Fault diagnosis method and device for high-speed rail running gear, electronic equipment and storage medium
Through a fault diagnosis method based on a multidimensional confidence rule base, combined with fuzzy C-mean clustering algorithm and evidence reasoning, and using an adaptive evolution strategy optimization algorithm, the problem of inaccurate complex fault diagnosis of high-speed railway walking department is solved, achieving higher diagnosis Accuracy and interpretability meet the safety requirements of high-speed railways.
Patent Information
- Application Number
- CN202411411520.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-11
- Publication Date
- 2024-11-12
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The prior art is difficult to accurately diagnose complex faults in high-speed railway traveling parts, especially in various and interrelated fault modes, which cannot provide sufficient diagnostic accuracy and cannot meet the high standards for high-speed railway safety.
The fault diagnosis method based on the multidimensional confidence rule base is adopted, and by obtaining running sample data, extracting feature quantities, building a confidence rule model, using the fuzzy C-mean clustering algorithm to determine the prerequisite attribute reference value, performing fusion inference based on evidence reasoning, and using projection association The adaptive evolution strategy optimization algorithm of the variance matrix updates the model parameters and improves the diagnostic accuracy.
It improves the accuracy and interpretability of fault diagnosis of high-speed railway travel department, can capture complex fault information more accurately, and meets the high standards for high-speed railway safety.
Smart Images

Figure CN118940077A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing, and in particular to a fault diagnosis method, device, electronic equipment and storage medium for a high-speed railway running gear. Background Art
[0002] The rapid development of high-speed rail technology has greatly promoted the progress of modern transportation and provided passengers with faster and more convenient travel options. However, as the speed of high-speed trains increases, their safety issues have increasingly attracted the attention of the public and industry experts. The safety of high-speed railways is not only the basis for ensuring the safe operation of trains, but also directly related to the life safety and comfort experience of passengers, as well as the public's trust and acceptance of the entire railway system.
[0003] The running gear refers to the lower part of the locomotive that guides the vehicle along the track and transfers the entire weight of the locomotive to the rails. It is the power heart of the high-speed rail. It converts electrical energy into mechanical energy to drive the train forward. The stability of this component directly affects the operating efficiency and safety of the train. Any slight performance deviation or failure may cause the train to run at an unstable speed or even cause safety accidents such as derailment or brake failure. In addition, a running gear motor failure may also cause the train to stop operating, affecting the efficiency and reputation of railway transportation. Therefore, it is particularly critical to maintain the performance and reliability of the running gear of high-speed trains.
[0004] From a technical perspective, the running gear of a high-speed train is a complex system that contains multiple highly interactive components, including motors, bogies, brake systems, suspension systems, etc. The good operation of each component is the prerequisite for ensuring the stability of the overall system. In such a complex system, each component may show different degrees of performance fluctuations due to changes in the external environment, such as temperature changes, humidity, mechanical wear, etc., making the dynamic and nonlinear characteristics of the system more obvious. Therefore, for such a system, it is difficult to meet the needs by relying solely on traditional fault diagnosis methods, especially when faced with situations where fault modes are diverse and interrelated. One-dimensional fault diagnosis methods cannot effectively capture and analyze complex fault information caused by the interaction of multiple factors, cannot provide sufficient diagnostic accuracy, and cannot meet the high standards of high-speed railway safety.
[0005] Therefore, how to improve the diagnostic accuracy and at the same time improve the interpretability of the high-speed railway running gear during the modeling process is the research content of establishing a high-speed railway running gear fault diagnosis model. Summary of the invention
[0006] The present invention provides a fault diagnosis method, device, electronic equipment and storage medium for a high-speed railway running gear, so as to solve the technical problem of inaccurate fault diagnosis of a high-speed railway running gear.
[0007] According to one aspect of the present invention, a method for diagnosing faults of a high-speed railway running gear is provided, comprising: obtaining operating sample data of the high-speed railway running gear, extracting feature quantities of the operating sample data, and constructing a fault diagnosis model based on a multidimensional confidence rule base, wherein the fault diagnosis model comprises a plurality of confidence rules defining a mapping relationship between premise attributes and fault conclusions; using a fuzzy C-means clustering algorithm to determine premise attribute reference values in the fault diagnosis model; using evidential reasoning to perform fusion reasoning on the confidence rules in the fault diagnosis model based on the premise attribute reference values; using an adaptive evolutionary strategy optimization algorithm of a projection covariance matrix to update parameters in the fault diagnosis model after fusion reasoning; collecting operating data of the high-speed railway running gear to be diagnosed, inputting the data into the updated fault diagnosis model, and outputting the fault diagnosis result of the high-speed railway running gear to be diagnosed.
[0008] Further, the extracting of the characteristic quantity of the running sample data and constructing a fault diagnosis model based on a multi-dimensional confidence rule base includes: Preprocessing the running sample data; The preprocessed running sample data is subjected to feature extraction, from which n premise attributes and health status levels for fault diagnosis are extracted, and the following fault diagnosis model is constructed: ;in, Indicates confidence rules; Indicates the first Prerequisite attributes; Indicates Rule No. Prerequisite attribute reference values; Indicates the level of health status; Indicates Results Confidence level; Indicates an “and” relationship; Indicates the number of evaluation results; Indicates The rule weight of the rule; Indicates The weight of the premise attributes; Determine the rule weights in the fault diagnosis model , Prerequisite attribute weight .
[0009] Further, the method of using the fuzzy C-means clustering algorithm to determine the reference value of the premise attribute in the fault diagnosis model includes: Initialize and set c initial reference values of premise attributes as initial cluster centers; Repeat the following steps to iteratively update the initial cluster center until the change of the cluster center is less than a preset threshold or reaches a preset number of iterations: Calculating the distance from the actual value of the premise attribute in the running sample data to the cluster center; According to the distance between the actual value of the premise attribute and the cluster center, the membership function is used to calculate the membership of the actual value relative to the cluster center. The expression is as follows: ;in, Indicates Data points and The membership degree of the cluster center, that is, The data point belongs to The degree of cluster centers; is the index of the data point, indicating the data points; is the index of the cluster center, indicating the Cluster centers; is the count index of the cluster center, representing the cluster center from 1 to c; Indicates Data points and The distance measure between cluster centers; Indicates The cluster centers and The distance measure between cluster centers; It is a fuzzy index or a weighted index; is the number of cluster centers; The cluster center is updated by weighted average algorithm according to the membership degree, and the expression is as follows: ;in, is the new weighted cluster center; Indicates The coordinates or attribute values of the data points; Represents the total number of all data points considered when calculating the cluster center; The cluster center after the iteration is completed is used as the reference value of the premise attribute in the fault diagnosis model.
[0010] Further, the adopting evidential reasoning to perform fusion reasoning on the confidence rules in the fault diagnosis model based on the premise attribute reference value includes: According to the premise attribute reference value and the premise attribute actual value in the running sample data, the matching degree of each premise attribute of each confidence rule in the fault diagnosis model is calculated, and the expression is as follows: ;in, Indicates Rule No. The matching degree of the premise attributes; and Represents two adjacent rules , The Prerequisite attribute reference values; Indicates the first The actual input value of the premise attribute; Indicates the rule number currently being processed; Indicates the number of premise attributes; The matching degree of each rule is calculated based on the matching degree of each premise attribute. The expression is as follows: ;in, Indicates The matching degree of the rules; Indicates The weight of the premise attributes; The activation weight of each rule is calculated according to the matching degree of each rule. The expression is as follows: ;in, Indicates The activation weight of the rule; Indicates The rule weight of the rule; Indicates The rule weight of the rule; Indicates The matching degree of the rules is used to sum up and represent the matching degree of multiple rules; Represents the number of rules in the fault diagnosis model; Based on evidence reasoning, the confidence rules of the fault diagnosis model are inferred by the following formula: ; ;in, Represents the fault diagnosis model output relative to the evaluation result The confidence level is determined by the rule weight , Prerequisite attributes Attribute weights and confidence The composed function.
[0011] Furthermore, the adaptive evolution strategy optimization algorithm using the projection covariance matrix is used to update the parameters in the fault diagnosis model after fusion reasoning, including: Constructing the objective function : ;in, Indicates the termination time of the iteration process; Indicates the current iteration step number; Indicates the starting time point of the iteration process; For the objective function The constraints make and As close as possible, the expression is as follows: ;in, Indicates The actual output parameter or weight of the rule.
[0012] According to another aspect of the present invention, a high-speed railway running gear fault diagnosis device is provided, comprising: A data acquisition module is used to collect sample data for building a model and to collect operating data of a high-speed railway running unit to be diagnosed; a data extraction module is used to extract feature quantities of sample data and operating data; a model construction module is used to build a fault diagnosis model based on a multidimensional confidence rule base, wherein the fault diagnosis model includes a number of confidence rules that define a mapping relationship between premise attributes and fault conclusions; a clustering algorithm module is used to determine premise attribute reference values in the fault diagnosis model using a fuzzy C-means clustering algorithm; a fusion reasoning module is used to perform fusion reasoning on the confidence rules in the fault diagnosis model based on the premise attribute reference values using evidence reasoning; an evolution strategy module is used to update the parameters in the fault diagnosis model after fusion reasoning using an adaptive evolution strategy optimization algorithm of a projection covariance matrix; a fault diagnosis module is used to input the operating data of the high-speed railway running unit to be diagnosed into the updated fault diagnosis model, and output the fault diagnosis result of the high-speed railway running unit to be diagnosed.
[0013] According to another aspect of the present invention, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the steps in the above method when executed by the processor.
[0014] According to another aspect of the present invention, a storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above method are implemented.
[0015] The present invention proposes a method, device, electronic device and storage medium for diagnosing faults of a high-speed railway running gear. By constructing a high-speed railway running gear fault diagnosis model based on a multidimensional confidence rule base, a plurality of rules for characterizing the mapping relationship between premise attributes and fault conclusions are established, so that various and interrelated fault causes can be comprehensively considered. The fuzzy C-means clustering algorithm is used to determine the reference values of the premise attributes in the diagnosis model. Compared with the method in which experts provide reference values in traditional calculations, the subjectivity of expert knowledge is reduced and the accuracy of model diagnosis is improved. Afterwards, the confidence rules in the model are fused and inferred based on evidence reasoning, and the adaptive evolution strategy optimization algorithm of the projection covariance matrix is used to update the parameters in the fault diagnosis model after fusion reasoning, and the parameters of the fault diagnosis model are optimized by the algorithm, so that the model diagnosis result is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings: Figure 1 is a flow chart of a high-speed railway running gear fault diagnosis method according to an embodiment of the present invention; Figure 2 It is a structural block diagram of a high-speed railway running gear fault diagnosis device according to an embodiment of the present invention. DETAILED DESCRIPTION
[0017] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only embodiments of a part of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present application. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.
[0018] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.
[0019] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0020] Embodiment 1: In this embodiment, a method for diagnosing faults of a high-speed railway running gear is provided. The method embodiment provided in this embodiment can be executed in a server, a computer, a mobile phone, or a similar computing device.
[0021] Figure 1 is a flow chart of a high-speed railway running gear fault diagnosis method according to an embodiment of the present invention. Figure 1 As shown, the process includes: S1: obtaining operation sample data of a high-speed railway running unit, extracting feature quantities of the operation sample data, and constructing a fault diagnosis model based on a multidimensional confidence rule base, wherein the fault diagnosis model includes a plurality of confidence rules defining a mapping relationship between premise attributes and fault conclusions; The fault diagnosis model based on multidimensional belief rule base is a system fault diagnosis method based on belief rule base (BRB). It analyzes the input high-speed railway running gear operation information by defining and combining a series of belief rules, and performs reasoning and calculation according to the rules in the rule base to obtain the fault diagnosis results and realize the diagnosis of high-speed railway running gear faults.
[0022] Among them, the confidence rule is the basic unit in the model, which describes the relationship between the premise attributes (features) and the fault conclusions (labels). Each rule contains a set of premise conditions and corresponding fault conclusions and confidence levels. These confidence rules are usually expressed in the form of "if...then...". The "if" part describes the premise attributes of the fault, and the "then" part describes the diagnostic conclusion of the fault.
[0023] The premise attributes are data used to characterize the operating status of the high-speed railway running gear, potential causes of faults, or any observable variables related to the occurrence of faults. They can be physical quantities (such as temperature, pressure, vibration frequency, etc.), statistics (such as mean, variance, kurtosis, etc.), transformed data (such as spectrum data after Fourier transformation), or any useful information extracted through data processing and feature engineering.
[0024] The fault conclusion refers to the current health status or fault degree of the high-speed railway running gear. This level can be discrete, such as "normal", "minor fault", "serious fault", etc., or continuous, such as a health index from 0 to 1, where 0 represents complete failure and 1 represents complete normality. It is usually determined based on domain knowledge, expert judgment or statistical analysis of historical data.
[0025] Furthermore, in this embodiment, the extracting of the characteristic quantity of the running sample data and constructing a fault diagnosis model based on a multi-dimensional confidence rule base, S1 includes: S11: preprocessing the running sample data; Preprocessing of sample data includes: identifying and processing outliers in the data, removing noise, filling missing values, standardizing or normalizing the data, etc.
[0026] By preprocessing the running sample data, the data quality can be ensured and the convergence speed and performance of the model can be improved.
[0027] S12: Extract features from the preprocessed running sample data to extract features used for fault diagnosis n The following fault diagnosis model is constructed based on the premise attributes and health status level: ;in, Indicates confidence rules; Indicates the first Prerequisite attributes; Indicates Rule No. Prerequisite attribute reference values; Indicates the level of health status; Indicates Results Confidence level; Indicates an “and” relationship; Indicates the number of evaluation results; Indicates The rule weight of the rule; Indicates The weight of the premise attributes; It should be noted that when When it was established, In contrast, when When it was established, rules are considered complete.
[0028] S13: Determine the rule weights in the fault diagnosis model , Prerequisite attribute weight .
[0029] Rule weights and premise attribute weights are core concepts in the Belief Rule Base (BRB) system. Together, they determine the accuracy and credibility of the model when making decisions or predictions.
[0030] Among them, the rule weight is used to quantify the importance or credibility of each rule in the model relative to other rules. In a confidence rule model, there are usually multiple rules, each of which gives a specific conclusion or prediction based on a different combination of premise attributes. Since different rules may be based on different experiences, knowledge or assumptions, their contribution to the overall decision or prediction may vary. The rule with a higher rule weight will have a greater influence when the model calculates the final result.
[0031] Premise Attribute Weight refers to the importance or influence of each premise attribute (or input variable, feature) in a rule. These attributes are the premise conditions of the rule and are used to determine whether the rule is applicable and the confidence of the rule. In a rule, different premise attributes may have different effects on the result of the rule. By assigning different weights to these attributes, the behavior of the rule can be more finely controlled to make it more in line with the actual situation.
[0032] By reasonably setting and adjusting the rule weights and premise attribute weights, a more accurate and reliable confidence rule model can be constructed to provide strong support for various decisions. Generally, these weights can be set or optimized based on expert knowledge, historical data or machine learning algorithms to improve the accuracy and robustness of the model.
[0033] This embodiment extracts key prerequisite attributes for fault diagnosis and corresponding health status level data from the data of the high-speed railway running gear under different operating conditions, including normal state and fault state, and constructs a fault diagnosis model based on a multidimensional confidence rule base. The model is used to perform fault diagnosis, comprehensively considering various fault causes, and making the diagnosis more accurate.
[0034] S2: using a fuzzy C-means clustering algorithm to determine the reference value of the premise attribute in the fault diagnosis model; The premise attribute reference value is a benchmark value used to compare with the premise attribute value in the data to be diagnosed. Each rule in the model contains a set of premise attributes and corresponding reference values. When evaluating whether the rule is activated, the premise attribute value in the high-speed railway running gear operation data to be diagnosed will be compared with the reference value defined in the rule to determine the degree of match or similarity. These reference values are usually determined based on domain knowledge and expert experience. Experts can directly set reference values for each premise attribute based on their in-depth understanding and experience of the system. However, this method relies on the subjective judgment of experts and is limited by the level of experts. In order to reduce the subjectivity of expert experience and improve the accuracy of the model, the fuzzy C-means clustering algorithm is used in this embodiment to determine the premise attribute reference value in the model.
[0035] Among them, the fuzzy C-means clustering algorithm (FCM) is a partition-based fuzzy clustering algorithm, which is mainly used for data clustering problems in unsupervised learning. The algorithm assigns fuzzy membership to each data point, divides the data points into different cluster centers, and finds the best clustering of data points by optimizing the objective function, thereby realizing clustering analysis of the data.
[0036] Specifically, in this embodiment, the fuzzy C-means clustering algorithm is used to determine the premise attribute reference value in the fault diagnosis model, and S2 includes: S21: Initialize and set c initial reference values of premise attributes as initial cluster centers; This step is the initialization step in the fuzzy C-means clustering algorithm. First, determine the number of clusters c, randomly select c cluster centers through initialization, and select a suitable fuzzy index (m, usually greater than 1).
[0037] S22: Repeat the following steps to iteratively update the initial cluster center until the change of the cluster center is less than a preset threshold or reaches a preset number of iterations: Calculating the distance from the actual value of the premise attribute in the running sample data to the cluster center; According to the distance between the actual value of the premise attribute and the cluster center, the membership function is used to calculate the membership of the actual value relative to the cluster center. The expression is as follows: ;in, Indicates Data points and The membership degree of the cluster center, that is, The data point belongs to The degree of cluster centers; is the index of the data point, indicating the data points; is the index of the cluster center, indicating the Cluster centers; is the count index of the cluster center, representing the cluster center from 1 to c; Indicates Data points and The distance measure between cluster centers; Indicates The cluster centers and The distance measure between cluster centers; It is a fuzzy index or a weighted index; is the number of cluster centers; The cluster center is updated by weighted average algorithm according to the membership degree, and the expression is as follows: ;in, is the new weighted cluster center; Indicates The coordinates or attribute values of the data points; Represents the total number of all data points considered when calculating the cluster center; In the Fuzzy C-means Clustering Algorithm (FCM), It represents a distance metric, which refers to the distance between two points (such as rules or premise attributes) in a certain space. For example, it can represent the distance between the premise attribute and a reference value, which is used to measure the matching degree between the premise attribute and the rule. is a fuzzy index or weighted index that controls the fuzziness of clustering. = 1, the clustering algorithm is hard clustering, increasing It can increase the fuzziness of the algorithm. Generally speaking, Greater than 1, usually between 1.5 and 2. c usually refers to the cluster center, which represents the center point of each cluster calculated in the fuzzy clustering algorithm. In the confidence rule model, c may represent the feature reference value, which is used to define the expected value or reference value of a premise attribute.
[0038] The degree of match between a premise attribute (such as motor temperature, speed, etc.) and the expected value in the rule can be measured by membership calculation. The higher the membership, the more the premise attribute conforms to the current rule, which helps to activate the most appropriate rule in the next reasoning step and obtain reliable diagnostic results.
[0039] This step is an iterative step of the fuzzy C-means clustering algorithm. By calculating the distance from each data point to all cluster centers and calculating the membership, the membership of the data point to each cluster center is inversely proportional to its distance to the center. The power is used as the weight to update the cluster center through the weighted average algorithm, so that each updated cluster center is the weighted average of its corresponding data points, and check whether the change of the new cluster center is less than a certain threshold or reaches the maximum number of iterations. If the conditions are met, the algorithm converges and outputs the clustering result; otherwise, continue to iterate with the new cluster center.
[0040] S23: Using the cluster center after the iteration as a reference value of the premise attribute in the fault diagnosis model.
[0041] This embodiment uses the fuzzy C-means clustering algorithm through iterative calculation to automatically classify the data set without human intervention. This high degree of automation makes the algorithm more efficient and convenient when processing large-scale data sets.
[0042] By adopting the fuzzy C-means clustering algorithm to determine the reference values of the premise attributes in the fault diagnosis model, the subjectivity of expert experience is reduced and the accuracy of the model is improved.
[0043] S3: Based on the premise attribute reference value, the confidence rules in the fault diagnosis model are integrated and reasoned by using evidential reasoning; Evidence Reasoning (ER) is a reasoning method for dealing with uncertain information. It describes the uncertainty of propositions from multiple perspectives by introducing concepts such as trust function, similarity function, and quasi-probability function. In this embodiment, evidence reasoning is used to integrate the confidence rules in the fault diagnosis model, and multiple uncertain and possibly conflicting confidence rules are combined to form a more reliable and accurate fault diagnosis conclusion. Specifically, it includes: S31: Calculate the matching degree of each premise attribute of each confidence rule in the fault diagnosis model according to the premise attribute reference value and the premise attribute actual value in the running sample data, and the expression is as follows: ;in, Indicates Rule No. The matching degree of the premise attributes; and Represents two adjacent rules , The Prerequisite attribute reference values; Indicates the first The actual input value of the premise attribute; Indicates the rule number currently being processed; Indicates the number of premise attributes; by comparing the actual input value with the rule reference value and The relationship between the premise attributes and the rules is used to determine the matching degree between the current premise attributes and the rules, which is then used in the fusion reasoning process based on evidence reasoning.
[0044] S32: Calculate the matching degree of each rule according to the matching degree of each premise attribute. The expression is as follows: ;in, Indicates The matching degree of the rules; Indicates The weight of the premise attributes; S33: Calculate the activation weight of each rule according to the matching degree of each rule. The expression is as follows: ;in, Indicates The activation weight of the rule; Indicates The rule weight of the rule; Indicates The rule weight of the rule; Indicates The matching degree of the rules is used to sum up and represent the matching degree of multiple rules; Represents the number of rules in the fault diagnosis model; S34: Based on evidence reasoning, the confidence rules of the fault diagnosis model are inferred by using the following formula: ; ;in, Represents the fault diagnosis model output relative to the evaluation result The confidence level indicates the degree of trust or certainty of the model in its output results, which is determined by the rule weight , Prerequisite attributes The attribute weight and the confidence corresponding to each premise attribute The composed function.
[0045] Hypothetical evaluation results The utility is ,but The expected utility is: .
[0046] In this example, the confidence rules in the fault diagnosis model are integrated and reasoned based on evidence reasoning. By comprehensively utilizing qualitative and quantitative information, an accurate diagnosis model is established, and the diagnosis decision process is optimized, which can further improve the accuracy and stability of the diagnosis model.
[0047] S4: Using the adaptive evolution strategy optimization algorithm of the projection covariance matrix, the parameters in the fault diagnosis model after fusion reasoning are updated; Evolutionary Strategy with CovarianceMatrix Adaptation (CMA-ES) is an evolutionary algorithm that searches for the optimal solution by iteratively updating the mean and covariance matrix of a multivariate normal distribution. It is suitable for solving continuous optimization problems and can search effectively in high-dimensional space. Specifically, it includes: S41: Constructing the objective function : ;in, Indicates the termination time of the iteration process; Indicates the current iteration step number; Indicates the starting time point of the iteration process; S42: the objective function The constraints make and As close as possible, the expression is as follows: ;in, Indicates The actual output parameters or weights of the rules, and It can be understood as The target output parameter of the rule is set by adding constraints to make the two values as close as possible.
[0048] The adaptive evolution strategy optimization algorithm of the projection covariance matrix is mainly an evolutionary algorithm for continuous optimization problems, and is particularly suitable for global optimization in high-dimensional space. In the fault diagnosis model, it is used to adjust the model parameters after fusion reasoning so that the model can more accurately reflect the actual situation. In this embodiment, although S41 and S42 do not explicitly display complex matrix operations, the core idea of the algorithm is still reflected in the optimization step, and the model parameters are updated through the evolution strategy to enhance the diagnostic effect.
[0049] From a process perspective, the role of S41 and S42 is to construct an objective function and optimize parameters to make the model output closer to the ideal state. The key here is to iteratively optimize the weights in the model (such as rule weights, premise attribute weights, etc.), and these weight optimizations are actually parameter optimization using the adaptive evolution strategy algorithm. These optimization steps are the practical application of the algorithm.
[0050] S5: Collect the operating data of the high-speed railway running gear to be diagnosed, input it into the updated fault diagnosis model, and output the fault diagnosis result of the high-speed railway running gear to be diagnosed.
[0051] It can be understood that in a specific implementation, the collected operating data of the high-speed railway running unit to be diagnosed also includes preprocessing of the data, such as identifying and processing outliers, removing noise, filling missing values, standardization or normalization, etc.
[0052] In this embodiment, the collected operating data of the high-speed railway running gear to be diagnosed is input into the optimized diagnostic model. The model calculates the confidence level of the running gear in different health status levels by evaluating the matching degree between the premise attributes in the operating data and the reference values in the rules, and combines the weights of the rules and the premise attributes, and finally gives a diagnostic result.
[0053] This embodiment constructs a high-speed railway running gear fault diagnosis model based on a multi-dimensional confidence rule base, establishes multiple rules that characterize the mapping relationship between premise attributes and fault conclusions, and can comprehensively consider various and interrelated fault causes; by adopting a fuzzy C-means clustering algorithm to determine the reference value of the premise attributes in the diagnosis model, the subjectivity of expert knowledge is reduced and the accuracy of model diagnosis is improved; by fusing the confidence rules in the model based on evidence reasoning, and using the adaptive evolution strategy optimization algorithm of the projection covariance matrix, the parameters in the fault diagnosis model after the fusion reasoning are updated, and the parameters of the fault diagnosis model are optimized by the algorithm, so that the model diagnosis result is more accurate.
[0054] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.
[0055] Example 2: In this example, a high-speed railway running gear fault diagnosis device is also provided to implement the above-mentioned examples and preferred implementation modes, which have been described and will not be repeated here. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following examples are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.
[0056] Figure 2 is a structural block diagram of a high-speed railway running gear fault diagnosis device according to an embodiment of the present invention. Figure 2As shown, the device includes: a data acquisition module 100, a data extraction module 200, a model building module 300, a clustering algorithm module 400, a fusion reasoning module 500, an evolution strategy module 600, and a fault diagnosis module 700, wherein: The data collection module 100 is used to collect sample data for building the model and the operation data of the high-speed railway running gear to be diagnosed; The data extraction module 200 is used to extract the characteristic quantities of the sample data and the operation data; A model building module 300 is used to build a fault diagnosis model based on a multi-dimensional confidence rule base, wherein the fault diagnosis model includes a plurality of confidence rules defining a mapping relationship between premise attributes and fault conclusions; A clustering algorithm module 400, for determining the premise attribute reference value in the fault diagnosis model by using a fuzzy C-means clustering algorithm; A fusion reasoning module 500, configured to perform fusion reasoning on the confidence rules in the fault diagnosis model based on the premise attribute reference value by using evidence reasoning; The evolution strategy module 600 is used to update the parameters in the fault diagnosis model after fusion reasoning by using the adaptive evolution strategy optimization algorithm of the projection covariance matrix; The fault diagnosis module 700 is used to input the operating data of the high-speed railway running unit to be diagnosed into the updated fault diagnosis model, and output the fault diagnosis result of the high-speed railway running unit to be diagnosed.
[0057] Further, the data extraction module includes: a preprocessing unit for preprocessing the running sample data; an extraction unit for extracting features from the preprocessed running sample data to extract n premise attributes and health status levels for fault diagnosis; the model construction module includes: a construction unit for constructing the following fault diagnosis model: ;in, Indicates confidence rules; Indicates the first Prerequisite attributes; Indicates Rule No. Prerequisite attribute reference values; Indicates the level of health status; Indicates Results confidence level; Indicates an “and” relationship; Indicates the number of evaluation results; Indicates The rule weight of the rule; Indicates The weight of the premise attributes; A weight assignment unit, used to determine the rule weights in the fault diagnosis model , Prerequisite attribute weight .
[0058] Furthermore, the clustering algorithm module includes: an initialization unit, used to initialize and set c initial reference values of premise attributes as initial cluster centers; an iteration unit, used to repeat the following steps to iteratively update the initial cluster centers until the change of the cluster centers is less than a preset threshold or reaches a preset number of iterations: calculating the distance from the actual value of the premise attribute in the running sample data to the cluster center; using a membership function to calculate the membership of the actual value relative to the cluster center according to the distance between the actual value of the premise attribute and the cluster center, the expression is as follows: ;in, Indicates Data points and The membership degree of the cluster center, that is, The data point belongs to The degree of cluster centers; is the index of the data point, indicating the data points; is the index of the cluster center, indicating the Cluster centers; is the count index of the cluster center, representing the cluster center from 1 to c; Indicates Data points and The distance measure between cluster centers; Indicates The cluster centers and The distance measure between cluster centers; It is a fuzzy index or a weighted index; is the number of cluster centers; The cluster center is updated by weighted average algorithm according to the membership degree, and the expression is as follows: ;in, is the new weighted cluster center; Indicates The coordinates or attribute values of the data points; Represents the total number of all data points considered when calculating the cluster center; The premise attribute reference value assignment unit is used to use the cluster center after the iteration as the premise attribute reference value in the fault diagnosis model.
[0059] Furthermore, the fusion reasoning module includes: a first calculation unit, which is used to calculate the matching degree of each premise attribute of each confidence rule in the fault diagnosis model according to the premise attribute reference value and the premise attribute actual value in the running sample data, and the expression is as follows: ;in, Indicates Rule No. The matching degree of the premise attributes; and Represents two adjacent rules , The Prerequisite attribute reference values; Indicates the first The actual input value of the premise attribute; Indicates the rule number currently being processed; Indicates the number of premise attributes; The second calculation unit is used to calculate the matching degree of each rule according to the matching degree of each premise attribute. The expression is as follows: ;in, Indicates The matching degree of the rules; Indicates The weight of the premise attributes; The third calculation unit is used to calculate the activation weight of each rule according to the matching degree of each rule. The expression is as follows: ;in, Indicates The activation weight of the rule; Indicates The rule weight of the rule; Indicates The rule weight of the rule; Indicates The matching degree of the rules is used to sum up and represent the matching degree of multiple rules; Represents the number of rules in the fault diagnosis model; The fourth computing unit is used to fuse and infer the confidence rules of the fault diagnosis model based on evidence reasoning using the following formula: ; ;in, Represents the fault diagnosis model output relative to the evaluation result The confidence level indicates the degree of trust or certainty of the model in its output results, which is determined by the rule weight , Prerequisite attributes The attribute weight and the confidence corresponding to each premise attribute The composed function.
[0060] Furthermore, the evolution strategy module includes: a construction unit for constructing an objective function : ;in, Indicates the termination time of the iteration process; Indicates the current iteration step number; Indicates the starting time point of the iteration process; Constraint unit, used to constrain the objective function The constraints make and As close as possible, the expression is as follows: ;in, Indicates The actual output parameter or weight of the rule.
[0061] It should be noted that the above modules can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.
[0062] Embodiment 3: An embodiment of the present invention further provides a storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.
[0063] Optionally, in this embodiment, the storage medium may be configured to store a computer program for performing the following steps: S1: obtaining operation sample data of a high-speed railway running unit, extracting feature quantities of the operation sample data, and constructing a fault diagnosis model based on a multidimensional confidence rule base, wherein the fault diagnosis model includes a plurality of confidence rules defining a mapping relationship between premise attributes and fault conclusions; S2: using a fuzzy C-means clustering algorithm to determine the reference value of the premise attribute in the fault diagnosis model; S3: Based on the premise attribute reference value, the confidence rules in the fault diagnosis model are integrated and reasoned by using evidential reasoning; S4: Using the adaptive evolution strategy optimization algorithm of the projection covariance matrix, the parameters in the fault diagnosis model after fusion reasoning are updated; S5: Collect the operating data of the high-speed railway running gear to be diagnosed, input it into the updated fault diagnosis model, and output the fault diagnosis result of the high-speed railway running gear to be diagnosed.
[0064] Optionally, in this embodiment, the above-mentioned storage medium may include but is not limited to: a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and other media that can store computer programs.
[0065] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.
[0066] Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments and optional implementation modes, and this embodiment will not be described in detail here.
[0067] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.
[0068] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.
[0069] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.
[0070] The units described as separate components may or may not be physically separated, and the components shown as units 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 units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0071] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or 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 software functional units.
[0072] If the integrated unit 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. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk, etc., which can store program code.
[0073] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.
Claims
1. A method for diagnosing faults of high-speed railway running gear. It is characterized in that include: Acquire operation sample data of a high-speed railway running unit, extract feature quantities of the operation sample data, and construct a fault diagnosis model based on a multidimensional confidence rule base, wherein the fault diagnosis model includes a plurality of confidence rules defining a mapping relationship between premise attributes and fault conclusions; Using fuzzy C-means clustering algorithm to determine the premise attribute reference value in the fault diagnosis model; Based on the premise attribute reference value, using evidential reasoning to perform fusion reasoning on the confidence rules in the fault diagnosis model; The adaptive evolution strategy optimization algorithm of the projection covariance matrix is used to update the parameters of the fault diagnosis model after fusion reasoning. The operating data of the high-speed railway running gear to be diagnosed is collected, input into the updated fault diagnosis model, and the fault diagnosis result of the high-speed railway running gear to be diagnosed is output.
2. The high-speed railway running gear fault diagnosis method according to claim 1, It is characterized in that The extracting of the characteristic quantity of the running sample data and constructing a fault diagnosis model based on a multi-dimensional confidence rule base includes: Preprocessing the running sample data; The preprocessed running sample data is subjected to feature extraction, from which n premise attributes and health status levels for fault diagnosis are extracted, and the following fault diagnosis model is constructed: ;in, Indicates confidence rules; Indicates the first Prerequisite attributes; Indicates Rule No. Prerequisite attribute reference values; Indicates the level of health status; Indicates Results Confidence level; Indicates an "and" relationship; Indicates the number of evaluation results; Indicates The rule weight of the rule; Indicates The weight of the premise attributes; Determine the rule weights in the fault diagnosis model , Prerequisite attribute weight .
3. The high-speed railway running gear fault diagnosis method according to claim 2, It is characterized in that The method of using the fuzzy C-means clustering algorithm to determine the premise attribute reference value in the fault diagnosis model includes: Initialize and set c initial reference values of premise attributes as initial cluster centers; Repeat the following steps to iteratively update the initial cluster center until the change of the cluster center is less than a preset threshold or reaches a preset number of iterations: Calculating the distance from the actual value of the premise attribute in the running sample data to the cluster center; According to the distance between the actual value of the premise attribute and the cluster center, the membership function is used to calculate the membership of the actual value relative to the cluster center. The expression is as follows: ;in, Indicates Data points and The membership degree of the cluster center, that is, The data point belongs to The degree of cluster centers; is the index of the data point, indicating the data points; is the index of the cluster center, indicating the Cluster centers; is the count index of the cluster center, representing the cluster center from 1 to c; Indicates Data points and The distance measure between cluster centers; Indicates The cluster centers and The distance measure between cluster centers; It is a fuzzy index or a weighted index; is the number of cluster centers; The cluster center is updated by weighted average algorithm according to the membership degree, and the expression is as follows: ;in, is the new weighted cluster center; Indicates The coordinates or attribute values of the data points; Represents the total number of all data points considered when calculating the cluster center; The cluster center after the iteration is completed is used as the reference value of the premise attribute in the fault diagnosis model.
4. The high-speed railway running gear fault diagnosis method according to claim 3, It is characterized in that The adopting evidential reasoning to perform fusion reasoning on the confidence rules in the fault diagnosis model based on the premise attribute reference value includes: According to the premise attribute reference value and the premise attribute actual value in the running sample data, the matching degree of each premise attribute of each confidence rule in the fault diagnosis model is calculated, and the expression is as follows: ;in, Indicates Rule No. The matching degree of the premise attributes; and Represents two adjacent rules , The Prerequisite attribute reference values; Indicates the first The actual input value of the premise attribute; Indicates the rule number currently being processed; Indicates the number of premise attributes; The matching degree of each rule is calculated based on the matching degree of each premise attribute. The expression is as follows: ;in, Indicates The matching degree of the rules; Indicates The weight of the premise attributes; The activation weight of each rule is calculated according to the matching degree of each rule. The expression is as follows: ;in, Indicates The activation weights of the rules; Indicates The rule weight of the rule; Indicates The rule weight of the rule; Indicates The matching degree of the rules is used to sum up and represent the matching degree of multiple rules; Represents the number of rules in the fault diagnosis model; Based on evidence reasoning, the confidence rules of the fault diagnosis model are inferred by the following formula: ; ;in, Represents the fault diagnosis model output relative to the evaluation result The confidence level is determined by the rule weight , Prerequisite attributes Attribute weights and confidence The composed function.
5. The high-speed railway running gear fault diagnosis method according to claim 4, It is characterized in that The adaptive evolution strategy optimization algorithm using the projection covariance matrix is used to update the parameters in the fault diagnosis model after fusion reasoning, including: Constructing the objective function : ;in, Indicates the termination time of the iteration process; Indicates the current iteration step number; Indicates the starting time point of the iteration process; For the objective function The constraints make and As close as possible, the expression is as follows: ;in, Indicates The actual output parameter or weight of the rule.
6. A high-speed railway running gear fault diagnosis device, It is characterized in that include: The data acquisition module is used to collect sample data for building the model and the operation data of the high-speed railway running gear to be diagnosed; A data extraction module is used to extract characteristic quantities of sample data and operation data; A model building module, used to build a fault diagnosis model based on a multi-dimensional confidence rule base, wherein the fault diagnosis model includes a plurality of confidence rules defining a mapping relationship between premise attributes and fault conclusions; A clustering algorithm module, used to determine the premise attribute reference value in the fault diagnosis model by using a fuzzy C-means clustering algorithm; A fusion reasoning module, used for performing fusion reasoning on the confidence rules in the fault diagnosis model based on the premise attribute reference value by using evidence reasoning; An evolution strategy module is used to update the parameters in the fault diagnosis model after fusion reasoning by using an adaptive evolution strategy optimization algorithm of the projection covariance matrix; The fault diagnosis module is used to input the operating data of the high-speed railway running gear to be diagnosed into the updated fault diagnosis model, and output the fault diagnosis result of the high-speed railway running gear to be diagnosed.
7. An electronic device, It is characterized in that The electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the computer program implements the steps of the method according to any one of claims 1 to 5 when executed by the processor.
8. A storage medium, It is characterized in that The storage medium stores a computer program, which implements the steps of the method according to any one of claims 1 to 5 when executed by a processor.
Citation Information
Patent Citations
High-speed train running gear system fault diagnosis method based on semi-quantitative information fusion
CN112990275A
Cited By
Fault diagnosis method and device based on knowledge extraction
CN119202762A
