Fault maintenance method and device, computer equipment and storage medium

Abstract

The invention relates to a fault maintenance method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: receiving a fault maintenance request, wherein the fault maintenance request comprises fault description information of fault equipment and a fault equipment identifier; the fault description information comprises fault occurrence time and a diagnosis fault code; querying a signal database according to the fault equipment identifier, and obtaining target signal data associated with the fault occurrence time; the signal database stores a corresponding relationship among the fault equipment identifier, the fault occurrence time and the target signal data; and inputting the acquired target signal data and the diagnosis fault code into a fault maintenance model, and outputting a recommended fault maintenance scheme by the fault maintenance model. By adopting the method, the adaptability of a fault maintenance scheme can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of fault diagnosis, and in particular to a fault repair method, apparatus, computer equipment, storage medium and computer program product. Background Art

[0002] Fault diagnosis of new energy vehicles is an important part of ensuring their stable operation and safety performance. With the popularization and application of new energy vehicles, fault diagnosis technology has also developed rapidly. The current status of fault diagnosis of new energy vehicles mainly includes two aspects. On the one hand, the fault types and causes of new energy vehicles are different from those of traditional fuel vehicles. The electric motors and battery systems used in new energy vehicles have unique working principles and characteristics, so it is necessary to develop corresponding fault diagnosis technologies in a targeted manner. On the other hand, fault diagnosis of new energy vehicles involves multiple systems and components, such as power batteries, motor control systems, electronic control units, etc., so it is necessary to use a variety of technical means for diagnosis.

[0003] In the related art, when a vehicle fails, the vehicle type and the fault type are input into a pre-trained model to obtain a corresponding fault solution. However, the input data of the model in this way is relatively limited, resulting in low adaptability of the fault solution. Summary of the invention

[0004] Based on this, it is necessary to provide a fault repair method, apparatus, computer equipment, computer-readable storage medium and computer program product that can improve the adaptability of fault repair solutions in order to address the above technical problems.

[0005] In a first aspect, the present application provides a fault repair method, comprising:

[0006] receiving a fault repair request, the fault repair request including fault description information of the faulty device and the faulty device identifier; the fault description information including the fault occurrence time and the diagnostic fault code;

[0007] According to the faulty device identification, query the signal database to obtain the target signal data associated with the fault occurrence time; the signal database stores the corresponding relationship between the faulty device identification, the fault occurrence time and the target signal data;

[0008] The acquired target signal data and diagnostic fault codes are input into the fault repair model, and the fault repair model outputs a recommended fault repair plan.

[0009] In one embodiment, the fault repair model includes a fault classification model and a repair solution generation model; the acquired target signal data and the diagnostic fault code are input into the fault repair model, and the fault repair model outputs a recommended fault repair solution, including:

[0010] Input the target signal data and diagnostic trouble codes into a fault classification model to obtain multiple maintenance guidance messages output by the fault classification model;

[0011] Input the multiple maintenance guidance messages into a maintenance plan generation model to obtain a recommended fault maintenance plan output by the maintenance plan generation model.

[0012] In one embodiment, inputting the multiple maintenance guidance messages into a maintenance plan generation model to obtain a recommended fault maintenance plan output by the maintenance plan generation model includes:

[0013] Taking the maintenance guidance message with the largest occupancy ratio among the multiple maintenance guidance messages as the target guidance message; inputting the target guidance message into the maintenance plan generation model to obtain a recommended fault maintenance plan output by the maintenance plan generation model.

[0014] In one embodiment, the training steps of the fault classification model include:

[0015] Obtain historical fault maintenance requests, where the historical fault maintenance requests include historical fault description information and historical fault device identifiers; the historical fault description information includes the historical fault occurrence time and historical diagnostic trouble codes;

[0016] Obtain historical maintenance texts corresponding to the historical fault maintenance requests;

[0017] According to the historical fault device identifier, query the signal database to obtain historical signal data associated with the historical fault occurrence time;

[0018] Train an initial decision tree model based on the historical diagnostic trouble codes, historical signal data, and historical maintenance texts to obtain a fault classification model.

[0019] In one embodiment, the initial decision tree model includes M initial decision trees, and the initial model parameters of the M initial decision trees are different; training the initial decision tree model based on the historical diagnostic trouble codes, historical signal data, and historical maintenance texts to obtain a fault classification model includes:

[0020] For each of the M initial decision trees, input the historical diagnostic trouble codes and historical signal data into the initial decision tree to obtain a predicted maintenance text output by the initial decision tree; calculate the model loss of the initial decision tree based on the historical maintenance text and the predicted maintenance text;

[0021] Based on the model losses corresponding to the M initial decision trees respectively, adjust the model parameters of the M initial decision trees until the M initial decision trees respectively meet the corresponding stopping conditions to obtain a fault classification model.

[0022] In one embodiment, the fault classification model includes M decision trees; the fault repair method further includes:

[0023] Receiving the actual fault repair plan corresponding to the recommended fault repair plan;

[0024] For each of the M decision trees, input the target signal data and diagnostic fault codes into the decision tree to obtain the updated predicted repair text output by the decision tree; based on the actual fault repair plan and the updated predicted repair text, calculate the updated model loss of the decision tree.

[0025] Based on the updated model losses corresponding to the M decision trees respectively, update the model parameters of the M decision trees to obtain an updated fault classification model.

[0026] In one embodiment, the steps for constructing the signal database include:

[0027] Obtaining the fault data uploaded by the vehicle when a fault occurs, where the fault data includes the real-time signal data collected when the vehicle has a fault, the time when the vehicle has a fault, and the device identifier of the vehicle;

[0028] Storing the correspondence relationship of the real-time signal data, time, and device identifier in the signal database.

[0029] In one embodiment, the fault description information further includes a fault description text; inputting the obtained target signal data and diagnostic fault codes into the fault repair model, and the fault repair model outputs a recommended fault repair plan, including:

[0030] Inputting the fault description text, target signal data, and diagnostic fault codes into the fault repair model, and the fault repair model outputs a recommended fault repair plan.

[0031] In a second aspect, the present application further provides a fault repair device, including:

[0032] A receiving module, configured to receive a fault repair request, where the fault repair request includes the fault description information of the faulty device and the faulty device identifier; the fault description information includes the fault occurrence time and diagnostic fault codes;

[0033] A query module, configured to query the signal database according to the faulty device identifier to obtain the target signal data associated with the fault occurrence time; the signal database stores the correspondence relationship of the faulty device identifier, fault occurrence time, and target signal data;

[0034] An output module, configured to input the obtained target signal data and diagnostic fault codes into the fault repair model, and the fault repair model outputs a recommended fault repair plan.

[0035] In a third aspect, the present application further provides a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0036] Receive a fault repair request, where the fault repair request includes fault description information of the faulty device and a faulty device identifier; the fault description information includes the fault occurrence time and a diagnostic fault code;

[0037] According to the faulty device identifier, query the signal database to obtain target signal data associated with the fault occurrence time; the signal database stores the corresponding relationship between the faulty device identifier, the fault occurrence time, and the target signal data;

[0038] Input the obtained target signal data and the diagnostic fault code into a fault repair model, and the fault repair model outputs a recommended fault repair solution.

[0039] In a fourth aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0040] Receive a fault repair request, where the fault repair request includes fault description information of the faulty device and a faulty device identifier; the fault description information includes the fault occurrence time and a diagnostic fault code;

[0041] According to the faulty device identifier, query the signal database to obtain target signal data associated with the fault occurrence time; the signal database stores the corresponding relationship between the faulty device identifier, the fault occurrence time, and the target signal data;

[0042] Input the obtained target signal data and the diagnostic fault code into a fault repair model, and the fault repair model outputs a recommended fault repair solution.

[0043] In a fifth aspect, the present application further provides a computer program product, including a computer program. When the computer program is executed by a processor, the following steps are implemented:

[0044] Receive a fault repair request, where the fault repair request includes fault description information of the faulty device and a faulty device identifier; the fault description information includes the fault occurrence time and a diagnostic fault code;

[0045] According to the faulty device identifier, query the signal database to obtain target signal data associated with the fault occurrence time; the signal database stores the corresponding relationship between the faulty device identifier, the fault occurrence time, and the target signal data;

[0046] Input the obtained target signal data and the diagnostic fault code into a fault repair model, and the fault repair model outputs a recommended fault repair solution.

[0047] The above-mentioned fault repair method, device, computer equipment, storage medium and computer program product receive a fault repair request, which includes the fault description information and the fault device identifier of the faulty device; the fault description information includes the fault occurrence time and the diagnostic fault code; according to the fault device identifier, query the signal database to obtain the target signal data associated with the fault occurrence time; the signal database stores the corresponding relationship between the fault device identifier, the fault occurrence time and the target signal data; input the obtained target signal data and the diagnostic fault code into the fault repair model, and the fault repair model outputs a recommended fault repair plan. This method of querying the target signal data associated with the fault occurrence time based on the fault description information and the fault device identifier of the faulty device, and inputting the target signal data and the diagnostic fault code in the fault description information into the fault repair model to obtain a recommended fault repair plan can perform fault diagnosis by combining the target signal data associated with the vehicle fault occurrence time and the diagnostic fault code. Compared with the method of performing fault diagnosis only based on the vehicle type and the fault type, the input data of the model is more comprehensive, and the obtained recommended fault repair plan has a higher adaptability to the current fault. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0049] Figure 1 It is an application environment diagram of the fault repair method in an embodiment;

[0050] Figure 2 It is a flowchart of the fault repair method in an embodiment;

[0051] Figure 3 It is a sub-flowchart of step 206 in an embodiment;

[0052] Figure 4 It is an overall flowchart of the fault repair method in an embodiment;

[0053] Figure 5 It is a flowchart of the fault repair method in another embodiment;

[0054] Figure 6 It is a structural block diagram of the fault repair device in an embodiment;

[0055] Figure 7 It is an internal structure diagram of the computer equipment in an embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[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] The fault repair method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the cloud server 102 communicates with the vehicle 104 through the network. The data storage system can store the data that the cloud server 102 needs to process. The data storage system can be integrated on the cloud server 102, and can also be placed on other network servers. The fault repair method provided in the embodiment of the present application can be executed by the cloud server 102, and the cloud server 102 receives a fault repair request, and the fault repair request includes the fault description information and the fault device identification of the faulty device; the fault description information includes the fault occurrence time and the diagnostic fault code; according to the faulty device identification, query the signal database to obtain the target signal data associated with the fault occurrence time; the signal database stores the corresponding relationship between the faulty device identification, the fault occurrence time and the target signal data; the acquired target signal data and the diagnostic fault code are input into the fault repair model, and the fault repair model outputs the recommended fault repair plan. Among them, the cloud server 102 can be implemented with an independent server or a server cluster consisting of multiple servers.

[0058] In an exemplary embodiment, Figure 2 As shown, a fault repair method is provided, which is applied to Figure 1 Taking the cloud server 102 in the example as an example, the method includes the following steps 202 to 206. Among them:

[0059] Step 202: Receive a fault repair request, where the fault repair request includes fault description information and a fault device identifier of the fault device; the fault description information includes a fault occurrence time and a diagnostic fault code.

[0060] The fault repair method can be applied to various types of fault repair scenarios of faulty equipment. For example, the faulty equipment can be a complete vehicle, a component in the vehicle (such as an engine, a power system, etc.), other mobile terminals or fixed terminals, etc.

[0061] After a fault occurs, a user or maintenance personnel of the faulty device usually initiates a fault repair request. There are many ways to initiate a fault repair request, for example, by logging into a maintenance website or a maintenance program of a mobile terminal.

[0062] The type of the fault repair request received by the cloud server can be a text type, an electronic form type, etc. According to the type of the fault repair request, a parsing method is determined, and the fault repair request is parsed by using the parsing method corresponding to the type of the fault repair request to obtain fault description information and a fault device identifier; the fault description information includes a fault occurrence time and a diagnostic fault code.

[0063] The fault description information is the information used to describe the fault condition of the fault device in the fault repair request. For example, it can include information such as the fault occurrence time, the diagnostic fault code, and the fault description text.

[0064] Among them, the fault occurrence time is the time when the fault device indicated in the fault repair request has a fault.

[0065] The diagnostic fault code can be a DTC (Diagnostic Trouble Code) code.

[0066] Each device has a unique device identifier. For example, the device identifier can be a device code. The fault device identifier is the device identifier of the fault device.

[0067] Step 204: According to the fault device identifier, query the signal database to obtain target signal data associated with the fault occurrence time; the signal database stores the corresponding relationship between the fault device identifier, the fault occurrence time, and the target signal data.

[0068] Among them, the signal database stores the signal data received by each device in real time at historical moments and the current moment, the fault device identifier of the fault device, and the reception time of each signal data. The signal database also stores data such as the corresponding relationship between the fault device identifier, the fault occurrence time, and the target signal data.

[0069] The target signal data is the signal data received by the fault device within a preset time period before and after the fault occurrence time. For example, if the fault occurrence time is T0, the target signal data can be the signal data in the time period from T0 - t1 to T0 + t2, where both t1 and t2 are greater than 0. The target signal data can be a voltage value, a current value, etc.

[0070] The cloud server queries the corresponding relationship between the fault device identifier, the fault occurrence time, and the target signal data in the signal database according to the fault device identifier, and obtains the target signal data associated with the fault occurrence time.

[0071] Step 206: Input the obtained target signal data and the diagnostic fault code into the fault repair model, and the fault repair model outputs a recommended fault repair plan.

[0072] Among them, the fault repair model is a trained machine learning model. In some embodiments, the fault repair model can be a machine learning model with a classification function. The fault repair model is trained based on historical diagnostic fault codes, historical signal data, and historical repair texts.

[0073] The cloud server inputs the diagnostic fault code and the target signal data into the trained fault repair model, and obtains the recommended fault repair solution output by the fault repair model.

[0074] The recommended fault repair solution is a text used to describe the fault repair method, and can be used to guide maintenance personnel to perform fault repair on the faulty equipment.

[0075] The method of using the fault repair model for fault diagnosis improves the output efficiency of the recommended fault repair solution on the one hand. On the other hand, by combining the target signal data associated with the fault occurrence time and the diagnostic fault code, rather than simply based on the original repair records, the generated recommended fault repair solution has a higher adaptability to the current fault, which is beneficial to increasing the possibility of successful fault repair.

[0076] In the above-mentioned fault repair method, by receiving a fault repair request, the fault repair request includes the fault description information of the faulty equipment and the faulty equipment identifier; the fault description information includes the fault occurrence time and the diagnostic fault code; according to the faulty equipment identifier, query the signal database to obtain the target signal data associated with the fault occurrence time; the signal database stores the corresponding relationship between the faulty equipment identifier, the fault occurrence time, and the target signal data; input the obtained target signal data and the diagnostic fault code into the fault repair model, and the fault repair model outputs the recommended fault repair solution. This method of querying the target signal data associated with the fault occurrence time based on the fault description information and the faulty equipment identifier of the faulty equipment, and inputting the target signal data and the diagnostic fault code in the fault description information into the fault repair model to obtain the recommended fault repair solution can perform fault diagnosis by combining the target signal data associated with the vehicle fault occurrence time and the diagnostic fault code. Compared with the method of performing fault diagnosis only based on the vehicle type and fault type, the input data of the model is more comprehensive, and the obtained recommended fault repair solution has a higher adaptability to the current fault.

[0077] In an exemplary embodiment, the fault repair model includes a fault classification model and a repair solution generation model; as Figure 3 shown, step 206 includes steps 302 to 304. Among them:

[0078] Step 302, input the target signal data and the diagnostic fault code into the fault classification model, and obtain multiple maintenance guidance information output by the fault classification model.

[0079] Step 304: Input multiple maintenance guidance messages into the maintenance plan generation model to obtain the recommended fault maintenance plan output by the maintenance plan generation model.

[0080] Among them, the fault classification model is a trained machine learning model with classification functions, such as a decision tree model or other classification models. After the target signal data and diagnostic fault codes are input into the fault classification model, multiple maintenance guidance messages are output.

[0081] The maintenance guidance message is a guidance text used to describe the fault maintenance method. Usually, it is a brief description of the fault maintenance method. For example, the fault maintenance text can be "Update the software of the fuel pump".

[0082] The maintenance plan generation model is a large language model, such as the GPT model, which is used to generate a recommended fault maintenance plan based on multiple maintenance guidance messages. Since the maintenance guidance messages output by the fault classification model are often incomprehensible to maintenance personnel or cannot be directly used as maintenance guidance, the cloud server inputs multiple maintenance guidance messages into the maintenance plan generation model to obtain an understandable and expressible fault maintenance plan.

[0083] The main difference between the recommended fault maintenance plan and the maintenance guidance message is that the recommended fault maintenance plan can be a detailed description of the fault maintenance method. For example, the recommended fault maintenance plan can be "Open the software of the fuel pump, enter the target page, modify the maximum allowable flow rate in the target page to 5, save the modified page, exit the page, and close the software of the fuel pump".

[0084] In this embodiment, first, multiple maintenance guidance messages are output through the fault classification model. The maintenance guidance signal is output based on the target signal data associated with the fault occurrence time and the diagnostic fault code, and has high accuracy. Then, the maintenance plan generation model outputs an understandable and expressible recommended fault maintenance plan based on the input multiple maintenance guidance messages, improving the adaptability of the fault maintenance plan to the current fault. At the same time, using the generated recommended fault maintenance plan for fault maintenance is beneficial to improving the guidance effect on maintenance personnel and increasing the success rate and efficiency of fault maintenance.

[0085] In an exemplary embodiment, inputting multiple maintenance guidance messages into the maintenance plan generation model to obtain the recommended fault maintenance plan output by the maintenance plan generation model includes: using the maintenance guidance message with the largest occupancy ratio among the multiple maintenance guidance messages as the target guidance message; inputting the target guidance message into the maintenance plan generation model to obtain the recommended fault maintenance plan output by the maintenance plan generation model.

[0086] Among them, multiple maintenance guidance messages may be different from each other, and there may also be the same maintenance guidance messages. The cloud server obtains the quantity of multiple maintenance guidance messages and the frequency of each maintenance guidance message appearing in the multiple maintenance guidance messages respectively, divides the frequency corresponding to each maintenance guidance message by the quantity of the multiple maintenance guidance messages, and obtains the occupancy ratio of each maintenance guidance message in the multiple maintenance guidance messages respectively.

[0087] The cloud server takes the maintenance guidance message with the largest occupancy ratio among the multiple maintenance guidance messages as the target guidance message, and inputs the target guidance message into the maintenance plan generation model to obtain the recommended fault maintenance plan output by the maintenance plan generation model.

[0088] In this embodiment, according to the occupancy ratio of each maintenance guidance message in the multiple maintenance guidance messages, the target guidance message with the largest occupancy ratio is screened out. Since the output ratio of the target guidance message is the highest, using the target guidance message as the input of the maintenance plan generation model is beneficial to generating a relatively accurate recommended fault maintenance plan.

[0089] In an exemplary embodiment, the training steps of the fault classification model include: obtaining historical fault maintenance requests, where the historical fault maintenance requests include historical fault description information and historical fault device identifiers; the historical fault description information includes the historical fault occurrence time and the historical diagnostic fault code; obtaining the historical maintenance text corresponding to the historical fault maintenance request; according to the historical fault device identifier, querying the signal database to obtain the historical signal data associated with the historical fault occurrence time; and training the initial decision tree model based on the historical diagnostic fault code, the historical signal data, and the historical maintenance text to obtain the fault classification model.

[0090] Among them, the historical fault maintenance request is a fault maintenance request initiated by a user or a maintenance personnel for a historical fault device at a historical moment.

[0091] The cloud server parses the historical fault maintenance request according to the type of the historical fault maintenance request and by using a corresponding parsing method to obtain the historical fault description information and the historical fault device identifier. The historical fault description information includes the historical fault occurrence time and the historical diagnostic fault code.

[0092] The historical fault description information is information used to describe the fault situation of the historical fault device. For example, it may include information such as the historical fault occurrence time, the historical diagnostic fault code, and the historical fault description text.

[0093] The historical fault occurrence time is the time when the historical fault device fails as indicated in the historical fault maintenance request.

[0094] The historical diagnostic fault code can be a DTC code. The historical fault device identifier is the device identifier of the historical fault device.

[0095] The historical maintenance text is the text of the maintenance method obtained from the maintenance personnel for the historical fault maintenance request.

[0096] The cloud server queries the correspondence between the fault device identifier, the fault occurrence time, and the target signal data in the signal database according to the historical fault device identifier, and obtains the historical signal data associated with the historical fault occurrence time. The historical signal data is the signal data received by the historical fault device within a preset time period before and after the historical fault occurrence time. For example, if the historical fault occurrence time is T1, the target signal data can be the signal data in the time period from T1 - t3 to T1 + t4, where both t3 and t4 are greater than 0. The historical signal data can be a voltage value, or a current value, etc.

[0097] The cloud server uses the historical diagnostic fault code and the historical signal data corresponding to the historical fault maintenance request as training samples, and the historical maintenance text as a training label to train the fault classification model. Taking the initial model of the fault classification model as a decision tree model as an example, the initial decision tree model is a pre-trained decision tree model, and the trained fault classification model is used to output maintenance guidance information.

[0098] In this embodiment, the training samples combine the historical fault data associated with the historical fault time and the historical diagnostic fault code, and use the historical maintenance text as a training label. The trained fault classification model can be continuously adjusted according to the increasing historical fault requests and the continuously updated signal database, which is beneficial to improving the adaptability of the fault maintenance plan to the current fault.

[0099] In an exemplary embodiment, the initial decision tree model includes M initial decision trees, and the initial model parameters of the M initial decision trees are different; based on the historical diagnostic fault code, the historical signal data, and the historical maintenance text, the initial decision tree model is trained to obtain the fault classification model, including: for each of the M initial decision trees, input the historical diagnostic fault code and the historical signal data into the initial decision tree to obtain the predicted maintenance text output by the initial decision tree; based on the historical maintenance text and the predicted maintenance text, calculate the model loss of the initial decision tree; based on the model losses corresponding to the M initial decision trees respectively, adjust the model parameters of the M initial decision trees until the M initial decision trees respectively meet the corresponding stop conditions to obtain the fault classification model.

[0100] Among them, the initial decision tree model includes M decision trees, and the initial model parameters of the M initial decision trees are different from each other, so that each initial decision tree can output different prediction results during the training process. Each initial decision tree can be trained independently, and after the training is completed, a fault classification model including M decision trees is obtained.

[0101] For the training process of each initial decision tree, the historical diagnostic fault codes and historical signal data are input into the initial decision tree, and the initial decision tree outputs the predicted maintenance text corresponding to the historical fault request. The historical maintenance text and the predicted maintenance text are brought into the model loss function to obtain the model loss of the initial decision tree.

[0102] Based on the model loss corresponding to each initial decision tree, the initial decision tree is trained independently until the model loss meets the preset stop condition, and a trained decision tree is obtained.

[0103] According to this method, the M initial decision trees are trained in sequence to obtain a fault classification model including M decision trees.

[0104] In this embodiment, by training the M initial decision trees separately, a fault classification model including M decision trees is obtained. Since the fault classification model includes M decision trees, when the trained fault classification model is used for fault diagnosis, M maintenance guidance messages can be output. The M maintenance guidance messages provide multiple possible methods for fault maintenance for fault maintenance personnel, which is beneficial to improving the success rate of fault maintenance. At the same time, the fault classification model is trained based on historical fault requests and data in the signal database. With the continuous enrichment of historical fault requests and the signal database, it is beneficial to improve the classification accuracy of the fault classification model.

[0105] In an exemplary embodiment, the fault classification model includes M decision trees; the fault maintenance method further includes: receiving the actual fault maintenance plan corresponding to the recommended fault maintenance plan; for each of the M decision trees, inputting the target signal data and the diagnostic fault code into the decision tree to obtain the updated predicted maintenance text output by the decision tree; calculating the updated model loss of the decision tree based on the actual fault maintenance plan and the updated predicted maintenance text; updating the model parameters of the M decision trees respectively based on the updated model losses corresponding to the M decision trees to obtain an updated fault classification model.

[0106] Among them, the cloud server pushes the generated recommended fault maintenance plan to the maintenance personnel to guide the maintenance personnel to repair the faulty equipment according to the recommended fault maintenance plan. During the maintenance process, the maintenance personnel adjust the recommended fault maintenance plan according to the actual maintenance situation, and the actual fault maintenance plan is sent to the cloud server by the maintenance personnel or other relevant authorized personnel.

[0107] The cloud server receives the actual fault repair plan corresponding to the recommended fault repair plan, and the actual fault repair plan is an optimized plan for the recommended fault repair plan. The cloud server will further adjust the fault classification model using the actual fault repair plan to improve the classification accuracy of the fault classification model.

[0108] Specifically, for each decision tree in the fault classification model, the cloud server inputs the target signal data and diagnostic fault codes into the decision tree as training samples, and obtains the updated predicted repair text output by the decision tree. The cloud server takes the actual fault repair plan as the training label and inputs it into the model loss function together with the updated predicted repair text to calculate the updated model loss of the decision tree.

[0109] Since the training label and model loss of the fault classification model have changed, the cloud server updates the model parameters of the corresponding decision tree based on the updated model loss to obtain the updated fault classification model.

[0110] In this embodiment, by receiving the actual fault repair plan output after the repair personnel adjust the recommended fault repair plan, and taking the actual fault repair plan as the training label of the fault classification model to update the model parameters of the fault classification model, it is beneficial to improve the model processing ability and classification accuracy of the fault classification model.

[0111] In an exemplary embodiment, the steps for constructing the signal database include: obtaining the fault data uploaded by the vehicle when a fault occurs, where the fault data includes the real-time signal data collected when the vehicle has a fault, the time when the vehicle has a fault, and the device identifier of the vehicle; storing the correspondence between the real-time signal data, the time, and the device identifier in the signal database.

[0112] Among them, the real-time signal data can be the signal data collected when the vehicle has a fault, or the signal data collected by the vehicle in real time. The real-time signal data can include the collected data of devices such as vehicle sensors, battery management systems, and charging systems, or the vehicle operation state data, for example, including data such as speed, acceleration, temperature, voltage, and current.

[0113] In some embodiments, signal data points are preset in the system for signal collection, and the signal data points are used to record the signal data that is crucial for fault repair, which is beneficial to collecting the real-time signal data related to fault repair.

[0114] Each real-time signal data corresponds to the device identifier of the vehicle and the time when the fault occurs. The cloud server writes the real-time signal data, the time when the vehicle has a fault, and the device identifier of the vehicle into the preset database, and writes the correspondence between the real-time signal data, the time, and the device identifier into the preset database to obtain the signal database.

[0115] In some embodiments, the cloud server can also perform data cleaning and preprocessing on the real-time signal data, remove abnormal data and noise data in the real-time signal data to obtain the cleaned data, perform normalization processing on the cleaned data to obtain the normalized data. Write the normalized data corresponding to the real-time signal data, the moment when the vehicle breaks down, and the device identifier of the vehicle into a preset database, and write the corresponding relationship between the real-time signal data, the moment, and the device identifier into the preset database to obtain a signal database.

[0116] In this embodiment, the data in the signal database mainly comes from the real-time signal data collected by the vehicle in real time. The data in the signal database can be updated in real time dynamically, with high real-time performance, which is beneficial to training a fault repair model with better diagnostic performance and generating a more accurate recommended fault repair plan.

[0117] In one of the embodiments, the fault description information further includes a fault description text; input the obtained target signal data and diagnostic fault code into the fault repair model, and the fault repair model outputs a recommended fault repair plan, including: input the fault description text, target signal data, and diagnostic fault code into the fault repair model, and the fault repair model outputs a recommended fault repair plan.

[0118] Among them, the fault description text is a description text of the fault details. For example, the fault description text can be "The instrument prompts a power system fault and the vehicle loses power".

[0119] The cloud server can also input the fault description text, target signal data, and diagnostic fault code into the fault repair model together, and the fault repair model outputs a recommended fault repair plan. Correspondingly, the fault repair model includes a fault classification model and a repair plan generation model. Input the fault description text, target signal data, and diagnostic fault code into the fault classification model to obtain multiple repair guidance information output by the fault classification model, and input the multiple repair guidance information into the repair plan generation model to obtain the recommended fault repair plan output by the repair plan generation model.

[0120] Among them, the fault classification model is trained based on historical fault description texts, historical diagnostic fault codes, historical signal data, and historical repair texts.

[0121] Specifically, the cloud server obtains historical fault repair requests, where the historical fault repair requests include historical fault description information and historical fault device identifiers; the historical fault description information includes the historical fault occurrence time, historical diagnostic fault codes, and historical fault description texts; obtains historical repair texts corresponding to the historical fault repair requests; queries the signal database according to the historical fault device identifiers to obtain historical signal data associated with the historical fault occurrence time; and trains the initial decision tree model based on the historical fault description texts, historical diagnostic fault codes, historical signal data, and historical repair texts to obtain a fault classification model.

[0122] Among them, training the initial decision tree model based on the historical fault description texts, historical diagnostic fault codes, historical signal data, and historical repair texts to obtain a fault classification model includes: for each of the M initial decision trees, inputting the historical fault description texts, historical diagnostic fault codes, and historical signal data into the initial decision tree to obtain the predicted repair text output by the initial decision tree; calculating the model loss of the initial decision tree based on the historical repair text and the predicted repair text; and adjusting the model parameters of the M initial decision trees respectively based on the model losses corresponding to the M initial decision trees until the M initial decision trees respectively meet the corresponding stopping conditions to obtain a fault classification model.

[0123] In this embodiment, since the fault description text includes the description text of the fault details, inputting the fault description text, target signal data, and diagnostic fault codes into the fault repair model is beneficial to improving the adaptability of the fault repair plan to the current fault when diagnosing faults for new vehicle models without a large amount of historical repair data.

[0124] To illustrate the fault repair method and effect in this solution in detail, the following uses a most detailed embodiment for illustration:

[0125] Taking the fault repair scenario of a new vehicle model as an example for illustration. As Figure 4 shown is the overall flowchart of the fault repair method in an embodiment. The fault repair method proposed in the embodiment of this application combines the feedback descriptions of the fault by users or maintenance personnel, the fault status, the fault manifestations, the signal data, and the diagnostic fault codes, and uses the trained fault repair model to perform fault detection and generate the optimal fault repair plan, improving the fault repair efficiency. Among them, since new vehicle models have fewer fault cases, traditional fault repair methods often have difficulty obtaining accurate fault repair plans when the number of fault cases is insufficient. However, the fault repair method provided in the embodiment of this application can generate an understandable and expressible fault repair plan, which is beneficial to providing more intuitive and accurate fault repair guidance for fault repair personnel and improving the fault repair efficiency.

[0126] As Figure 5The figure shows a schematic flow diagram of a fault repair method in an embodiment. The fault repair method includes three key steps: fault discovery, fault diagnosis, and fault repair.

[0127] Fault discovery is based on the feedback of fault repair requests, and real-time obtains the vehicle's performance, real-time signal data, and signal data related to fault repair. Fault diagnosis is to perform fault diagnosis based on a trained fault classification model to obtain multiple repair guidance messages. Fault repair refers to that the repair plan generation model intelligently generates a recommended fault repair plan with higher adaptability to the current fault according to multiple repair guidance messages, and performs fault repair based on the recommended fault repair plan, which improves the repair success rate, repair efficiency, and repair effect.

[0128] The specific implementation process is as follows:

[0129] 1. The cloud server obtains the fault data uploaded by the vehicle when a fault occurs. The fault data includes the real-time signal data collected when the vehicle has a fault, the time when the vehicle has a fault, and the device identifier of the vehicle, and stores the corresponding relationship of the real-time signal data, time, and device identifier in the signal database. The signal data can include: the Chinese and English names of the fault, the fault explanation, the fault location, the fault degradation action, the fault precondition, the fault recovery condition, and the basic signals that need to be queried for this fault.

[0130] 2. Obtain real-time signal data: By real-time monitoring devices such as vehicle sensors, battery management systems, and charging systems, collect the vehicle's operating state data, including data such as speed, acceleration, temperature, voltage, and current.

[0131] 3. Data cleaning and preprocessing: Clean and preprocess the collected real-time signal data, remove abnormal data and noise data, and perform normalization processing on the data.

[0132] 4. Record high-frequency query signals, make buried points at the front end of the system used by the user, and record the target signal data queried when querying diagnostic fault codes.

[0133] 5. Perform database storage processing on the target signal data, and extract features from the target signal data, and extract key indicators for problem solving, such as information such as the threshold of the signal.

[0134] 6. Train a large number of processing results to establish a fault repair model. The fault repair model includes a fault classification model and a repair plan generation model. The training samples of the fault classification model include historical fault description texts, historical fault occurrence times, and historical fault device identifiers, and the training labels include historical repair texts.

[0135] The fault classification model can be a decision tree model. A decision tree is a method of analysis based on known feature values by constructing a tree-shaped decision structure. Through training, a fault classification model including multiple decision trees is obtained, multiple maintenance guidance information is output, and the target guidance information is found.

[0136] In order to provide more detailed maintenance guidance to maintenance personnel, the target guidance information is output to the maintenance plan generation model. The maintenance plan generation model can be a GPT generation model to generate an understandable and expressible recommended fault maintenance plan.

[0137] In some embodiments, the cloud server receives the actual fault maintenance plan obtained based on the recommended fault maintenance plan, uses the actual fault maintenance plan as the training label of the fault classification model, and makes relevant corrections to the model parameters of the fault classification model to obtain an updated fault classification model, improving the model's processing ability and classification accuracy.

[0138] The above-mentioned fault maintenance method includes receiving a fault maintenance request, where the fault maintenance request includes the fault description information and the fault device identifier of the faulty device; the fault description information includes the fault occurrence time and the diagnostic fault code; according to the fault device identifier, query the signal database to obtain the target signal data associated with the fault occurrence time; the signal database stores the corresponding relationship between the fault device identifier, the fault occurrence time and the target signal data; input the obtained target signal data and the diagnostic fault code into the fault maintenance model, and the fault maintenance model outputs a recommended fault maintenance plan. This method of querying the target signal data associated with the fault occurrence time based on the fault description information and the fault device identifier of the faulty device, and inputting the target signal data and the diagnostic fault code in the fault description information into the fault maintenance model to obtain a recommended fault maintenance plan can perform fault diagnosis by combining the target signal data associated with the vehicle fault occurrence time and the diagnostic fault code. Compared with the method of performing fault diagnosis only based on the vehicle type and fault type, the input data of the model is more comprehensive, and the obtained recommended fault maintenance plan has a higher adaptability to the current fault. The above solution solves the problems of few fault cases and low reliability of fault maintenance plans for new vehicle models and vehicle manufacturers with few fault cases, and has high fault diagnosis and fault maintenance efficiency, high accuracy, and saves fault diagnosis and fault maintenance costs.

[0139] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear indication in this article, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0140] Based on the same inventive concept, an embodiment of the present application also provides a fault repair device for implementing the above-mentioned fault repair method. The solution provided by this device for solving problems is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the following fault repair devices can refer to the limitations on the fault repair method in the above text, and will not be repeated here.

[0141] In an exemplary embodiment, as Figure 6 shown, a fault repair device 100 is provided, including: a receiving module 120, a query module 140, and an output module 160, where:

[0142] The receiving module 120 is configured to receive a fault repair request, where the fault repair request includes fault description information of a faulty device and a faulty device identifier; the fault description information includes the fault occurrence time and a diagnostic fault code;

[0143] The query module 140 is configured to query a signal database according to the faulty device identifier to obtain target signal data associated with the fault occurrence time; the signal database stores the corresponding relationship between the faulty device identifier, the fault occurrence time, and the target signal data;

[0144] The output module 160 is configured to input the obtained target signal data and the diagnostic fault code into a fault repair model, and the fault repair model outputs a recommended fault repair solution.

[0145] The above-mentioned fault repair device receives a fault repair request, which includes fault description information of a faulty device and a faulty device identifier; the fault description information includes the fault occurrence time and a diagnostic fault code; according to the faulty device identifier, it queries a signal database to obtain target signal data associated with the fault occurrence time; the signal database stores the corresponding relationship between the faulty device identifier, the fault occurrence time and the target signal data; it inputs the obtained target signal data and the diagnostic fault code into a fault repair model, and the fault repair model outputs a recommended fault repair solution. This method of querying the target signal data associated with the fault occurrence time based on the fault description information of the faulty device and the faulty device identifier, and inputting the target signal data and the diagnostic fault code in the fault description information into the fault repair model to obtain a recommended fault repair solution can perform fault diagnosis by combining the target signal data associated with the vehicle fault occurrence time and the diagnostic fault code. Compared with the method of performing fault diagnosis only based on the vehicle type and fault type, the input data of the model is more comprehensive, and the obtained recommended fault repair solution has a higher adaptability to the current fault.

[0146] In one embodiment, the fault repair model includes a fault classification model and a repair solution generation model; when inputting the obtained target signal data and the diagnostic fault code into the fault repair model, and the fault repair model outputs a recommended fault repair solution, the output module 160 is further configured to: input the target signal data and the diagnostic fault code into the fault classification model to obtain multiple repair guidance information output by the fault classification model; input the multiple repair guidance information into the repair solution generation model to obtain the recommended fault repair solution output by the repair solution generation model.

[0147] In one embodiment, when inputting the multiple repair guidance information into the repair solution generation model to obtain the recommended fault repair solution output by the repair solution generation model, the output module 160 is further configured to: use the repair guidance information with the largest occupancy ratio among the multiple repair guidance information as the target guidance information; input the target guidance information into the repair solution generation model to obtain the recommended fault repair solution output by the repair solution generation model.

[0148] In one embodiment, in terms of the training of the fault classification model, the fault repair device 100 further includes a training module, and the training module is further configured to: obtain historical fault repair requests, where the historical fault repair requests include historical fault description information and historical faulty device identifiers; the historical fault description information includes the historical fault occurrence time and historical diagnostic fault codes; obtain historical repair texts corresponding to the historical fault repair requests; according to the historical faulty device identifiers, query the signal database to obtain historical signal data associated with the historical fault occurrence time; based on the historical diagnostic fault codes, historical signal data and historical repair texts, train an initial decision tree model to obtain the fault classification model.

[0149] In one embodiment, the initial decision tree model includes M initial decision trees, and the initial model parameters of the M initial decision trees are different from each other; based on historical diagnostic fault codes, historical signal data, and historical maintenance texts, the initial decision tree model is trained to obtain a fault classification model. The training module is further configured to: for each of the M initial decision trees, input the historical diagnostic fault codes and historical signal data into the initial decision tree to obtain the predicted maintenance text output by the initial decision tree; calculate the model loss of the initial decision tree based on the historical maintenance text and the predicted maintenance text; adjust the model parameters of the M initial decision trees respectively based on the model losses corresponding to the M initial decision trees until the M initial decision trees respectively meet the corresponding stop conditions, so as to obtain the fault classification model.

[0150] In one embodiment, the fault classification model includes M decision trees; the fault repair device 100 further includes an adjustment module, and the adjustment module is further configured to: receive the actual fault repair plan corresponding to the recommended fault repair plan; for each of the M decision trees, input the target signal data and the diagnostic fault code into the decision tree to obtain the updated predicted maintenance text output by the decision tree; calculate the updated model loss of the decision tree based on the actual fault repair plan and the updated predicted maintenance text; update the model parameters of the M decision trees respectively based on the updated model losses corresponding to the M decision trees to obtain the updated fault classification model.

[0151] In one embodiment, in terms of the construction of the signal database, the fault repair device 100 further includes a construction module, and the construction module is further configured to: obtain the fault data uploaded by the vehicle when a fault occurs, where the fault data includes the real-time signal data collected when the vehicle has a fault, the time when the vehicle has a fault, and the device identifier of the vehicle; store the corresponding relationship between the real-time signal data, the time, and the device identifier in the signal database.

[0152] In one embodiment, the fault description information further includes a fault description text; input the obtained target signal data and diagnostic fault code into the fault repair model, and the output module 160 is further configured to: input the fault description text, the target signal data, and the diagnostic fault code into the fault repair model, and output the recommended fault repair plan by the fault repair model.

[0153] Each module in the above-mentioned fault repair device can be implemented in whole or in part by software, hardware, and their combination. The above-mentioned modules can be embedded in the processor of the computer device in hardware form or be independent of it, or be stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to the above-mentioned modules.

[0154] In an exemplary embodiment, a computer device is provided. The computer device may be a cloud server, and its internal structure diagram may be as shown in Figure 7 . The computer device includes a processor, a memory, an input / output interface, and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the correspondence between the faulty device identifier, the fault occurrence time, and the target signal data. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner. The wireless manner can be implemented through WIFI, a mobile cellular network, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, a fault repair method is implemented.

[0155] Those skilled in the art can understand that Figure 7 the structure shown in is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0156] In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0157] In an embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by the processor, the steps in the above method embodiments are implemented.

[0158] In an embodiment, a computer program product is provided, including a computer program. When the computer program is executed by the processor, the steps in the above method embodiments are implemented.

[0159] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0160] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.

[0161] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0162] The above-described embodiments merely represent several implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A fault repair method, characterized in that, The method includes: Receiving a fault repair request, where the fault repair request includes fault description information of a faulty device and a faulty device identifier; the fault description information includes the fault occurrence time and a diagnostic fault code; Querying a signal database according to the faulty device identifier to obtain target signal data associated with the fault occurrence time; the signal database stores the corresponding relationship between the faulty device identifier, the fault occurrence time, and the target signal data; Inputting the obtained target signal data and the diagnostic fault code into a fault repair model, and outputting a recommended fault repair plan by the fault repair model.

2. The method according to claim 1, wherein The fault repair model includes a fault classification model and a repair plan generation model; the step of inputting the obtained target signal data and the diagnostic fault code into the fault repair model and outputting a recommended fault repair plan by the fault repair model includes: Inputting the target signal data and the diagnostic fault code into the fault classification model to obtain multiple maintenance guidance information output by the fault classification model; Inputting the multiple maintenance guidance information into the repair plan generation model to obtain a recommended fault repair plan output by the repair plan generation model.

3. The method according to claim 2, characterized in that, The step of inputting the multiple maintenance guidance information into the repair plan generation model to obtain a recommended fault repair plan output by the repair plan generation model includes: Taking the maintenance guidance information with the largest occupancy ratio among the multiple maintenance guidance information as the target guidance information; inputting the target guidance information into the repair plan generation model to obtain a recommended fault repair plan output by the repair plan generation model.

4. The method according to claim 2, characterized in that The training steps of the fault classification model include: Obtaining historical fault repair requests, where the historical fault repair requests include historical fault description information and historical faulty device identifiers; the historical fault description information includes the historical fault occurrence time and historical diagnostic fault codes; Obtaining historical repair texts corresponding to the historical fault repair requests; Querying the signal database according to the historical faulty device identifier to obtain historical signal data associated with the historical fault occurrence time; Training an initial decision tree model based on the historical diagnostic fault code, the historical signal data, and the historical repair text to obtain a fault classification model.

5. The method according to claim 4, wherein The initial decision tree model includes M initial decision trees, and the initial model parameters of the M initial decision trees are different; the step of training the initial decision tree model based on the historical diagnostic fault code, the historical signal data, and the historical repair text to obtain a fault classification model includes: For each of the M initial decision trees, inputting the historical diagnostic fault code and the historical signal data into the initial decision tree to obtain a predicted repair text output by the initial decision tree; calculating a model loss of the initial decision tree based on the historical repair text and the predicted repair text; Adjusting the model parameters of the M initial decision trees respectively based on the model losses corresponding to the M initial decision trees until the M initial decision trees respectively meet the corresponding stopping conditions to obtain a fault classification model.

6. The method according to claim 2, characterized in that, The fault classification model includes M decision trees; the method further includes: Receiving the actual fault repair plan corresponding to the recommended fault repair plan; For each of the M decision trees, inputting the target signal data and the diagnostic fault code into the decision tree to obtain the updated predicted repair text output by the decision tree; calculating the updated model loss of the decision tree based on the actual fault repair plan and the updated predicted repair text; Based on the updated model losses corresponding to the M decision trees respectively, updating the model parameters of the M decision trees to obtain an updated fault classification model.

7. The method according to claim 1, characterized in that, The steps for constructing the signal database include: Obtaining the fault data uploaded by the vehicle when a fault occurs, where the fault data includes the real-time signal data collected when the vehicle has a fault, the time when the vehicle has a fault, and the device identifier of the vehicle; Storing the corresponding relationship among the real-time signal data, the time, and the device identifier in the signal database.

8. The method according to claim 1, characterized in that The fault description information further includes a fault description text; the step of inputting the obtained target signal data and the diagnostic fault code into the fault repair model and outputting a recommended fault repair plan by the fault repair model includes: Inputting the fault description text, the target signal data, and the diagnostic fault code into the fault repair model and outputting a recommended fault repair plan by the fault repair model.

9. A fault repair device, characterized in that, The device includes: A receiving module, configured to receive a fault repair request, where the fault repair request includes the fault description information of the faulty device and the faulty device identifier; the fault description information includes the fault occurrence time and the diagnostic fault code; A query module, configured to query the signal database according to the faulty device identifier to obtain the target signal data associated with the fault occurrence time; the signal database stores the corresponding relationship among the faulty device identifier, the fault occurrence time, and the target signal data; An output module, configured to input the obtained target signal data and the diagnostic fault code into the fault repair model and output a recommended fault repair plan by the fault repair model.

10. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 8.

12. A computer program product, comprising a computer program, characterized in that, When this computer program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 8.

Images