized in that: The prediction method comprises: Acquiring the running status data of the target subway train in the target time period; the running status data includes parking confirmation parameters; According to the parking confirmation parameter, the parking status data corresponding to the longest parking time period of the target subway train is selected from the running status data; Calculate and obtain effective data of the main air path pressure during the longest parking period based on the parking state data; Input the effective data of the main air path pressure into the leakage model, and output the leakage state of the braking main air path pressure of the target subway train in the target period; the leakage model is a mathematical model constructed by a two-dimensional threshold array and a two-dimensional attribute array, each threshold in the two-dimensional threshold array represents the holding time threshold of the main air path pressure at a set temperature and a set pressure, the attribute value of the two-dimensional attribute array includes true value and false value, and the dimension of the two-dimensional threshold array includes temperature and pressure; The parking confirmation parameters include train speed, cab activation coefficient and air compressor working coefficient; The step of selecting the parking status data corresponding to the longest parking time period of the target subway train from the running status data according to the parking confirmation parameter comprises: According to the train speed, the driver's cab activation coefficient and the air compressor working coefficient, intercepting at least one alternative parking time period from the target time period; or, The operating status data also includes time parameters and reference parameters, the reference parameters include the brake main air path pressure; the parking confirmation parameters include train speed, driver's cab activation coefficient and air compressor working coefficient; According to the train speed, the driver's cab activation coefficient and the air compressor working coefficient, intercepting at least one preliminary parking time period from the target time period; Comparing every two adjacent pressure change values ​​of the brake main air path pressure with a first threshold, and determining at least one alternative parking time period from the primary parking time period according to the comparison result; Selecting the longest time period from the candidate parking time periods as the longest parking time period; The parking state data corresponding to the longest parking time period is intercepted from the running state data.

2. The method for predicting brake main air path pressure leakage according to claim 1, characterized in that: The step of calculating and obtaining effective data of the main air path pressure during the longest parking period based on the parking state data comprises: Extracting a plurality of different air pressure values ​​from the parking state data, and deleting the maximum and minimum air pressure values ​​to generate an initial air pressure array; Eliminate the air pressure values ​​corresponding to the sampling time intervals greater than the second threshold from the initial air pressure array, and generate a pressure holding set and timeout judgment information; generating a candidate pressure maintenance sample set and pressure maintenance preset judgment information according to the pressure maintenance set and the parking state data; Comparing the number of samples in the candidate pressure maintenance sample set with a preset number; If the number of samples is not greater than the preset number, the candidate pressure maintenance sample set is directly used as the effective pressure maintenance sample set, and the candidate pressure maintenance sample judgment information and the effective data of the main air path pressure are generated; If the number of samples is greater than the preset number, the samples in the candidate pressure maintenance set are sorted in descending order of pressure maintenance duration, and a valid pressure maintenance sample set that meets the preset number is obtained by screening, and the candidate pressure maintenance sample judgment information and the main air path pressure valid data are generated; The effective pressure holding sample set includes at least one of a pressure holding value, a holding temperature value, a pressure holding time and a starting time point.

3. The method for predicting brake main air path pressure leakage according to claim 2, characterized in that: The step of generating a candidate pressure maintenance sample set according to the pressure maintenance set and the parking state data comprises: selecting at least one target pressure maintenance value from the set of pressure maintenance values; After comparing the difference between the air pressure value in the parking state data and the target pressure holding value with a third threshold, a plurality of candidate target pressure holding periods are intercepted from the longest parking period according to the comparison result; Selecting a longest target pressure holding period from the candidate target pressure periods, and determining whether the longest target pressure holding period meets a preset condition; If satisfied, generating the candidate pressure maintenance sample set according to the target pressure maintenance state data corresponding to the longest target pressure maintenance period intercepted from the parking state data; If not, the step of selecting at least one target pressure holding value from the pressure holding value set is performed again until all pressure holding values ​​in the pressure holding value set are traversed.

4. The method for predicting brake main air path pressure leakage according to claim 1, characterized in that: The prediction method obtains the leakage model by training through the following steps, including: Collecting pressure maintenance data of multiple normal main air path pressures of subway trains in a historical period; The pressure maintenance data is input into the initial model for training to obtain the leakage model.

5. The method for predicting brake main air path pressure leakage according to claim 4, characterized in that: The pressure maintenance data includes standard parameters, and the step of inputting the pressure maintenance data into the initial model for training to obtain the leakage model includes: Select the pressure maintenance data whose standard parameters are within the set range as the effective pressure maintenance data; Based on the pressure holding time in the effective pressure holding data, the pressure holding time threshold of the two-dimensional threshold array in the initial model is corrected to generate the leakage model.

6. The method for predicting brake main air path pressure leakage according to claim 3, characterized in that: After obtaining the running status data of the target subway train in the target time period, the prediction method further includes: Preprocessing the running status data; the preprocessing includes performing integrity detection and abnormal data correction on the running status data, and generating preprocessing judgment information; Among them, the main air path pressure valid data includes the effective pressure maintenance sample set, the pre-processing judgment information, the timeout judgment information, the pressure maintenance preset judgment information and the alternative pressure maintenance sample judgment information.

7. A prediction system for brake main air path pressure leakage, characterized in that: The prediction system comprises: An acquisition module, used to acquire the running status data of the target subway train in the target period; the running status data includes parking confirmation parameters; A screening module, configured to screen out parking status data corresponding to the longest parking time period of the target subway train from the running status data according to the parking confirmation parameter; A calculation module, configured to calculate and obtain effective data of the main air path pressure during the longest parking period based on the parking state data; A prediction module is used to input the effective data of the main air path pressure into a leakage model, and output the leakage state of the braking main air path pressure of the target subway train in the target period; the leakage model is a mathematical model constructed by a two-dimensional threshold array and a two-dimensional attribute array, each threshold in the two-dimensional threshold array represents the holding time threshold of the main air path pressure at a set temperature and a set pressure, the attribute value of the two-dimensional attribute array includes true value and false value, and the dimension of the two-dimensional threshold array includes temperature and pressure; The parking confirmation parameters include train speed, cab activation coefficient and air compressor working coefficient; the screening module includes: A first interception unit is used to intercept at least one alternative parking time period from the target time period according to the train speed, the driver's cab activation coefficient and the air compressor working coefficient; or, The operating status data also includes time parameters and reference parameters, the reference parameters include the brake main air path pressure, and the parking confirmation parameters include train speed, driver's cab activation coefficient and air compressor working coefficient; the screening module includes: A second interception unit is used to intercept at least one preliminary parking time period from the target time period according to the train speed, the driver's cab activation coefficient and the air compressor working coefficient; a first comparison unit, configured to compare every two adjacent pressure change values ​​of the brake main air path pressure with a first threshold value, and determine at least one candidate parking time period from the primary parking time period according to the comparison result; A first screening unit, configured to screen out a longest period from the candidate parking time periods as the longest parking time period; The third intercepting unit is used to intercept the parking state data corresponding to the longest parking time period from the running state data.

8. The system for predicting brake main air path pressure leakage according to claim 7, characterized in that: The computing module comprises: An extraction unit, used to extract a plurality of different air pressure values ​​from the parking state data, and delete the maximum and minimum air pressure values ​​to generate an initial air pressure array; A removal unit, used to remove the air pressure values ​​corresponding to the sampling time intervals greater than the second threshold from the initial air pressure array, and generate a pressure holding set and timeout judgment information; A generating unit, configured to generate a candidate pressure maintenance sample set and pressure maintenance preset judgment information according to the pressure maintenance set and the parking state data; A second comparison unit, used for comparing the number of samples in the candidate pressure maintenance sample set with a preset number; A first processing unit, configured to directly use the candidate pressure maintenance sample set as a valid pressure maintenance sample set if the number of samples is not greater than the preset number, and generate candidate pressure maintenance sample judgment information and the main air path pressure valid data; A second processing unit is configured to, if the number of samples is greater than the preset number, sort the samples in the candidate pressure maintenance set in descending order of pressure maintenance duration, screen and obtain a valid pressure maintenance sample set that meets the preset number, and generate the candidate pressure maintenance sample judgment information and the main air path pressure valid data; The effective pressure holding sample set includes at least one of a pressure holding value, a holding temperature value, a pressure holding time and a starting time point.

9. The system for predicting brake main air path pressure leakage according to claim 8, characterized in that: The generating unit is specifically used for: selecting at least one target pressure maintenance value from the set of pressure maintenance values; After comparing the difference between the air pressure value in the parking state data and the target pressure holding value with a third threshold, a plurality of candidate target pressure holding periods are intercepted from the longest parking period according to the comparison result; Selecting a longest target pressure holding period from the candidate target pressure periods, and determining whether the longest target pressure holding period meets a preset condition; If satisfied, generating the candidate pressure maintenance sample set according to the target pressure maintenance state data corresponding to the longest target pressure maintenance period intercepted from the parking state data; If not, the step of selecting at least one target pressure holding value from the pressure holding value set is performed again until all pressure holding values ​​in the pressure holding value set are traversed.

10. The system for predicting brake main air path pressure leakage according to claim 7, characterized in that: The prediction system obtains the leakage model by training the following modules, including: A collection module, used for collecting pressure maintenance data of a plurality of normal main air path pressures of a subway train in a historical period; The training module is used to input the pressure maintenance data into the initial model for training to obtain the leakage model.

11. The system for predicting brake main air path pressure leakage according to claim 10, characterized in that: The pressure maintenance data includes standard parameters, and the training module includes: A selection unit, used for selecting the pressure maintenance data whose standard parameters are within a set range as effective pressure maintenance data; A correction unit is used to correct the pressure holding time threshold of the two-dimensional threshold array in the initial model based on the pressure holding time in the effective pressure holding data to generate the leakage model.

12. The system for predicting brake main air path pressure leakage according to claim 8, characterized in that: The prediction system also includes: A preprocessing module, used for preprocessing the running status data; the preprocessing includes performing integrity detection and abnormal data correction on the running status data, and generating preprocessing judgment information; Among them, the main air path pressure valid data includes the effective pressure maintenance sample set, the pre-processing judgment information, the timeout judgment information, the pressure maintenance preset judgment information and the alternative pressure maintenance sample judgment information.

13. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method for predicting brake main air path pressure leakage according to any one of claims 1 to 6 is implemented.

14. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for predicting the brake main air path pressure leakage according to any one of claims 1 to 6 is implemented.

Images