Heavy haul railway track maintenance decision-making method, device, equipment and medium

Abstract

The invention relates to the technical field of track maintenance, and provides a heavy haul railway track maintenance decision-making method, device, equipment and medium, and the method comprises the steps: storing historical detection data of a heavy haul railway track in a grid, and carrying out the standardization processing of the historical detection data in the grid; calculating a weight coefficient corresponding to the standardized historical detection data, and constructing a track comprehensive evaluation model of the heavy haul railway track according to the standardized historical detection data and the weight coefficient; training a track comprehensive evaluation model according to the track sample data, and calculating a comprehensive score of the heavy haul railway track corresponding to preset to-be-analyzed track data by using the trained track comprehensive evaluation model; classifying the maintenance categories of the heavy haul railway track by using the classification model and the comprehensive score to obtain a maintenance decision category; and determining the maintenance state of the heavy haul railway track according to the maintenance decision category. According to the invention, the accuracy of heavy haul railway track maintenance decision making can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of track maintenance decision-making, and in particular to a heavy-load railway track maintenance decision-making method, device, equipment and medium. Background Art

[0002] As the volume of railway freight increases, the load on the tracks will also increase. Tracks that have been used for a long time will gradually show signs of wear, deformation, or other aging phenomena, which may affect the safe operation of trains. Therefore, the tracks need to be inspected and maintained regularly to ensure their load-bearing capacity and safety.

[0003] The existing heavy-haul railway maintenance decision-making technology mainly evaluates the quality of railway tracks through static and dynamic monitoring data of the track, and formulates corresponding maintenance strategies in a graded manner. In practical applications, the evaluation based on static and dynamic monitoring data of the track has resulted in some data not being included in the management standard system, resulting in insufficient data utilization, and the evaluation of track health status is mainly based on thresholds and the measurement method is relatively discrete, resulting in low accuracy when making heavy-haul railway track maintenance decisions. Summary of the invention

[0004] In view of the above problems, embodiments of the present invention provide a heavy-duty railway track maintenance decision-making method, device, equipment and medium.

[0005] In a first aspect, an embodiment of the present invention provides a heavy-haul railway track maintenance decision-making method, comprising:

[0006] Collecting historical inspection data of heavy-duty railway tracks, storing the historical inspection data in a pre-built grid, and performing standardization processing on the historical inspection data stored in the grid;

[0007] Calculating the weight coefficient corresponding to the standardized historical detection data, and constructing a track comprehensive evaluation model for the heavy-duty railway track according to the standardized historical detection data and the weight coefficient;

[0008] The track comprehensive evaluation model is trained according to the preset track sample data, and the comprehensive score of the heavy-load railway track corresponding to the preset track data to be analyzed is calculated using the trained track comprehensive evaluation model;

[0009] Using the pre-trained classification model and the comprehensive score to classify the maintenance category of the heavy-duty railway track, and obtain the maintenance decision category of the heavy-duty railway track;

[0010] A maintenance status of the heavy haul railway track is determined according to the maintenance decision category.

[0011] According to an embodiment of the present invention, the historical detection data includes track flaw detection data, vehicle dynamics data and track quality index, wherein the vehicle dynamics data includes roadbed settlement data, vehicle body vertical acceleration, wavelength sensitivity, driving speed and structural vibration damage data.

[0012] According to an embodiment of the present invention, the step of standardizing the historical detection data stored in the grid includes:

[0013] The historical detection data stored in the grid is converted into a feature vector according to the preset grid column attributes;

[0014] The characteristic vector is standardized, wherein the formula for the standardized processing includes a positive indicator formula and a negative indicator formula, wherein the positive indicator formula is:

[0015]

[0016] Among them, x i ′ j is the normalized eigenvalue corresponding to the i-th row and j-th column in the grid cell, x ij is the eigenvalue corresponding to the i-th row and j-th column in the grid cell, x nj is the eigenvalue corresponding to the nth row and jth column in the grid cell, min(x 1j ,x 2j ,…,x nj ) is the minimum eigenvalue in the eigenvector, max(x 1j ,x 2j ,…,x nj ) is the maximum eigenvalue in the eigenvector;

[0017] The negative indicator formula is:

[0018]

[0019] Among them, x i ′ j is the normalized eigenvalue corresponding to the i-th row and j-th column in the grid cell, x ij is the eigenvalue corresponding to the i-th row and j-th column in the grid cell, x nj is the eigenvalue corresponding to the nth row and jth column in the grid cell, min(x 1j ,x 2j ,…,x nj ) is the minimum eigenvalue in the eigenvector, max(x 1j ,x 2j ,…,x nj ) is the maximum eigenvalue in the eigenvector.

[0020] According to an embodiment of the present invention, the calculating of the weight coefficient corresponding to the standardized historical detection data includes:

[0021] Calculate the ratio of the eigenvectors corresponding to the standardized historical detection data:

[0022]

[0023] Among them, y ij is the ratio corresponding to the i-th row and j-th column in the eigenvector, x i ′ j is the standardized eigenvalue corresponding to the i-th row and j-th column in the eigenvector, and m is the total number of historical detection data after standardization;

[0024] The information entropy of the feature vector is calculated according to the ratio:

[0025]

[0026] Among them, e j is the information entropy of the feature information of the jth column, y ij is the ratio corresponding to the i-th row and j-th column in the feature vector, m is the total number of historical detection data after standardization, and ln is a logarithmic function;

[0027] The weight coefficient corresponding to the standardized historical detection data is calculated according to the information entropy:

[0028]

[0029] Among them, W j is the weight coefficient corresponding to the standardized historical detection data corresponding to the jth column, e j is the information entropy of the j-th column feature information, and n is the total number of indicators.

[0030] According to an embodiment of the present invention, the track comprehensive evaluation model of the heavy-duty railway track is constructed according to the standardized historical detection data and the weight coefficient, including:

[0031] The track comprehensive evaluation model of the heavy-duty railway track is obtained by weighted summing the standardized historical detection data and the weight coefficients, wherein the track comprehensive evaluation model is:

[0032]

[0033] Among them, S is the comprehensive score in the track comprehensive evaluation model, x i ′ jis the standardized historical detection data corresponding to the i-th row and j-th column in the grid cell, W j is the weight coefficient corresponding to the standardized historical detection data corresponding to the jth column.

[0034] According to an embodiment of the present invention, the method of classifying the maintenance category of the heavy-haul railway track by using the pre-trained classification model and the comprehensive score to obtain the maintenance decision category of the heavy-haul railway track includes:

[0035] Extracting maintenance classification thresholds from a pre-trained classification model;

[0036] When the comprehensive score is greater than or equal to the maintenance classification threshold, determining the maintenance decision category of the heavy-load railway track as a preset first maintenance category;

[0037] When the comprehensive score is less than the maintenance classification threshold, the maintenance decision category of the heavy-load railway track is determined as a preset second maintenance category.

[0038] According to an embodiment of the present invention, before the maintenance category of the heavy-haul railway track is classified by using the pre-trained classification model and the comprehensive score to obtain the maintenance decision category of the heavy-haul railway track, the method further includes:

[0039] Based on the historical test data acquired in advance and the comprehensive score of the heavy-load railway track corresponding to the historical test data;

[0040] Using the comprehensive score to train a preset classification model to obtain a maintenance classification threshold corresponding to the classification model;

[0041] A trained classification model is generated according to the maintenance classification threshold.

[0042] In a second aspect, an embodiment of the present invention provides a heavy-duty railway track maintenance decision-making device, characterized in that it includes:

[0043] A standardization processing module, used for collecting historical detection data of heavy-load railway tracks, storing the historical detection data in a pre-built grid, and performing standardization processing on the historical detection data stored in the grid;

[0044] A track comprehensive evaluation model construction module is used to calculate the weight coefficient corresponding to the standardized historical detection data, and to construct a track comprehensive evaluation model for heavy-duty railway tracks based on the standardized historical detection data and the weight coefficient;

[0045] A comprehensive score calculation module, used to train the track comprehensive evaluation model according to preset track sample data, and use the trained track comprehensive evaluation model to calculate the comprehensive score of the heavy-load railway track corresponding to the preset track data to be analyzed;

[0046] A maintenance category classification module, used to classify the maintenance category of the heavy-duty railway track using a pre-trained classification model and the comprehensive score to obtain a maintenance decision category of the heavy-duty railway track;

[0047] The maintenance status determination module is used to determine the maintenance status of the heavy-load railway track according to the maintenance decision category.

[0048] In a third aspect, an embodiment of the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the multi-fuel cell system fan control method described in the above aspect.

[0049] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the heavy-load railway track maintenance decision-making method described in the above aspects.

[0050] Compared with the prior art, the above technical solution of the present invention has the following beneficial effects:

[0051] The embodiment of the present invention adds data on track flaw detection and vehicle dynamics to the evaluation process, which can improve the utilization rate of data, improve the monitoring and prediction of track health status, and adopt the classification algorithm in artificial intelligence and train the classification algorithm to establish a numerical range for preventive maintenance before reaching the damage threshold, so as to facilitate the guidance of operating units in making maintenance decisions. Therefore, the heavy-duty railway track maintenance decision-making method, device, equipment and medium proposed by the present invention can solve the problem of low accuracy when making heavy-duty railway track maintenance decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0053] Figure 1 A flowchart showing a heavy-haul railway track maintenance decision-making method according to Embodiment 1 of the present invention is shown;

[0054] Figure 2 A schematic diagram of a pre-built grid according to the first embodiment of the present invention is shown;

[0055] Figure 3 A functional module diagram of a heavy-duty railway track maintenance decision-making device according to a third embodiment of the present invention is shown;

[0056] Figure 4 A schematic diagram of the structure of an electronic device for implementing the heavy-load railway track maintenance decision-making method according to the fourth embodiment of the present invention is shown. DETAILED DESCRIPTION

[0057] The present disclosure is further described below in conjunction with the embodiments shown in the accompanying drawings.

[0058] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0059] The present invention proposes a heavy-duty railway track maintenance decision-making method based on artificial intelligence technology. Based on the theory of artificial intelligence algorithms and combined with the method of calculating weight coefficients, a track comprehensive evaluation model is built to achieve automatic classification of heavy-duty railway track maintenance decisions. Compared with traditional methods, the heavy-duty railway track maintenance decision-making technology has higher accuracy and reduces human subjectivity, and has great potential and application prospects in railway track maintenance decision-making applications.

[0060] Embodiment 1

[0061] like Figure 1 As shown, the present invention proposes a heavy-duty railway track maintenance decision-making method, comprising the following steps:

[0062] S1. Collect historical detection data of heavy-duty railway tracks, store the historical detection data in a pre-built grid, and perform standardization on the historical detection data stored in the grid.

[0063] In the embodiment of the present invention, the railway track profession has accumulated a large amount of track geometry parameters, track flaw detection operations, and wheel-rail relationship vehicle dynamics test data. By collecting track flaw detection and vehicle dynamics data in the historical inspection data of heavy-load railway tracks, the utilization rate of the data can be improved, and the track health status can be monitored with comprehensive data.

[0064] In an embodiment of the present invention, the historical detection data includes track flaw detection data, vehicle dynamics data and track quality index, wherein the vehicle dynamics data includes roadbed settlement data, vehicle vertical acceleration, wavelength sensitivity, driving speed and structural vibration damage data.

[0065] In detail, the track flaw detection data can be obtained by using flaw detection vehicles, such as the GTC80 rail flaw detection vehicle. These vehicles are equipped with advanced ultrasonic detection systems, which can conduct comprehensive inspections of the rails during continuous driving and arrange regular inspections during periods when trains are less operational, such as at night or in the early morning, to minimize interference with normal train operations. The detection range of vehicle dynamics damage to train tracks includes subgrade settlement data, vehicle body vertical acceleration, wavelength sensitivity, driving speed, and structural vibration damage data. Subgrade settlement data refers to uneven settlement of the subgrade that will lead to uneven track surface. This unevenness will have an adverse effect on train operation, increase the dynamic force between the wheel and rail, and thus cause damage to the track structure. The vertical acceleration of the vehicle body is affected by the subgrade settlement. The amplitude has a greater impact, indicating that during the operation of the train, the up and down vibration of the car body may cause impact on the track and cause damage; the dynamic response of the car body is more sensitive to uneven road settlement with a wavelength of less than 20 meters, indicating that short-wavelength unevenness is more easily perceived by the vehicle and may cause corresponding damage; with the increase of driving speed, the dynamic response of the vehicle will increase, and the wheel-rail force will increase significantly, which may aggravate the degree of track damage; using structural vibration damage identification technology to monitor the structural health of rail vehicles is an important technology, which can help to timely discover potential problems of the track and prevent the occurrence and development of damage; the track quality index (TQI) reflects the smoothness of the track state through the statistical results of dynamic detection data such as height, track direction, gauge, level and triangular pit.

[0066] Specifically, the data collection time is used as the pre-constructed grid row attribute, and the various track flaw detection data, roadbed settlement data, vehicle vertical acceleration, wavelength sensitivity, driving speed, structural vibration damage, and track quality index are used as pre-constructed grid column attributes, such as Figure 2 As shown in the figure, it is a schematic diagram of a pre-constructed grid, in which various data are collected based on the date, such as various data of track flaw detection, roadbed settlement data, vehicle vertical acceleration, wavelength sensitivity, driving speed, structural vibration damage, and track quality index data are collected every day, and the collected data are stored in the grid unit according to the pre-constructed grid data, so that the collected historical detection data is stored in the pre-constructed grid.

[0067] Furthermore, the collected historical detection data are not uniform in different units and scales. Therefore, it is necessary to convert the historical detection data of different units and scales into a uniform scale so that the historical detection data can be comparable during analysis and the distribution of the historical detection data can be more even.

[0068] In an embodiment of the present invention, the historical detection data is subjected to standard processing, that is, the historical detection data is normalized, and the data range of the standardized historical detection data will usually be compressed to between [0,1], so that different features can be compared at the same scale, and when training the machine learning model, the performance degradation of the algorithm due to different feature value scales can be avoided.

[0069] In the embodiment of the present invention, the process of standardizing the historical detection data stored in the grid includes:

[0070] The historical detection data stored in the grid is converted into a feature vector according to the preset grid column attributes;

[0071] The characteristic vector is standardized, wherein the formula for the standardized processing includes a positive indicator formula and a negative indicator formula, wherein the positive indicator formula is:

[0072]

[0073] Among them, x i ′ j is the normalized eigenvalue corresponding to the i-th row and j-th column in the grid cell, x ij is the eigenvalue corresponding to the i-th row and j-th column in the grid cell, x nj is the eigenvalue corresponding to the nth row and jth column in the grid cell, min(x 1j ,x 2j ,…,x nj ) is the minimum eigenvalue in the eigenvector, max(x 1j ,x 2j ,…,x nj ) is the maximum eigenvalue in the eigenvector;

[0074] The negative indicator formula is:

[0075]

[0076] Among them, x i ′ j is the normalized eigenvalue corresponding to the i-th row and j-th column in the grid cell, x ij is the eigenvalue corresponding to the i-th row and j-th column in the grid cell, x nj is the eigenvalue corresponding to the nth row and jth column in the grid cell, min(x 1j ,x 2j ,…,x nj ) is the minimum eigenvalue in the eigenvector, max(x 1j ,x 2j ,…,x nj) is the maximum eigenvalue in the eigenvector.

[0077] In detail, before data standardization, the historical detection data in the grid is first converted into a feature vector according to the preset grid column attributes. The feature vector is constructed based on the eigenvalue of each grid cell and represents the data of the grid cell in some dimensions, such as (x 1j ,x 2j ,…,x nj ) represents the characteristic values ​​of various track flaw detection data corresponding to the first column in the grid at different times, and the goal of standardization is to convert the characteristic values ​​to a unified scale. The standardization process includes two standardization methods: positive indicators and negative indicators.

[0078] Specifically, each eigenvalue in the grid can be standardized by a positive indicator formula, or by a negative indicator formula. The positive indicator is used for features whose larger eigenvalues ​​are more favorable, such as performance or efficiency, and the closer the standardized value is to 1, the better the feature is; the negative indicator is used for features whose smaller eigenvalues ​​are more favorable, and the closer the standardized value is to 1, the better the feature is, such as cost or error rate; for example, (x 1j ,x 2j ,…,x nj ) represents the characteristic values ​​of the various data of the track flaw detection corresponding to the first column in the grid at different times. 11 To standardize the corresponding eigenvalues, we need to select (x 1j ,x 2j ,…,x nj ) and then select the positive index formula or negative index formula for x 11 The corresponding eigenvalues ​​are standardized to obtain the standardized value of each eigenvalue.

[0079] Furthermore, through the collected historical inspection data, it is also necessary to evaluate the importance of each feature to the final result of the heavy-duty railway track. Therefore, it is necessary to calculate the weight coefficient corresponding to each historical inspection data. A higher weight means that the feature plays a more critical role in the model.

[0080] S2. Calculate the weight coefficient corresponding to the standardized historical detection data, and construct a comprehensive evaluation model for the heavy-duty railway track based on the standardized historical detection data and the weight coefficient.

[0081] In the embodiment of the present invention, the weight coefficient is used to indicate the importance of each feature. A higher weight coefficient indicates that the feature has a larger amount of information in the data and has a greater impact on the final result.

[0082] In the embodiment of the present invention, the weight coefficient corresponding to the historical detection data after the normalization processing is calculated, including:

[0083] Calculate the ratio of the eigenvectors corresponding to the standardized historical detection data:

[0084]

[0085] Among them, y ij is the ratio corresponding to the i-th row and j-th column in the eigenvector, x i ′ j is the standardized eigenvalue corresponding to the i-th row and j-th column in the eigenvector, and m is the total number of historical detection data after standardization;

[0086] The information entropy of the feature vector is calculated according to the ratio:

[0087]

[0088] Among them, e j is the information entropy of the feature information of the jth column, y ij is the ratio corresponding to the i-th row and j-th column in the feature vector, m is the total number of historical detection data after standardization, and ln is a logarithmic function;

[0089] The weight coefficient corresponding to the standardized historical detection data is calculated according to the information entropy:

[0090]

[0091] Among them, W j is the weight coefficient corresponding to the standardized historical detection data corresponding to the jth column, e j is the information entropy of the j-th column feature information, and n is the total number of indicators.

[0092] In detail, the weight coefficient W j is the information entropy based on the feature j Calculated to indicate the importance of each feature, information entropy e j It measures the uniformity or uncertainty of the jth column feature. The higher the information entropy, the more uniform the distribution of the feature and the smaller its contribution to the model. The lower the information entropy, the more concentrated the distribution of the feature, and the feature may be more discriminative. Thus, the weight coefficient corresponding to each data item of track flaw detection data, roadbed settlement data, vehicle vertical acceleration, wavelength sensitivity, driving speed, structural vibration damage, and track quality index in the historical detection data is obtained. For example, the weight coefficient corresponding to each data item of track flaw detection is W. 1 , the weight coefficient corresponding to the roadbed settlement data is W 2, the weight coefficient corresponding to the vertical acceleration of the vehicle body is W 3 By comparing the weight coefficients of different features, we can decide which features are most important for model analysis, and thus perform feature selection or weighted processing.

[0093] Furthermore, a comprehensive evaluation model for heavy-duty railway tracks can be constructed based on the weight coefficient corresponding to each indicator in the historical detection data. The comprehensive evaluation model can comprehensively consider multiple factors to help predict and identify potential safety hazards, thereby improving the safety of railway operations.

[0094] In the embodiment of the present invention, the track comprehensive evaluation model refers to a comprehensive score obtained by combining multiple detection data and features through weighted calculation to judge the overall quality and safety of heavy-duty railway tracks.

[0095] In an embodiment of the present invention, the track comprehensive evaluation model of the heavy-duty railway track is constructed according to the standardized historical detection data and the weight coefficient, including:

[0096] The track comprehensive evaluation model of the heavy-duty railway track is obtained by weighted summing the standardized historical detection data and the weight coefficients, wherein the track comprehensive evaluation model is:

[0097]

[0098] Among them, S is the comprehensive score in the track comprehensive evaluation model, x i ′ j is the standardized historical detection data corresponding to the i-th row and j-th column in the grid cell, W j is the weight coefficient corresponding to the standardized historical detection data corresponding to the jth column.

[0099] In detail, the standardized data and weight coefficients are used to construct the rail comprehensive evaluation model. The comprehensive score S of the model is obtained by weighted summing each standardized data point and its corresponding weight coefficient, that is, according to the determined indicators and weights, the rail comprehensive evaluation model is constructed using the weighted summation method, weighted sum = (value 1 × weight 1) + (value 2 × weight 2) + ... + (value n × weight n).

[0100] For example, if the various track flaw detection data, roadbed settlement data, vehicle vertical acceleration, wavelength sensitivity, driving speed, structural vibration damage, and track quality index collected on August 5, 2024 are used, then based on the weight coefficient corresponding to each index, the score of the heavy-duty railway track corresponding to August 5, 2024 can be obtained, and the comprehensive performance of the track can be evaluated based on the score, so as to decide whether the track needs maintenance on that day.

[0101] Furthermore, in order to make the constructed track comprehensive evaluation model more accurately reflect the actual status of the track in practical applications, it is necessary to use preset track sample data to optimize the weight coefficients and other parameters in the model, so that the trained track comprehensive evaluation model can more accurately predict potential track failures or problems, provide early warning information, and help formulate preventive maintenance plans.

[0102] S3. The track comprehensive evaluation model is trained according to the preset track sample data, and the comprehensive score of the heavy-load railway track corresponding to the preset track data to be analyzed is calculated using the trained track comprehensive evaluation model.

[0103] In an embodiment of the present invention, the track sample data refers to a large amount of historical sample data collected, including track flaw detection, vehicle dynamics, and track quality index (TQI) data, and the corresponding weight coefficient is calculated based on the track sample data. The weight coefficient usually represents the degree of influence of each feature on the comprehensive score, so that the score corresponding to each indicator data is obtained according to the weight coefficient.

[0104] In detail, the track flaw detection, vehicle dynamics, track quality index (TQI) data and their corresponding score labels are used as track sample data, and the sample data is divided into a training set and a test set. Usually, the training set accounts for 70%-80% of the total data, and the test set accounts for 20%-30%. The data of the training set is input into the track comprehensive evaluation model, and the trained model is evaluated using the test set. The prediction accuracy, error rate and other indicators of the model are calculated to determine the performance of the model, and then the trained track evaluation model is used to score the heavy-duty railway track corresponding to the track data to be analyzed.

[0105] In the embodiment of the present invention, the comprehensive score refers to the result of combining multiple evaluation indicators, which is used to measure the quality of heavy-duty railway tracks.

[0106] In an embodiment of the present invention, the method of calculating the comprehensive score of the heavy-load railway track corresponding to the preset track data to be analyzed by using the trained track comprehensive evaluation model includes: obtaining a weight coefficient corresponding to the track data to be analyzed; inputting the track data to be analyzed and the weight coefficient into the trained track comprehensive evaluation model to obtain the comprehensive score of the heavy-load railway track corresponding to the track data to be analyzed.

[0107] In detail, the track data to be analyzed refers to various track flaw detection data, roadbed settlement data, vehicle vertical acceleration, wavelength sensitivity, driving speed, structural vibration damage, and track quality index collected in real time on heavy-load railway tracks, and then the weight coefficient of each indicator is calculated. The corresponding value of each indicator and the corresponding weight coefficient of each indicator are input into the trained track comprehensive evaluation model, and the comprehensive score of the heavy-load railway track corresponding to the track data to be analyzed can be obtained.

[0108] Furthermore, the scoring data obtained from the comprehensive track evaluation model is input into the artificial intelligence algorithm, and the artificial intelligence algorithm is trained to establish the numerical range of preventive maintenance required before the damage threshold is reached, so as to guide the operating unit to carry out preventive maintenance.

[0109] S4. Classify the maintenance category of the heavy-load railway track using the pre-trained classification model and the comprehensive score to obtain the maintenance decision category of the heavy-load railway track.

[0110] In an embodiment of the present invention, a classification model is used to classify maintenance decisions for heavy-duty railway tracks according to a comprehensive score. The comprehensive score usually reflects the overall condition of the track, and the classification model can map this condition to a specific maintenance category. The maintenance decision categories include normal, light maintenance, moderate maintenance, and heavy maintenance. The normal category means that the track is in good condition and does not require immediate maintenance; light maintenance means that the track has minor problems and requires regular inspection and maintenance; moderate maintenance means that the track has obvious problems and requires more comprehensive inspection and repair; heavy maintenance means that the track has serious problems and requires comprehensive maintenance or replacement.

[0111] In the embodiment of the present invention, the maintenance category of the heavy-haul railway track is classified by using the pre-trained classification model and the comprehensive score to obtain the maintenance decision category of the heavy-haul railway track, including:

[0112] Extracting maintenance classification thresholds from a pre-trained classification model;

[0113] When the comprehensive score is greater than or equal to the maintenance classification threshold, determining the maintenance decision category of the heavy-load railway track as a preset first maintenance category;

[0114] When the comprehensive score is less than the maintenance classification threshold, the maintenance decision category of the heavy-load railway track is determined as a preset second maintenance category.

[0115] In detail, in the pre-trained classification model, there are usually one or more thresholds for classifying the comprehensive score into different maintenance categories. For example, the classification model (such as decision tree, logistic regression, support vector machine, etc.) uses a specific threshold to determine the category, and then extracts the maintenance classification threshold from the training parameters of the model, and then compares the comprehensive score with the extracted classification threshold to determine the maintenance category, and the first maintenance category refers to the normal category, which does not require maintenance; the second maintenance category refers to the maintenance required, including light maintenance, moderate maintenance and severe maintenance.

[0116] Specifically, the comprehensive score obtained from the actual heavy-load railway track state, when the comprehensive score is greater than or equal to the maintenance classification threshold: the maintenance decision category is determined as the preset first maintenance category; when the comprehensive score is less than the maintenance classification threshold: the maintenance decision category is determined as the preset second maintenance category. If the maintenance classification threshold of the model is 0.9, if the comprehensive score is 0.95, the maintenance decision category is determined as the normal category corresponding to the first maintenance category; if the comprehensive score is 0.75, the maintenance decision category is determined as light maintenance in the second maintenance category. The performance of the model can be checked regularly, and retrained or the threshold can be adjusted according to new data to ensure the classification accuracy and the effectiveness of the maintenance decision, where the maintenance category in the second maintenance category can be assumed to be 0.7 to 0.9 for light maintenance, 0.5 to 0.8 for moderate maintenance, and 0 to 0.5 for moderate maintenance.

[0117] Furthermore, the maintenance status of the heavy-duty railway track can be determined according to the classified maintenance decision categories, and then the corresponding maintenance plan can be taken for the heavy-duty railway track according to the maintenance status to ensure that the track is within the safe operating range, avoid potential failures or accidents, and improve maintenance efficiency.

[0118] S5. Determine the maintenance status of the heavy-load railway track according to the maintenance decision category.

[0119] In the embodiment of the present invention, the maintenance status of the heavy-duty railway track can be determined according to the maintenance decision category, that is, when the maintenance decision category is the normal category, the score is higher than the maintenance classification threshold, the track is in good condition and no maintenance is required, and the maintenance status is no maintenance required; the maintenance decision category is light maintenance, although there are some problems with the track, but the problems are minor, it is recommended to conduct regular inspections and small-scale repairs, and the maintenance status is light maintenance; the maintenance decision category is moderate maintenance, the track problems are more serious, and more comprehensive maintenance and repair are required to prevent further deterioration, and the maintenance status is moderate maintenance; the maintenance decision category is heavy maintenance, the track state is very poor, and large-scale repairs or replacements are required immediately to ensure safety, and the maintenance decision state is heavy maintenance. Through accurate status assessment, more effective long-term maintenance plans can be formulated to extend the service life of the track, thereby supporting the sustainable development of railway transportation, and maintenance resources and priorities can be reasonably allocated through clear division of maintenance categories. For tracks that are heavily maintained, more resources are allocated first to ensure that problems are solved in a timely manner; and tracks that are lightly maintained can be arranged in subsequent maintenance plans.

[0120] Embodiment 2

[0121] In order to more clearly understand the present invention, a second embodiment is used below to further explain the situation when the classification model is constructed in the embodiment of the present invention.

[0122] In an embodiment of the present invention, the classification model includes supervised learning. In supervised learning, the algorithm learns from labeled training data and then predicts unseen data. Common supervised learning algorithms include linear regression, support vector machine (SVM), logistic regression, decision tree, random forest and gradient enhancement, etc. Semi-supervised learning and reinforcement learning can also be used. Semi-supervised learning is between supervised learning and unsupervised learning, and uses a small amount of labeled data and a large amount of unlabeled data for learning.

[0123] In the embodiment of the present invention, before the maintenance category of the heavy-haul railway track is classified by using the pre-trained classification model and the comprehensive score to obtain the maintenance decision category of the heavy-haul railway track, the method further includes:

[0124] Based on the historical test data acquired in advance and the comprehensive score of the heavy-load railway track corresponding to the historical test data;

[0125] Using the comprehensive score to train a preset classification model to obtain a maintenance classification threshold corresponding to the classification model;

[0126] A trained classification model is generated according to the maintenance classification threshold.

[0127] In detail, historical inspection data and the comprehensive scores of heavy-duty railway tracks corresponding to these data are obtained. The historical inspection data and comprehensive score data are the basis for model training, and these historical data and comprehensive scores are used to train the preset classification model. During the training process, the model will learn how to classify according to the score, and the maintenance classification threshold is determined during the model training process. The maintenance classification threshold is the standard used by the model to distinguish different maintenance categories. Finally, the training results and the determined threshold are used to generate a trained classification model, and the classification model can be used for actual track maintenance decision classification.

[0128] The embodiment of the present invention adds data on track flaw detection and vehicle dynamics to the evaluation process, which can improve the utilization rate of data, improve the monitoring and prediction of track health status, and adopt the classification algorithm in artificial intelligence and train the classification algorithm to establish a numerical range for preventive maintenance before reaching the damage threshold, so as to facilitate the guidance of operating units in making maintenance decisions. Therefore, the heavy-duty railway track maintenance decision-making method, device, equipment and medium proposed by the present invention can solve the problem of low accuracy when making heavy-duty railway track maintenance decisions.

[0129] Embodiment 3

[0130] like Figure 3 As shown, this embodiment also provides a functional module diagram of a heavy-load railway track maintenance decision-making device.

[0131] The heavy-duty railway track maintenance decision device 100 described in this embodiment can be installed in an electronic device. According to the functions to be implemented, the heavy-duty railway track maintenance decision device 100 may include a standardization processing module 101, a track comprehensive evaluation model construction module 102, a comprehensive score calculation module 103, a maintenance category classification module 104 and a maintenance status determination module 105. The module described in the present invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0132] In this embodiment, the functions of each module / unit are as follows:

[0133] The standardization processing module 101 is used to collect historical detection data of heavy-duty railway tracks, store the historical detection data in a pre-constructed grid, and perform standardization processing on the historical detection data stored in the grid;

[0134] The track comprehensive evaluation model building module 102 is used to calculate the weight coefficient corresponding to the standardized historical detection data, and build a track comprehensive evaluation model for heavy-duty railway tracks according to the standardized historical detection data and the weight coefficient;

[0135] The comprehensive score calculation module 103 is used to train the track comprehensive evaluation model according to the preset track sample data, and use the trained track comprehensive evaluation model to calculate the comprehensive score of the heavy-load railway track corresponding to the preset track data to be analyzed;

[0136] The maintenance category classification module 104 is used to classify the maintenance category of the heavy-duty railway track using the pre-trained classification model and the comprehensive score to obtain the maintenance decision category of the heavy-duty railway track;

[0137] The maintenance status determination module 105 is used to determine the maintenance status of the heavy-load railway track according to the maintenance decision category.

[0138] In detail, each module described in the heavy-duty railway track maintenance decision-making device 100 described in the embodiment of the present invention adopts the same technical means as the heavy-duty railway track maintenance decision-making method described in Example 1 and Example 2 when used, and can produce the same technical effects, which will not be repeated here.

[0139] Embodiment 4

[0140] like Figure 4 As shown, this embodiment also provides a computer electronic device, which may include a processor 10, a memory 11, a communication bus 12 and a communication interface 13, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as a heavy-load railway track maintenance decision program.

[0141] In some embodiments, the processor 10 may be composed of an integrated circuit, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips. The processor 10 is the control core (ControlUnit) of the electronic device, and uses various interfaces and lines to connect various components of the entire electronic device, and executes or executes programs or modules stored in the memory 11 (for example, executing a heavy-load railway track maintenance decision program, etc.), and calls data stored in the memory 11 to execute various functions of the electronic device and process data.

[0142] The memory 11 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (for example, an SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. In some embodiments, the memory 11 may be an internal storage unit of an electronic device, such as a mobile hard disk of the electronic device. In other embodiments, the memory 11 may also be an external storage device of an electronic device, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device. Further, the memory 11 may also include both an internal storage unit of the electronic device and an external storage device. The memory 11 may not only be used to store application software and various types of data installed in the electronic device, such as the code of the heavy-duty railway track maintenance decision program, but also be used to temporarily store data that has been output or is to be output.

[0143] The communication bus 12 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is configured to realize connection and communication between the memory 11 and at least one processor 10, etc.

[0144] The communication interface 13 is used for communication between the above-mentioned electronic device and other devices, including a network interface and a user interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device and other electronic devices. The user interface may be a display (Display), an input unit (such as a keyboard (Keyboard)), and optionally, the user interface may also be a standard wired interface, a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode, organic light-emitting diode) touch device, etc. Among them, the display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device and to display a visual user interface.

[0145] The figure only shows an electronic device with components. Those skilled in the art will understand that the structure shown in the figure does not constitute a limitation on the electronic device, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0146] For example, although not shown, the electronic device may also include a power source (such as a battery) for supplying power to each component. Preferably, the power source may be logically connected to the at least one processor 10 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

[0147] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

[0148] The heavy-load railway track maintenance decision program stored in the memory 11 in the electronic device is a combination of multiple instructions. When running in the processor 10, it can achieve:

[0149] Collecting historical inspection data of heavy-duty railway tracks, storing the historical inspection data in a pre-built grid, and performing standardization processing on the historical inspection data stored in the grid;

[0150] Calculating the weight coefficient corresponding to the standardized historical detection data, and constructing a track comprehensive evaluation model for the heavy-duty railway track according to the standardized historical detection data and the weight coefficient;

[0151] The track comprehensive evaluation model is trained according to the preset track sample data, and the comprehensive score of the heavy-load railway track corresponding to the preset track data to be analyzed is calculated using the trained track comprehensive evaluation model;

[0152] Using the pre-trained classification model and the comprehensive score to classify the maintenance category of the heavy-duty railway track, and obtain the maintenance decision category of the heavy-duty railway track;

[0153] A maintenance status of the heavy haul railway track is determined according to the maintenance decision category.

[0154] Specifically, the specific implementation method of the processor 10 for the above instructions can refer to the description of the relevant steps in the corresponding embodiment of the accompanying drawings, which will not be repeated here.

[0155] Furthermore, if the module / unit integrated in the electronic device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0156] Embodiment 5

[0157] This embodiment provides a storage medium storing a computer program. When the computer program is executed by a processor, the steps of the heavy-duty railway track maintenance decision-making method described above are implemented.

[0158] These program codes can also be loaded onto a computer or other programmable data processing device so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions executed on the computer or other programmable device for implementing the process. Figure 1 The steps of a specified function in a process or multiple processes.

[0159] Storage media include permanent and non-permanent, removable and non-removable media, and can be implemented by any method or technology to store information. Information can be computer-readable instructions, data structures, modules of programs or other data. Examples of storage media can include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission medium that can be used to store information that can be accessed by a computing device.

[0160] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0161] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0162] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0163] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0164] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is not limited only according to the above description, and it is intended that all changes within the meaning and scope of equivalent elements within the scope of protection are included in the present invention.

[0165] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0166] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

[0167] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A heavy-haul railway track maintenance decision-making method, characterized in that: The method comprises: Collecting historical inspection data of heavy-duty railway tracks, storing the historical inspection data in a pre-built grid, and performing standardization processing on the historical inspection data stored in the grid; Calculating the weight coefficient corresponding to the standardized historical detection data, and constructing a track comprehensive evaluation model for the heavy-duty railway track according to the standardized historical detection data and the weight coefficient; The track comprehensive evaluation model is trained according to the preset track sample data, and the comprehensive score of the heavy-load railway track corresponding to the preset track data to be analyzed is calculated using the trained track comprehensive evaluation model; Using the pre-trained classification model and the comprehensive score to classify the maintenance category of the heavy-duty railway track, and obtain the maintenance decision category of the heavy-duty railway track; A maintenance status of the heavy haul railway track is determined according to the maintenance decision category.

2. The heavy-haul railway track maintenance decision-making method according to claim 1, characterized in that: The historical detection data includes track flaw detection data, vehicle dynamics data and track quality index, wherein the vehicle dynamics data includes roadbed settlement data, vehicle body vertical acceleration, wavelength sensitivity, driving speed and structural vibration damage data.

3. The heavy-haul railway track maintenance decision-making method according to claim 1, characterized in that: The process of standardizing the historical detection data stored in the grid includes: The historical detection data stored in the grid is converted into a feature vector according to the preset grid column attributes; The characteristic vector is standardized, wherein the formula for the standardized processing includes a positive indicator formula and a negative indicator formula, wherein the positive indicator formula is: Among them, x i ′ j is the normalized eigenvalue corresponding to the i-th row and j-th column in the grid cell, x ij is the eigenvalue corresponding to the i-th row and j-th column in the grid cell, x nj is the eigenvalue corresponding to the nth row and jth column in the grid cell, min(x 1j ,x 2j ,…,x nj ) is the minimum eigenvalue in the eigenvector, max(x 1j ,x 2j ,…,x nj ) is the maximum eigenvalue in the eigenvector; The negative indicator formula is: Among them, x i ′ j is the normalized eigenvalue corresponding to the i-th row and j-th column in the grid cell, x ij is the eigenvalue corresponding to the i-th row and j-th column in the grid cell, x nj is the eigenvalue corresponding to the nth row and jth column in the grid cell, min(x 1j ,x 2j ,…,x nj ) is the minimum eigenvalue in the eigenvector, max(x 1j ,x 2j ,…,x nj ) is the maximum eigenvalue in the eigenvector.

4. The heavy-haul railway track maintenance decision-making method according to claim 1, characterized in that: The calculation of the weight coefficient corresponding to the standardized historical detection data includes: Calculate the ratio of the eigenvectors corresponding to the standardized historical detection data: Among them, y ij is the ratio corresponding to the i-th row and j-th column in the eigenvector, x i ′ j is the standardized eigenvalue corresponding to the i-th row and j-th column in the eigenvector, and m is the total number of historical detection data after standardization; The information entropy of the feature vector is calculated according to the ratio: Among them, e j is the information entropy of the feature information of the jth column, y ij is the ratio corresponding to the i-th row and j-th column in the feature vector, m is the total number of historical detection data after standardization, and ln is a logarithmic function; The weight coefficient corresponding to the standardized historical detection data is calculated according to the information entropy: Among them, W j is the weight coefficient corresponding to the standardized historical detection data corresponding to the jth column, e j is the information entropy of the j-th column feature information, and n is the total number of indicators.

5. The heavy-haul railway track maintenance decision-making method according to claim 1, characterized in that: The track comprehensive evaluation model of the heavy-duty railway track is constructed according to the standardized historical detection data and the weight coefficient, including: The track comprehensive evaluation model of the heavy-duty railway track is obtained by weighted summing the standardized historical detection data and the weight coefficients, wherein the track comprehensive evaluation model is: Among them, S is the comprehensive score in the track comprehensive evaluation model, x i ′ j is the standardized historical detection data corresponding to the i-th row and j-th column in the grid cell, W j is the weight coefficient corresponding to the standardized historical detection data corresponding to the jth column.

6. The heavy-haul railway track maintenance decision-making method according to claim 1, characterized in that: The maintenance category of the heavy-duty railway track is classified by using the pre-trained classification model and the comprehensive score to obtain the maintenance decision category of the heavy-duty railway track, including: Extracting maintenance classification thresholds from a pre-trained classification model; When the comprehensive score is greater than or equal to the maintenance classification threshold, determining the maintenance decision category of the heavy-load railway track as a preset first maintenance category; When the comprehensive score is less than the maintenance classification threshold, the maintenance decision category of the heavy-load railway track is determined as a preset second maintenance category.

7. The heavy-haul railway track maintenance decision-making method according to claim 6, characterized in that: Before the maintenance category of the heavy-haul railway track is classified by using the pre-trained classification model and the comprehensive score to obtain the maintenance decision category of the heavy-haul railway track, the method further includes: Based on the historical test data acquired in advance and the comprehensive score of the heavy-load railway track corresponding to the historical test data; Using the comprehensive score to train a preset classification model to obtain a maintenance classification threshold corresponding to the classification model; A trained classification model is generated according to the maintenance classification threshold.

8. A heavy-duty railway track maintenance decision-making device, characterized in that: The device comprises: A standardization processing module, used for collecting historical detection data of heavy-load railway tracks, storing the historical detection data in a pre-built grid, and performing standardization processing on the historical detection data stored in the grid; A track comprehensive evaluation model construction module is used to calculate the weight coefficient corresponding to the standardized historical detection data, and to construct a track comprehensive evaluation model for heavy-duty railway tracks based on the standardized historical detection data and the weight coefficient; A comprehensive score calculation module, used to train the track comprehensive evaluation model according to preset track sample data, and use the trained track comprehensive evaluation model to calculate the comprehensive score of the heavy-load railway track corresponding to the preset track data to be analyzed; A maintenance category classification module, used to classify the maintenance category of the heavy-duty railway track using a pre-trained classification model and the comprehensive score to obtain a maintenance decision category of the heavy-duty railway track; The maintenance status determination module is used to determine the maintenance status of the heavy-load railway track according to the maintenance decision category.

9. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the heavy-haul railway track maintenance decision-making method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the heavy-load railway track maintenance decision-making method described in any one of claims 1 to 7 are implemented.

Images