Vehicle part fault prediction method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a vehicle part fault prediction method and device, electronic equipment and a storage medium, and belongs to the technical field of vehicle fault prediction. The method comprises the following steps: acquiring historical part operation data; performing fault identification on the historical part operation data to obtain historical fault part information; performing associated information mining on the historical fault part information to obtain historical fault part associated information; performing fault information clustering on the historical fault part information to obtain historical fault part clustering information; performing parameter optimization on a preset part fault prediction model according to the historical fault part information, the historical fault part association information and the historical fault part clustering information to obtain a target part fault prediction model; and obtaining current fault part information, and performing fault prediction on the current fault part information through the target part fault prediction model. According to the embodiment of the invention, the prediction accuracy of the vehicle part fault can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of vehicle fault prediction, and in particular to a vehicle parts fault prediction method and device, electronic equipment and storage medium. Background Art

[0002] Since vehicle parts are usually interrelated and dependent on each other, when a part fails, it will affect the normal operation of the parts associated with it, causing the associated parts to fail. Therefore, how to improve the accuracy of predicting vehicle parts failures has become an urgent problem to be solved. Summary of the invention

[0003] The main purpose of the embodiments of the present application is to propose a vehicle parts failure prediction method and device, electronic equipment and storage medium, aiming to improve the prediction accuracy of vehicle parts failure.

[0004] To achieve the above-mentioned purpose, a first aspect of an embodiment of the present application proposes a vehicle parts failure prediction method, the method comprising:

[0005] Obtain the historical operation data of the preset vehicle and obtain the historical parts operation data;

[0006] Performing fault identification on the historical parts operation data to obtain historical fault parts information;

[0007] Performing associated information mining on the historical faulty parts information to obtain historical faulty parts associated information;

[0008] Performing fault information clustering on the historical fault parts information to obtain historical fault parts clustering information;

[0009] Optimizing parameters of a preset part failure prediction model according to the historical fault part information, the historical fault part association information, and the historical fault part clustering information to obtain a target part failure prediction model;

[0010] Fault information of a target vehicle is acquired to obtain current faulty parts information, and fault prediction is performed on the current faulty parts information using the target parts fault prediction model.

[0011] In some embodiments, the optimizing the parameters of the preset part failure prediction model according to the historical fault part information, the historical fault part association information and the historical fault part clustering information to obtain the target part failure prediction model includes:

[0012] Performing feature fusion on the historical faulty parts association information and the historical faulty parts clustering information through a preset feature fusion model to obtain faulty parts feature information;

[0013] The parameters of the preset part failure prediction model are optimized according to the faulty part feature information and the historical faulty part information to obtain the target part failure prediction model.

[0014] In some embodiments, the step of performing feature fusion on the historical faulty parts association information and the historical faulty parts clustering information through a preset feature fusion model to obtain the faulty parts feature information includes:

[0015] Convolving the historical faulty parts associated information through a preset convolution layer to obtain associated faulty parts feature information;

[0016] Convolving the historical faulty parts clustering information through the preset convolution layer to obtain convolved faulty parts feature information;

[0017] Performing feature splicing on the associated faulty part feature information and the convolved faulty part feature information to obtain aggregated faulty part feature information;

[0018] The aggregated faulty part feature information is subjected to feature generation through a preset fully connected layer to obtain the faulty part feature information.

[0019] In some embodiments, the target part fault prediction model includes a long-short memory sub-model, a self-attention sub-model, and a regression sub-model; and performing fault prediction on the current fault part information by using the target part fault prediction model includes:

[0020] Generate the time series feature of the current fault part information by using the long-short memory sub-model to obtain the time series feature of the current vehicle part;

[0021] Performing feature encoding on the time series features of the current vehicle parts through the self-attention sub-model to obtain the current vehicle part coding features;

[0022] The regression sub-model is used to perform regression prediction on the current vehicle part code feature.

[0023] In some embodiments, the performing associated information mining on the historical faulty parts information to obtain the historical faulty parts associated information includes:

[0024] Performing frequency statistics on the historical faulty parts information to obtain parts failure frequency data;

[0025] The part failure frequency data is screened according to a preset frequent item set threshold to obtain a frequent failure part item set;

[0026] Generate association rules based on the frequent faulty parts item set to obtain faulty parts association rules;

[0027] The faulty parts association rules are screened to obtain the historical faulty parts association information.

[0028] In some embodiments, the historical faulty parts information includes a plurality of fault record information; and clustering the historical faulty parts information to obtain the historical faulty parts clustering information includes:

[0029] Calculate the Euclidean distance between any two fault record information to obtain the distance matrix;

[0030] Classifying the fault record information according to the distance matrix to obtain fault classification information;

[0031] The fault classification information is screened by silhouette coefficient to obtain the historical fault parts clustering information.

[0032] In some embodiments, the performing fault identification on the historical part operation data to obtain historical fault part information includes:

[0033] Performing mean calculation on the historical parts operation data to obtain operation mean data;

[0034] Calculating the standard deviation of the historical parts operation data to obtain operation standard deviation data;

[0035] Performing aggregation calculation according to the operation mean data, the operation standard deviation data and the historical part operation data to obtain an operation data offset;

[0036] The historical faulty part information is filtered out from the historical part operation data according to the operation data offset.

[0037] To achieve the above-mentioned purpose, a second aspect of an embodiment of the present application provides a vehicle parts failure prediction device, the device comprising:

[0038] A data acquisition module is used to acquire historical operation data of a preset vehicle and obtain historical parts operation data;

[0039] A fault identification module is used to identify faults on the historical parts operation data to obtain historical fault parts information;

[0040] An information mining module is used to mine related information on the historical faulty parts information to obtain related information on the historical faulty parts;

[0041] An information clustering module, used for clustering the historical faulty parts information to obtain historical faulty parts clustering information;

[0042] A parameter optimization module, used to optimize the parameters of a preset part failure prediction model according to the historical fault part information, the historical fault part association information and the historical fault part clustering information, so as to obtain a target part failure prediction model;

[0043] The information acquisition module is used to acquire the fault information of the target vehicle, obtain the current fault part information, and perform fault prediction on the current fault part information through the target part fault prediction model.

[0044] To achieve the above-mentioned purpose, the third aspect of an embodiment of the present application proposes an electronic device, which includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method described in the first aspect is implemented.

[0045] To achieve the above-mentioned purpose, the fourth aspect of an embodiment of the present application proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in the first aspect is implemented.

[0046] The present application proposes a vehicle part fault prediction method and device, electronic device and storage medium, which obtain historical part operation data, then perform fault identification on the historical part operation data to obtain historical fault part information, then perform correlation information mining on the historical fault part information to obtain historical fault part correlation information, and perform fault information clustering on the historical part fault information to obtain historical fault part clustering information, and then optimize the parameters of a preset part fault prediction model through the historical fault part clustering information and the historical fault part correlation information, so that the preset part fault prediction model can perceive the fault characteristics of the part from two dimensions of clustering information and correlation information, and enable the preset part fault prediction model to have good prediction ability for part failure, finally obtain current fault part information, and perform fault prediction on the current fault part information through the target part fault prediction model, thereby improving the prediction accuracy of vehicle part failure. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 is a flow chart of a vehicle parts failure prediction method provided by an embodiment of the present application;

[0048] Figure 2 yes Figure 1 Flow chart of step S102 in FIG.

[0049] Figure 3 yes Figure 1 Flow chart of step S103 in FIG.

[0050] Figure 4 yes Figure 1Flow chart of step S104 in FIG.

[0051] Figure 5 yes Figure 1 Flow chart of step S105 in FIG.

[0052] Figure 6 yes Figure 5 Flow chart of step S501 in FIG.

[0053] Figure 7 yes Figure 1 Flow chart of step S106 in FIG.

[0054] Figure 8 is a schematic diagram of the structure of a vehicle parts failure prediction device provided in an embodiment of the present application;

[0055] Fig. 9 It is a schematic diagram of the hardware structure of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0056] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0057] It should be noted that, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0058] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0059] First, some nouns involved in this application are analyzed:

[0060] Clustering: is a technique for grouping objects in a data set according to their similarities, with the goal of making objects in the same group as similar as possible, and objects in different groups as different as possible. Clustering algorithms divide objects into several different groups by analyzing the features or attributes between objects, and the objects in each group are relatively close in features.

[0061] Temporal Features: refers to the time-related features in the data, which are used to describe the laws and trends of data changes over time. Temporal features include time points, time intervals, trend components, periodic components, seasonal components, etc., which are used to analyze and model the dynamic behavior of time series data and reveal long-term trends, periodic fluctuations and short-term changes in the data.

[0062] Silhouette Coefficient: It is an indicator used to evaluate the effect of data clustering, which measures the compactness of samples in their cluster and the separation from other clusters. The silhouette coefficient is based on the average distance between the sample and other samples in the cluster and the average distance between the sample and the sample in the nearest cluster. The silhouette coefficient is used to determine whether the sample is reasonably assigned to its cluster and the separation effect between clusters. The value of the silhouette coefficient reflects the quality of clustering. A high value means that the sample is relatively compact in its cluster and has a good separation effect with samples in other clusters.

[0063] Since vehicle parts are usually interrelated and dependent on each other, when a part fails, it will affect the normal operation of the parts associated with it, causing the associated parts to fail. Therefore, how to improve the accuracy of predicting vehicle parts failures has become an urgent problem to be solved.

[0064] Based on this, the embodiments of the present application provide a vehicle parts failure prediction method and device, an electronic device and a storage medium, aiming to improve the prediction accuracy of vehicle parts failure.

[0065] A vehicle parts failure prediction method and device, electronic device and storage medium provided in the embodiments of the present application are specifically described through the following embodiments. First, the vehicle parts failure prediction method in the embodiments of the present application is described.

[0066] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. 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.

[0067] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0068] The vehicle parts fault prediction method provided in the embodiment of the present application relates to the field of vehicle fault prediction technology. The vehicle parts fault prediction method provided in the embodiment of the present application can be applied to a terminal, can also be applied to a server side, and can also be software running in a terminal or a server side. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc.; the server side can be configured as an independent physical server, or a server cluster or a distributed system composed of multiple physical servers, and can also be configured as a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application that implements the vehicle parts fault prediction method, etc., but is not limited to the above forms.

[0069] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0070] Figure 1 is an optional flowchart of the vehicle parts failure prediction method provided in the embodiment of the present application, Figure 1 The method may include but is not limited to steps S101 to S106.

[0071] Step S101, obtaining historical operation data of a preset vehicle and obtaining historical parts operation data;

[0072] Step S102, performing fault identification on historical parts operation data to obtain historical fault parts information;

[0073] Step S103, performing correlation information mining on historical faulty parts information to obtain historical faulty parts correlation information;

[0074] Step S104, clustering the historical faulty parts information to obtain historical faulty parts clustering information;

[0075] Step S105, optimizing parameters of a preset part failure prediction model according to historical fault part information, historical fault part association information, and historical fault part clustering information to obtain a target part failure prediction model;

[0076] Step S106, acquiring the fault information of the target vehicle, obtaining the current faulty parts information, and performing fault prediction on the current faulty parts information through the target parts fault prediction model.

[0077] Steps S101 to S106 shown in the embodiment of the present application are obtained by acquiring historical part operation data, then performing fault identification on the historical part operation data to obtain historical fault part information, then performing associated information mining on the historical fault part information to obtain historical fault part associated information, and performing fault information clustering on the historical part fault information to obtain historical fault part clustering information, and then performing parameter optimization on a preset part fault prediction model through the historical fault part clustering information and the historical fault part associated information, so that the preset part fault prediction model can perceive the fault characteristics of the part from two dimensions of clustering information and associated information, and enable the preset part fault prediction model to have good prediction capabilities for part failures, and finally obtain current fault part information, and perform fault prediction on the current fault part information through the target part fault prediction model, thereby improving the prediction accuracy of vehicle part failures.

[0078] In step S101 of some embodiments, obtaining the historical operation data of the preset vehicle is to obtain the operation record data of all vehicles during the operation process, including vehicle speed, engine temperature, fuel consumption, parts working condition, parts maintenance records and other data, and extracting the operation data of the parts from the historical operation data, that is, the historical parts operation data, including usage time, working conditions and fault records, for example: "["2024-01-15","Engine","Normal","None"],["2024-02-20","Brake system","Severe wear","None"],["2024-03-10","Transmission system","Normal","Clutch slip"],["2024-04-05","Air filter","Frequent abnormal noise","None"]".

[0079] See also Figure 2 In some embodiments, step S102 may include but is not limited to steps S201 to S204:

[0080] Step S201, performing mean calculation on historical part operation data to obtain operation mean data;

[0081] Step S202, calculating the standard deviation of historical part operation data to obtain operation standard deviation data;

[0082] Step S203, performing aggregation calculation according to the operation mean data, the operation standard deviation data and the historical part operation data to obtain the operation data offset;

[0083] Step S204, filtering out historical faulty parts information from historical parts operation data according to the operation data offset.

[0084] Steps S201 to S204 shown in the embodiment of the present application obtain the operating mean data and the operating standard deviation data of the historical part operating data, and perform aggregation calculations based on the operating mean data, the operating standard deviation data and the historical part operating data to obtain the operating data offset. Finally, the historical faulty part information is filtered out from the historical part operating data based on the operating data offset, so as to accurately identify the historical faulty part information from the historical part operating data.

[0085] In step S201 of some embodiments, the historical part operation data includes multiple records of multiple parts, for example, ["2024-01-15","Engine","Normal","None"], ["2024-02-20","Engine","Severe wear","None"], ["2024-03-10","Engine","Normal","Clutch slip"], ["2024-04-05","Engine","Frequent abnormal noise","None"], ["2024-01-15","Brake system","Normal","None"], ["2024-02-20","Brake system","Severe wear","None"], ["2024-03-10","Brake system","Normal","None"] and ["2024-04-05","Brake system","Normal","None"].

[0086] Then, the historical parts operation data is converted into indicators, and ["2024-01-15","Engine",1,0],["2024-02-20","Engine",3,0],["2024-03-10","Engine",1,1],["2024-04-05","Engine",4,0] are obtained. Then, the mean of the engine is calculated, and the mean of the working condition rating is 2.25, and the mean of the fault record rating is 0.25, that is, ["Engine",2.25,0.25]. Finally, the mean of the working condition and the mean of the fault record rating of each part are calculated in turn.

[0087] In step S202 of some embodiments, the standard deviation of the data after the index conversion of the historical part operation data is calculated, for example, ["2024-01-15", "engine", 1, 0], ["2024-02-20", "engine", 3, 0], ["2024-03-10", "engine", 1, 1], ["2024-04-05", "engine", 4, 0]. The standard deviation of the working condition and fault records is calculated to obtain ["engine", 1.46, 0.43].

[0088] In step S203 of some embodiments, the aggregation calculation is as shown in formula (1):

[0089]

[0090] Among them, z is the operating data offset, x is the historical parts operating data, μ is the operating mean data, and σ is the operating standard deviation data.

[0091] In step S204 of some embodiments, a comparison is made between a preset offset range threshold and the operating data offset, and data whose operating data offset is not within the offset range threshold is screened out as historical fault part information.

[0092] See also Figure 3 In some embodiments, step S103 may include but is not limited to steps S301 to S304:

[0093] Step S301, performing frequency statistics on historical faulty parts information to obtain part failure frequency data;

[0094] Step S302, filtering the part failure frequency data according to a preset frequent item set threshold to obtain a frequent failure part item set;

[0095] Step S303, generating association rules based on the frequent item set of faulty parts to obtain association rules for faulty parts;

[0096] Step S304, screening the faulty parts association rules to obtain historical faulty parts association information.

[0097] In steps S301 to S304 shown in the embodiment of the present application, frequency statistics are performed on historical faulty parts information to obtain part failure frequency data, and then the part failure frequency data is filtered according to a preset frequent item set threshold to obtain a frequent faulty parts item set, and association rules are generated based on the frequent faulty parts item set to obtain faulty parts association rules, and finally the faulty parts association rules are filtered to obtain historical faulty parts association information, thereby mining the association relationship between part failures on the vehicle from the historical faulty parts information.

[0098] In step S301 of some embodiments, the historical fault parts information includes parts that have been repaired in one maintenance, such as ("engine, brake system, steering system"), ("engine, brake system"), ("brake system, transmission system"), ("engine, transmission system, fuel system"), ("brake system"), ("transmission system"), ("engine"), ("brake system, steering system, transmission system"), ("engine, brake system"), and ("fuel system").

[0099] Then the historical faulty parts information is used to generate frequent two-item sets, and the number of times is counted to obtain the part failure frequency data, for example, {engine, brake system}: 3 times, {engine, transmission system}: 1 time, {brake system, transmission system}: 2 times, {brake system, steering system}: 2 times, {transmission system, steering system}: 1 time, {engine, fuel system}: 1 time, {brake system, fuel system}: 1 time, {engine, steering system}: 1 time, {fuel system, steering system}: 0 times.

[0100] In step S302 of some embodiments, the part failure frequency data is screened according to a preset frequent item set threshold, and data greater than the preset frequent item set threshold is screened out to obtain multiple frequent failure part item sets. In one embodiment, the preset frequent item set threshold is 2, and the part failure frequency data is screened according to the preset frequent item set threshold to obtain {engine, brake system}: 3 times, {brake system, transmission system}: 2 times, and {brake system, steering system}: 2 times.

[0101] In step S303 of some embodiments, association rules are generated based on frequent faulty parts item sets to obtain faulty parts association rules. For example, based on {engine, brake system}: 3 times, {brake system, transmission system}: 2 times and {brake system, steering system}: 2 times, faulty parts association rules (engine, brake system), (brake system, engine), (brake system, transmission system), (transmission system, brake system), (brake system, steering system), (steering system, brake system) are generated.

[0102] In step S304 of some embodiments, the confidence is calculated based on the faulty parts association rule. For example, based on the faulty parts association rule (engine, brake system): 3 times, the ratio of (engine, brake system): 3 times and (engine): 5 times is obtained as 0.6, and then 0.6 is compared with the preset confidence threshold to determine whether there is an association between the damage of (engine, brake system). For example, if the preset confidence threshold is 0.5, then 0.6 is greater than 0.5, and it is determined that when the engine is damaged, the brake system has a greater probability of being damaged. It should be noted that (engine): 5 times is the number of times the engine has been damaged in the historical faulty parts information.

[0103] The confidence level of each faulty parts association rule is calculated and compared with the preset confidence threshold to determine the historical faulty parts association information.

[0104] It should be noted that since vehicle failures often do not occur in isolation, multiple parts may fail at the same time. By mining the associated information of historical faulty parts, the associated information of historical faulty parts can be obtained, and based on the historical faulty parts information, the prediction ability of the preset part failure prediction model for faulty parts can be improved.

[0105] See also Figure 4 In some embodiments, the historical fault part information includes a plurality of fault record information, and step S104 may include but is not limited to steps S401 to S403:

[0106] Step S401, calculating the Euclidean distance between any two fault record information to obtain a distance matrix;

[0107] Step S402, classifying the fault record information according to the distance matrix to obtain fault classification information;

[0108] Step S403, screening the fault classification information by silhouette coefficient to obtain historical fault parts clustering information.

[0109] In steps S401 to S403 shown in the embodiment of the present application, the Euclidean distance between any two fault record information is calculated to obtain a distance matrix, and then the fault record information is classified according to the distance matrix to obtain fault classification information. Finally, the fault classification information is screened by contour coefficient to obtain historical fault part clustering information, thereby completing the clustering of the historical fault part information and providing clustering features for the subsequent process to improve the fault prediction capability of the target part fault prediction model for the parts.

[0110] In step S401 of some embodiments, the fault record information is a maintenance record, including time and maintenance parts, for example ("2023-01-15", "engine, brake system, steering system"). The time in the fault record information is then normalized, specifically by obtaining the oldest maintenance time recorded in multiple fault record information, and calculating the difference between the time in each fault record information and the oldest maintenance time, and then normalizing the difference. Then, the fault records in which only one part is repaired in the fault record information are deleted, and the fault records in which the number of maintenance parts is greater than two are converted into multiple fault records in which only two parts are repaired. Finally, encoding is performed, and the Euclidean distance is calculated to obtain a distance matrix.

[0111] In one embodiment, the fault maintenance records include ("2023-01-15", "engine, brake system, steering system"), ("2023-02-28", "engine, brake system"), ("2023-03-20", "brake system, transmission system"), ("2023-05-05", "engine, transmission system, fuel system"), ("2023-06-18", "brake system"), ("2023-07-22", "transmission system"), ("2023-08-30", "engine"), ("2023-10-09", "brake system, steering system, transmission system"), ("2023-11-11", "engine, brake system") and ("2023-12-20", "fuel system").

[0112] After normalizing the time, the fault maintenance records include (0.0,"engine, brake system, steering system"), (0.121,"engine, brake system"), (0.176,"brake system, transmission system"), (0.303,"engine, transmission system, fuel system"), (0.424,"brake system"), (0.518,"transmission system"), (0.625,"engine"), (0.735,"brake system, steering system, transmission system"), (0.826,"engine, brake system"), and (0.933,"fuel system").

[0113] The fault record data after deleting the fault records in which only one part was repaired include (0.0,"engine, brake system, steering system"), (0.121,"engine, brake system"), (0.176,"brake system, transmission system"), (0.303,"engine, transmission system, fuel system"), (0.735,"brake system, steering system, transmission system") and (0.826,"engine, brake system").

[0114] Then the fault records with more than two repair parts are converted into multiple fault records with only two parts repaired. The fault record data include (0.0,"engine, brake system"), (0.0,"engine, steering system"), (0.0,"brake system, steering system"), (0.121,"engine, brake system"), (0.176,"brake system, transmission system"), (0.303,"engine, transmission system"), (0.303,"engine, fuel system"), (0.303,"transmission system, fuel system"), (0.735,"brake system, steering system"), (0.735,"brake system, transmission system"), (0.735,"steering system, transmission system") and (0.826,"engine, brake system").

[0115] Then, the converted multiple fault record data are encoded, and the Euclidean distance between any two fault record information is calculated to obtain a distance matrix.

[0116] In step S402 of some embodiments, the fault information is classified based on the distance matrix. Specifically, the fault information is hierarchically clustered according to the distance matrix to obtain a fault information classification tree, that is, fault classification information.

[0117] In step S403 of some embodiments, the fault classification information is a fault information classification tree with a tree structure, and the fault information classification tree contains multiple branches. First, the number of branches of the classification tree is determined, and the number of branches is used as the maximum number of clusters. First, starting from 1 cluster, the fault classification information is clustered, and then the silhouette coefficient under a single cluster is calculated, and then the silhouette coefficient when two clusters are calculated, and this process is continued until the maximum number of clusters is reached. Under each cluster configuration, the corresponding silhouette coefficient is calculated, and finally, the number of clusters with the highest silhouette coefficient is selected as the optimal number of clusters, and the fault classification information is clustered according to the optimal number of clusters, so as to obtain the historical fault parts clustering information.

[0118] See also Figure 5 In some embodiments, step S105 includes but is not limited to steps S501 to S502:

[0119] Step S501, performing feature fusion on historical faulty parts association information and historical faulty parts clustering information through a preset feature fusion model to obtain faulty parts feature information;

[0120] Step S502, optimizing the parameters of a preset part failure prediction model according to the faulty part feature information and the historical faulty part information, to obtain a target part failure prediction model.

[0121] In steps S501 to S502 shown in the embodiment of the present application, feature fusion is performed on historical faulty part association information and historical faulty part clustering information through a preset feature fusion model to obtain faulty part feature information, and then parameters of a preset part fault prediction model are optimized based on the faulty part feature information and historical faulty part information to obtain a target part fault prediction model, thereby improving the prediction capability of the target part fault prediction model for faulty parts.

[0122] See also Figure 6 In some embodiments, step S501 includes but is not limited to steps S601 to S604:

[0123] Step S601, convolving the historical faulty parts association information through a preset convolution layer to obtain the associated faulty parts feature information;

[0124] Step S602, convolving the historical faulty parts clustering information through a preset convolution layer to obtain convolved faulty parts feature information;

[0125] Step S603, performing feature concatenation on the associated faulty part feature information and the convoluted faulty part feature information to obtain aggregated faulty part feature information;

[0126] Step S604, performing feature generation on the aggregated faulty part feature information through a preset fully connected layer to obtain the faulty part feature information.

[0127] In steps S601 to S604 shown in the embodiment of the present application, historical faulty parts association information is convolved through a preset convolution layer to obtain associated faulty parts feature information, and then historical faulty parts clustering information is convolved to obtain convolved faulty parts feature information, and the associated faulty parts feature information and the convolved faulty parts feature information are feature concatenated to obtain aggregated faulty parts feature information, and finally, feature generation is performed on the aggregated faulty parts feature information through a preset fully connected layer to obtain faulty parts feature information that can express potential associations between part failures.

[0128] In step S601 of some embodiments, the preset convolution layer includes a convolution kernel and an activation function, the historical faulty parts association information is input into the convolution kernel, and the output of the convolution kernel is input into the activation function to obtain detailed feature information between the faulty parts, and to obtain associated faulty parts feature information that can express the interaction between the faulty parts.

[0129] In step S602 of some embodiments, the preset convolution layer includes a convolution kernel and an activation function, the historical faulty parts clustering information is input into the convolution kernel, and the output of the convolution kernel is input into the activation function, so as to mine the clustering features in the historical faulty parts clustering information and obtain the convolution faulty parts feature information.

[0130] In step S603 of some embodiments, the historical faulty parts association information and the historical faulty parts clustering information are convolved to the same feature dimension. In the case of the same feature dimension, the historical faulty parts association information and the historical faulty parts clustering information are directly spliced ​​together to obtain aggregated faulty parts feature information.

[0131] In step S604 of some embodiments, the preset fully-connected layer includes a first fully-connected unit, a second fully-connected unit and an activation unit, and the aggregated faulty part feature information is input into the first fully-connected unit and the second fully-connected unit, and the output of the first fully-connected unit is input into the second fully-connected unit, and the output of the second fully-connected unit is input into the activation unit to obtain the final output, i.e., the faulty part feature information.

[0132] In step S502 of some embodiments, the faulty part feature information is input into a preset part fault prediction model to obtain predicted faulty part information, and a loss value is calculated based on the predicted faulty part information and historical faulty part information to obtain a predicted loss value. Finally, the preset part fault prediction model is parameter optimized according to the predicted loss value to obtain a target part fault prediction model.

[0133] See also Figure 7 In some embodiments, the target part fault prediction model includes a long short memory sub-model, a self-attention sub-model and a regression sub-model. Step S106 may include but is not limited to steps S701 to S703:

[0134] Step S701, generating time series features for current fault part information through a long-short memory sub-model to obtain the time series features of current vehicle parts;

[0135] Step S702, feature encoding the current vehicle part temporal features through the self-attention sub-model to obtain the current vehicle part coding features;

[0136] Step S703, performing regression prediction on the current vehicle part code feature through the regression sub-model.

[0137] In steps S701 to S703 shown in the embodiment of the present application, the time series features of the current faulty parts information are generated through the long-short memory sub-model to obtain the time series features of the current vehicle parts, and then the time series features of the current vehicle parts are feature encoded through the self-attention sub-model to obtain the current vehicle part coding features. Finally, the current vehicle part coding features are regressed and predicted through the regression sub-model to achieve accurate fault prediction of the target vehicle.

[0138] In step S701 of some embodiments, the long short-term memory sub-model is a neural network architecture for processing sequence data, which is used to capture the time series characteristics of the fault information of the target vehicle, thereby improving the prediction accuracy of the faulty parts.

[0139] In step S702 of some embodiments, the self-attention sub-model is a neural network module for capturing the temporal dependency between the timing features of the current vehicle parts. The self-attention sub-model includes an input layer, a linear transformation layer, a self-attention layer, and an output layer. The input layer obtains the timing features of the current vehicle parts and passes them to the linear transformation layer, which is then converted into a temporal dependency feature that preliminarily captures the timing features of the current vehicle parts through linear transformation, and is input into the self-attention layer to deeply perceive the temporal dependency between the features, and finally obtains the current vehicle part coding features through the output layer output, thereby improving the prediction accuracy of the faulty parts.

[0140] In step S703 of some embodiments, the regression sub-model is a deep neural network including multiple hidden layers, and the multiple hidden layers are used to capture the nonlinear relationship between the current vehicle coding features and finally output the fault prediction of the parts.

[0141] See also Figure 8 The embodiment of the present application also provides a vehicle parts failure prediction device, which can implement the above-mentioned vehicle parts failure prediction method, and the device includes:

[0142] The data acquisition module 801 is used to acquire the historical operation data of a preset vehicle and obtain the historical parts operation data;

[0143] A fault identification module 802 is used to identify faults on historical parts operation data to obtain historical fault parts information;

[0144] The information mining module 803 is used to mine the historical faulty parts information for related information to obtain the historical faulty parts related information;

[0145] The information clustering module 804 is used to cluster the historical faulty parts information to obtain the historical faulty parts clustering information;

[0146] The parameter optimization module 805 is used to optimize the parameters of the preset part failure prediction model according to the historical fault part information, the historical fault part association information and the historical fault part clustering information to obtain the target part failure prediction model;

[0147] The information acquisition module 806 is used to acquire the fault information of the target vehicle, obtain the current faulty part information, and perform fault prediction on the current faulty part information through the target part fault prediction model.

[0148] The specific implementation of the vehicle part failure prediction device is basically the same as the specific implementation of the above-mentioned vehicle part failure prediction method, and will not be repeated here.

[0149] The embodiment of the present application also provides an electronic device, the electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the above-mentioned vehicle parts fault prediction method when executing the computer program. The electronic device can be any intelligent terminal including a tablet computer, a vehicle-mounted computer, etc.

[0150] See also Fig. 9 , Fig. 9 The hardware structure of an electronic device of another embodiment is illustrated, and the electronic device includes:

[0151] The processor 901 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application;

[0152] The memory 902 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 902 can store an operating system and other application programs. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 902, and the processor 901 calls and executes the vehicle parts fault prediction method of the embodiment of this application;

[0153] Input / output interface 903, used to implement information input and output;

[0154] Communication interface 904, used to realize communication interaction between the device and other devices, which can be realized through wired mode (such as USB, network cable, etc.) or wireless mode (such as mobile network, WI FI, Bluetooth, etc.);

[0155] A bus 905 that transmits information between various components of the device (e.g., the processor 901, the memory 902, the input / output interface 903, and the communication interface 904);

[0156] The processor 901 , the memory 902 , the input / output interface 903 and the communication interface 904 are connected to each other in communication within the device via a bus 905 .

[0157] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above-mentioned vehicle part failure prediction method is implemented.

[0158] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0159] The vehicle part fault prediction method, vehicle part fault prediction device, electronic device and storage medium provided in the embodiments of the present application obtain historical part operation data, then perform fault identification on the historical part operation data to obtain historical fault part information, then perform correlation information mining on the historical fault part information to obtain historical fault part correlation information, and perform fault information clustering on the historical part fault information to obtain historical fault part clustering information, and then perform parameter optimization on a preset part fault prediction model through the historical fault part clustering information and the historical fault part correlation information, so that the preset part fault prediction model can perceive the fault characteristics of the part from two dimensions of clustering information and correlation information, and enable the preset part fault prediction model to have good prediction capabilities for part faults, and finally obtain current fault part information, and perform fault prediction on the current fault part information through a target part fault prediction model, thereby improving the prediction accuracy of vehicle part faults.

[0160] The embodiments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0161] Those skilled in the art will appreciate that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0162] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place, or they may be 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.

[0163] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0164] The terms "first", "second", "third", "fourth", etc. (if any) in the specification 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, system, 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.

[0165] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0166] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation, such as 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 devices or units, which can be electrical, mechanical or other forms.

[0167] The units described above 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.

[0168] 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.

[0169] 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 is essentially 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, and the computer software product is stored in a storage medium, including multiple instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store programs.

[0170] The preferred embodiments of the present application are described above with reference to the accompanying drawings, but the scope of the rights of the present application is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present application should be within the scope of the rights of the present application.

Claims

1. A vehicle parts failure prediction method, characterized in that: The method comprises: Obtain the historical operation data of the preset vehicle and obtain the historical parts operation data; Performing fault identification on the historical parts operation data to obtain historical fault parts information; Performing associated information mining on the historical faulty parts information to obtain historical faulty parts associated information; Performing fault information clustering on the historical fault parts information to obtain historical fault parts clustering information; Optimizing parameters of a preset part failure prediction model according to the historical fault part information, the historical fault part association information, and the historical fault part clustering information to obtain a target part failure prediction model; Fault information of a target vehicle is acquired to obtain current faulty parts information, and fault prediction is performed on the current faulty parts information using the target parts fault prediction model.

2. The method according to claim 1, characterized in that The method of optimizing the parameters of a preset part failure prediction model according to the historical fault part information, the historical fault part association information and the historical fault part clustering information to obtain a target part failure prediction model includes: Performing feature fusion on the historical faulty parts association information and the historical faulty parts clustering information through a preset feature fusion model to obtain faulty parts feature information; The parameters of the preset part failure prediction model are optimized according to the faulty part feature information and the historical faulty part information to obtain the target part failure prediction model.

3. The method according to claim 2, characterized in that The method of fusing the historical faulty parts association information and the historical faulty parts clustering information by using a preset feature fusion model to obtain the faulty parts feature information includes: Convolving the historical faulty parts associated information through a preset convolution layer to obtain associated faulty parts feature information; Convolving the historical faulty parts clustering information through the preset convolution layer to obtain convolved faulty parts feature information; Performing feature concatenation on the associated faulty part feature information and the convolved faulty part feature information to obtain aggregated faulty part feature information; The aggregated faulty part feature information is subjected to feature generation through a preset fully connected layer to obtain the faulty part feature information.

4. The method according to any one of claims 1 to 3, characterized in that: The target part fault prediction model includes a long-short memory sub-model, a self-attention sub-model and a regression sub-model; the fault prediction of the current fault part information by using the target part fault prediction model includes: Generate the time series feature of the current fault part information by using the long-short memory sub-model to obtain the time series feature of the current vehicle part; Performing feature encoding on the time series features of the current vehicle parts through the self-attention sub-model to obtain the current vehicle part coding features; The regression sub-model is used to perform regression prediction on the current vehicle part code feature.

5. The method according to any one of claims 1 to 3, characterized in that: The performing of associated information mining on the historical faulty parts information to obtain the historical faulty parts associated information includes: Performing frequency statistics on the historical faulty parts information to obtain parts failure frequency data; The part failure frequency data is screened according to a preset frequent item set threshold to obtain a frequent failure part item set; Generate association rules based on the frequent faulty parts item set to obtain faulty parts association rules; The faulty parts association rules are screened to obtain the historical faulty parts association information.

6. The method according to any one of claims 1 to 3, characterized in that: The historical fault part information includes a plurality of fault record information; the historical fault part information is clustered to obtain the historical fault part clustering information, including: Calculate the Euclidean distance between any two fault record information to obtain the distance matrix; Classifying the fault record information according to the distance matrix to obtain fault classification information; The fault classification information is screened by silhouette coefficient to obtain the historical fault parts clustering information.

7. The method according to any one of claims 1 to 3, characterized in that: The performing fault identification on the historical part operation data to obtain historical fault part information includes: Performing mean calculation on the historical parts operation data to obtain operation mean data; Calculating the standard deviation of the historical parts operation data to obtain operation standard deviation data; Performing aggregation calculation according to the operation mean data, the operation standard deviation data and the historical part operation data to obtain an operation data offset; The historical faulty part information is filtered out from the historical part operation data according to the operation data offset.

8. A vehicle parts failure prediction device, characterized in that: The device comprises: A data acquisition module is used to acquire historical operation data of a preset vehicle and obtain historical parts operation data; A fault identification module is used to identify faults on the historical parts operation data to obtain historical fault parts information; An information mining module is used to mine related information on the historical faulty parts information to obtain related information on the historical faulty parts; An information clustering module, used for clustering the historical faulty parts information to obtain historical faulty parts clustering information; A parameter optimization module, used to optimize the parameters of a preset part failure prediction model according to the historical fault part information, the historical fault part association information and the historical fault part clustering information, so as to obtain a target part failure prediction model; The information acquisition module is used to acquire the fault information of the target vehicle, obtain the current fault part information, and perform fault prediction on the current fault part information through the target part fault prediction model.

9. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the vehicle part failure prediction method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the vehicle parts failure prediction method according to any one of claims 1 to 7 is implemented.

Images