Vehicle maintenance information pushing method and device, electronic equipment and readable storage medium

Abstract

The invention provides a vehicle maintenance information pushing method and device, electronic equipment and a readable storage medium, relates to the technical field of computers, and aims to solve the technical problem that a vehicle maintenance process depends on experience of a maintainer. The pushing method comprises the steps of obtaining a maintenance record database, obtaining a plurality of historical maintenance work orders from the maintenance record database, performing vectorization processing on a vehicle model, a mileage and a fault type for each historical maintenance work order to obtain a first vector, and establishing a corresponding relationship between the first vector and maintenance information; obtaining fault description information of the to-be-maintained vehicle, determining a fault type of the to-be-maintained vehicle, and carrying out vectorization processing on the vehicle type, the first mileage and the fault type of the to-be-maintained vehicle to obtain a second vector; performing similarity calculation on the second vector and each first vector, and determining similar vectors; and pushing maintenance information corresponding to the similar vector according to the corresponding relationship. The method can overcome the problem that the user returns to the factory for many times.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a vehicle maintenance information push method, device, electronic device and readable storage medium. Background Art

[0002] When users need to repair their vehicles, the repairmen usually replace the vehicle parts based on their own experience. However, due to the different models, the repair data is very large and errors are inevitable, such as taking the wrong parts or not replacing the parts that need to be replaced, etc., which causes users to return to the factory many times and low repair efficiency. Summary of the invention

[0003] The present invention aims to solve or improve the technical problem in the related art that the vehicle maintenance process relies on the experience of the maintenance technician, resulting in multiple returns to the factory.

[0004] A first aspect of the present invention provides a vehicle maintenance information push method.

[0005] A second aspect of the present invention provides a vehicle maintenance information push device.

[0006] A third aspect of the present invention provides an electronic device.

[0007] A fourth aspect of the present invention provides a readable storage medium.

[0008] The vehicle maintenance information push method provided by the present invention includes: obtaining a maintenance record database, and obtaining multiple historical maintenance work orders from the maintenance record database, wherein each historical maintenance work order includes a vehicle model, mileage, fault type and maintenance information; for each historical maintenance work order, vectorizing the vehicle model, mileage, and fault type to obtain a first vector, and establishing a corresponding relationship between the first vector and the maintenance information, so that for multiple historical maintenance work orders, multiple first vectors and multiple corresponding relationships can be obtained; obtaining fault description information of a vehicle to be repaired; analyzing the fault description information, and determining the fault type of the vehicle to be repaired; obtaining the vehicle model and the first mileage of the vehicle to be repaired; vectorizing the vehicle model, the first mileage and the fault type of the vehicle to be repaired to obtain a second vector; calculating the similarity between the second vector and each first vector respectively; sorting the multiple first vectors in order of similarity from large to small, and recording the first preset number of first vectors as similar vectors; and pushing the maintenance information corresponding to the similar vectors according to the corresponding relationship.

[0009] The vehicle maintenance information push method provided by the present invention comprises: obtaining a maintenance record database, and obtaining a plurality of historical maintenance work orders from the maintenance record database, wherein each historical maintenance work order comprises a vehicle model, mileage, fault type and maintenance information, wherein the maintenance information also refers to information in the maintenance process of the vehicle fault, such as maintenance man-hours, maintenance tools, maintenance parts name, maintenance parts model, maintenance steps, etc.; then for each historical maintenance work order, the vehicle model, mileage, and fault type are vectorized by a vector processing model to obtain a first vector, and a corresponding relationship between the first vector and the maintenance information is established, so that for a plurality of historical maintenance work orders, a plurality of first vectors and a plurality of corresponding relationships can be obtained, and then the vehicle to be repaired is obtained. Fault description information, analyze the fault description information, and determine the fault type of the vehicle to be repaired; in the process of analyzing the fault description information, for example, keyword technology can be used to determine the fault type of the vehicle to be repaired, and then obtain the vehicle model and the first mileage of the vehicle to be repaired; the vehicle model, the first mileage and the fault type of the vehicle to be repaired are vectorized to obtain a second vector; the second vector and each first vector are respectively calculated for similarity; multiple first vectors are sorted in order of similarity from large to small, and the greater the similarity, the more similar the vehicle model, mileage and fault type are to the current fault, and the first preset number of first vectors are recorded as similar vectors; then, according to the corresponding relationship, the maintenance information corresponding to the similar vector is pushed. The present invention is based on historical maintenance work orders, and each historical maintenance work order is vectorized, and then according to the vehicle model, mileage and fault type of the vehicle to be repaired, the most similar historical maintenance work order to this time is determined, so as to match the corresponding maintenance information, so that the maintenance master can use the pushed maintenance information as a reference, so that there will be no problems such as taking the wrong parts and not replacing the parts that need to be replaced, thereby improving the maintenance efficiency and overcoming the problem of users returning to the factory many times.

[0010] In addition, it should be noted that the reason why the present invention uses vehicle model, mileage and fault type as the basis for promotion is that for vehicles of the same model, mileage has a certain impact on the promotion information. For example, abnormal engine noise. If the mileage is short, within 50,000 kilometers, then it may be caused by insufficient engine oil. If the mileage is long, more than 150,000 kilometers, then it is necessary to significantly replace engine-related parts, such as bearings, valves, etc. Therefore, for different mileages, the corresponding pushed maintenance information is also different.

[0011] In some technical solutions, optionally, the preset number may be greater than or equal to 5 and less than or equal to 10. For example, 5.

[0012] In some technical solutions, optionally, the vehicle maintenance information push method also includes: after obtaining the input of first maintenance information for the vehicle to be repaired; forming a first maintenance work order based on the first maintenance information, the model of the vehicle to be repaired, the first mileage and the fault type; inputting the first maintenance work order into the maintenance record database to update the maintenance record database.

[0013] In this technical solution, for the fault of the vehicle to be repaired, the maintenance master can perform maintenance based on his own experience, or he can repair the vehicle according to the pushed maintenance information, and then the maintenance master can feedback the actual maintenance information, wherein the first maintenance information is the maintenance information input by the maintenance master for the fault of the vehicle to be repaired, so that the device can form a first maintenance work order based on the first maintenance information, the model of the vehicle to be repaired, the first mileage and the fault type. The first maintenance work order can also be understood as the maintenance work order for this maintenance, and the first maintenance work order is input into the maintenance record database to update the maintenance record database. The present invention can input all records of this maintenance into the maintenance record database in the form of a work order to update the maintenance record database, so that after each maintenance, the device can be gradually improved.

[0014] In some technical solutions, optionally, the vehicle maintenance information push method further includes: obtaining an evaluation score for the pushed maintenance information; and associating the evaluation score with a corresponding similarity vector.

[0015] In this embodiment, the maintenance technician can score the pushed maintenance information, and then the vehicle maintenance information push device can associate the evaluation score with the corresponding similarity vector, so that in the subsequent maintenance process, the user can intuitively understand the accuracy of the maintenance information corresponding to each similarity vector.

[0016] In some technical schemes, optionally, the vehicle maintenance information push method also includes: based on the model of the vehicle to be repaired and the first mileage, extracting a second maintenance work order from multiple historical maintenance work orders, the model of which is the same as the model of the vehicle to be repaired and the mileage is a second mileage, wherein the second mileage is greater than the first mileage; extracting the fault type and corresponding maintenance information in the second maintenance work order; using the fault type and corresponding maintenance information in the second maintenance work order as potential risk information, and pushing the potential risk information.

[0017] In this embodiment, the present invention can also achieve early prediction of fault trends based on mileage. The difference between the second mileage and the first mileage can be greater than or equal to 10,000 km and less than or equal to 30,000 km, for example, 20,000 km. In this way, based on the current mileage of the vehicle to be repaired, the possible fault risk after 20,000 km can be predicted in advance.

[0018] In some technical solutions, optionally, the maintenance information includes one or a combination of the following: maintenance man-hours, maintenance tools, maintenance part name, maintenance part model, and maintenance steps.

[0019] In this embodiment, based on the pushed maintenance information, the maintenance technician can intuitively understand the maintenance tools, maintenance parts names, maintenance parts models and other information, and there will be no problems such as taking the wrong parts or not replacing the parts that need to be replaced. This improves the maintenance efficiency and overcomes the problem of users returning to the factory multiple times.

[0020] In some technical solutions, optionally, when the maintenance information includes maintenance steps, the step of pushing the maintenance information corresponding to the similarity vector specifically includes: displaying the maintenance steps corresponding to the similarity vector by means of video.

[0021] In this embodiment, the maintenance steps are displayed by means of video, so that the car owner can easily perform the maintenance, thereby improving the maintenance efficiency.

[0022] In some technical solutions, optionally, the vehicle maintenance information push device can exchange information with the accessories purchasing device. When the maintenance information includes the maintenance part name and the maintenance part model, the vehicle maintenance information push method also includes: after receiving the purchase instruction, sending the maintenance part name and the maintenance part model to the accessories purchasing device.

[0023] In this embodiment, after receiving the purchase instruction, the repair part name and the repair part model can be directly sent to the parts purchasing device, thereby realizing automatic purchase and improving maintenance efficiency.

[0024] The second aspect of the present invention provides a vehicle maintenance information push device, comprising: a data acquisition unit, used to acquire a maintenance record database, and acquire multiple historical maintenance work orders from the maintenance record database, wherein each historical maintenance work order includes a vehicle model, mileage, fault type and maintenance information; a vector conversion unit, used to vectorize the vehicle model, mileage, and fault type for each historical maintenance work order to obtain a first vector, and establish a corresponding relationship between the first vector and the maintenance information, so that for multiple historical maintenance work orders, multiple first vectors and multiple corresponding relationships can be obtained; the data acquisition unit is also used to acquire Fault description information of a vehicle to be repaired; a determination unit, used to analyze the fault description information and determine the fault type of the vehicle to be repaired; the data acquisition unit is also used to obtain the model and first mileage of the vehicle to be repaired; the determination unit is also used to vectorize the model, first mileage and fault type of the vehicle to be repaired to obtain a second vector; respectively calculate the similarity between the second vector and each first vector; sort the multiple first vectors in descending order of similarity, and record the first preset number of first vectors as similar vectors; a push unit, used to push the maintenance information corresponding to the similar vectors according to the corresponding relationship.

[0025] The vehicle maintenance information push device provided by the present invention can implement the vehicle maintenance information push method of any technical solution of the first aspect of the present invention, and therefore has all the beneficial effects of the vehicle maintenance information push method.

[0026] The third aspect of the present invention provides an electronic device, comprising: a memory and a processor, the memory storing a program, and the processor implementing the steps of the vehicle maintenance information push method as the technical solution of the first aspect of the present application when executing the program.

[0027] The electronic device provided by the present invention can implement the vehicle maintenance information push method of any technical solution of the first aspect of the present invention, and therefore has all the beneficial effects of the vehicle maintenance information push method.

[0028] The fourth aspect of the present invention provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the vehicle maintenance information push method as described in the technical solution of the first aspect of the present application are implemented.

[0029] The readable storage medium provided by the present invention can implement the vehicle maintenance information push method of any technical solution of the first aspect of the present invention, and therefore has all the beneficial effects of the vehicle maintenance information push method.

[0030] The vehicle maintenance information push method proposed in the present invention has achieved significant improvement in technical effect, which is mainly reflected in the following aspects: 1. Improve the accuracy of maintenance push notifications: This invention uses a large model to deeply reason and analyze vehicle maintenance data, and can accurately identify common maintenance reasons and locations for different brands and models in various mileage intervals. Combined with vectorized matching technology, the system can intelligently push the maintenance content that best matches the vehicle information input by the user, greatly improving the accuracy and pertinence of maintenance push notifications, and avoiding push errors caused by fixed rules or simple statistical analysis in traditional systems.

[0031] 2. Optimize data utilization efficiency: Traditional maintenance push systems are difficult to effectively process massive, multi-dimensional maintenance data. This invention significantly improves data storage and query efficiency through vector processing and the application of vector database. Efficient data processing capabilities not only speed up the system's response speed, but also improve the accuracy of matching results, providing technical support for large-scale data analysis and intelligent push.

[0032] 3. Enhance the learning and adaptive capabilities of the system: The large model in the present invention can continuously learn and update. By continuously acquiring and analyzing new maintenance data, the system self-optimizes its reasoning ability and push algorithm to ensure that the push results are continuously improved as the data accumulates. Compared with the traditional static rule system, the present invention has stronger adaptive capabilities, can dynamically respond to the changing needs of different brands, models and usage environments, and continuously improve service quality.

[0033] 4. Unified processing of multi-brand and multi-model data: In view of the heterogeneity of maintenance data of different brands and models, the present invention realizes unified processing of multi-brand and multi-model data through vectorized representation. The vectorized model can overcome the differences in data sources and realize unified analysis and push of various types of data, solving the shortcomings of traditional systems in data integration and providing the possibility for wide application in various vehicle after-sales service scenarios.

[0034] 5. Improve user after-sales service experience: Through intelligent maintenance push and fault prediction, the present invention can significantly improve the quality and efficiency of vehicle after-sales service and provide users with more professional and reliable maintenance suggestions. This innovative technical effect not only helps users save maintenance time and costs, but also enhances their trust and satisfaction with after-sales service.

[0035] In summary, the present invention overcomes the limitations of traditional systems through advanced AI (Artificial Intelligence) reasoning and vectorization technology, brings significant technological improvements to the field of vehicle after-sales service, and has important practical value and broad application prospects.

[0036] Additional aspects and advantages of the present invention will become apparent in the following description or may be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The above and / or additional aspects and advantages of the embodiments of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which: Figure 1 One of the flow charts of the vehicle maintenance information push method provided by an embodiment of the present invention is shown; Figure 2 The second flowchart of the vehicle maintenance information push method provided by the embodiment of the present invention is shown; Figure 3 A corresponding diagram showing the relationship between the execution subject and steps of the vehicle maintenance information push method provided by an embodiment of the present invention; Figure 4 A structural block diagram of a vehicle maintenance information push device provided by an embodiment of the present invention is shown; Figure 5 A structural block diagram of an electronic device provided by an embodiment of the present invention is shown.

[0038] in, Figure 4 and Figure 5 The corresponding relationship between the component names and numbers in is as follows: 1 a vehicle maintenance information push device, 12 a data acquisition unit, 14 a vector conversion unit, 16 a determination unit, 18 a push unit, 2 an electronic device, 22 a memory, 24 a processor. DETAILED DESCRIPTION

[0039] In order to more clearly understand the above aspects, features and advantages of the embodiments of the present invention, the embodiments of the present invention are optionally described in detail in conjunction with the accompanying drawings and specific implementation methods. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0040] In the following description, many specific details are set forth to facilitate a full understanding of the embodiments according to the present invention. However, the embodiments according to the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the embodiments according to the present invention is not limited to the specific embodiments disclosed below.

[0041] First of all, it is important to understand that with the rapid development and popularization of the automotive industry, the types and number of vehicles are increasing, and the resulting vehicle maintenance data is also increasing. In particular, a large amount of complex and diverse maintenance data is generated in different brands and models and in different mileage ranges. Existing maintenance push systems usually rely on rule-based logic or simple statistical analysis methods, which often fail to fully explore the potential information and patterns in the data when faced with massive and multi-dimensional maintenance data.

[0042] Due to the limitations of algorithms, traditional maintenance push systems are unable to accurately push maintenance content, let alone predict the potential failure trends of vehicles in the future. This method relies on manual experience or static rules and lacks deep learning and intelligent reasoning capabilities for data, resulting in inaccurate push results and low maintenance efficiency, which in turn affects users' after-sales service experience.

[0043] Therefore, the market urgently needs a new intelligent method that can make full use of the current growing vehicle maintenance data and combine it with advanced artificial intelligence technology to conduct in-depth reasoning and analysis of the data. By introducing a large model to learn vehicle maintenance data and combining it with a vector model to accurately match the cause of vehicle failure, the appropriate maintenance content, including repair parts and working hours, can be intelligently pushed. This new method can not only improve the accuracy and efficiency of maintenance services, but also significantly improve the quality of after-sales services, bringing significant innovation and value enhancement to the vehicle maintenance industry.

[0044] like Figure 1 As shown, the first aspect of the present invention provides a method for pushing vehicle maintenance information, including: S102: Obtain a maintenance record database, and obtain multiple historical maintenance work orders from the maintenance record database, wherein each historical maintenance work order includes vehicle model, mileage, fault type and maintenance information; S104: for each historical maintenance work order, vector processing is performed on the vehicle model, mileage, and fault type to obtain a first vector, and a corresponding relationship between the first vector and the maintenance information is established, so that for multiple historical maintenance work orders, multiple first vectors and multiple corresponding relationships can be obtained; S106: Obtaining fault description information of the vehicle to be repaired; S108: Analyze the fault description information and determine the fault type of the vehicle to be repaired; S110: Acquire the model and first mileage of the vehicle to be repaired; perform vector processing on the model, first mileage and fault type of the vehicle to be repaired to obtain a second vector; S112: Calculate the similarity between the second vector and each first vector respectively, sort the multiple first vectors in descending order of similarity, and record the first preset number of first vectors as similar vectors; S114: Pushing maintenance information corresponding to the similar vectors according to the corresponding relationship.

[0045] The vehicle maintenance information push method provided by the present invention comprises: obtaining a maintenance record database, and obtaining a plurality of historical maintenance work orders from the maintenance record database, wherein each historical maintenance work order comprises a vehicle model, mileage, fault type and maintenance information, wherein the maintenance information also refers to information in the maintenance process of a vehicle fault, such as maintenance man-hours, maintenance tools, maintenance parts name, maintenance parts model, maintenance steps, etc.; then for each historical maintenance work order, the vehicle model, mileage, and fault type are vectorized by a vector processing model to obtain a first vector, and a corresponding relationship between the first vector and the maintenance information is established, so that for a plurality of historical maintenance work orders, a plurality of first vectors and Multiple corresponding relationships, then obtain the fault description information of the vehicle to be repaired, analyze the fault description information, and determine the fault type of the vehicle to be repaired; in the process of analyzing the fault description information, for example, keyword technology can be adopted to determine the fault type of the vehicle to be repaired, and then obtain the model and first mileage of the vehicle to be repaired; the model, first mileage and fault type of the vehicle to be repaired are vectorized to obtain the second vector; the similarity between the second vector and each first vector is calculated respectively; according to the order of similarity from large to small, multiple first vectors are sorted, and the first preset number of first vectors are recorded as similar vectors; then according to the corresponding relationship, the maintenance information corresponding to the similar vector is pushed. The present invention is based on historical maintenance work orders, and each historical maintenance work order is vectorized, and then according to the model, mileage and fault type of the vehicle to be repaired, the most similar historical maintenance work order is determined to match the corresponding maintenance information, so that the maintenance master can use the pushed maintenance information as a reference, so that there will be no problems such as taking the wrong parts and not replacing the parts that need to be replaced, which improves the maintenance efficiency and overcomes the problem of users returning to the factory many times.

[0046] In some technical solutions, optionally, the preset number may be greater than or equal to 5 and less than or equal to 10. For example, 5.

[0047] In some technical solutions, optionally, the vehicle maintenance information push method also includes: after obtaining the input of first maintenance information for the vehicle to be repaired; forming a first maintenance work order based on the first maintenance information, the model of the vehicle to be repaired, the first mileage and the fault type; inputting the first maintenance work order into the maintenance record database to update the maintenance record database.

[0048] In this technical solution, for the fault of the vehicle to be repaired, the maintenance master can perform maintenance based on his own experience, or he can repair the vehicle according to the pushed maintenance information, and then the maintenance master can feedback the actual maintenance information, wherein the first maintenance information is the maintenance information input by the maintenance master for the fault of the vehicle to be repaired, so that the device can form a first maintenance work order based on the first maintenance information, the model of the vehicle to be repaired, the first mileage and the fault type, and input it into the maintenance record database to update the maintenance record database. The present invention can input all records of this maintenance into the maintenance record database in the form of a work order to update the maintenance record database, so that after each maintenance, the device can be gradually improved.

[0049] In some technical solutions, optionally, the vehicle maintenance information push method further includes: obtaining an evaluation score for the pushed maintenance information; and associating the evaluation score with a corresponding similarity vector.

[0050] In this embodiment, the maintenance technician can score the pushed maintenance information, and then the device can associate the evaluation score with the corresponding similarity vector, so that the user can intuitively understand the accuracy of the maintenance information corresponding to each similarity vector.

[0051] In some technical schemes, optionally, the vehicle maintenance information push method also includes: based on the model of the vehicle to be repaired and the first mileage, extracting a second maintenance work order from multiple historical maintenance work orders, the model of which is the same as the model of the vehicle to be repaired and the mileage is a second mileage, wherein the second mileage is greater than the first mileage; extracting the fault type and corresponding maintenance information in the second maintenance work order; using the fault type and corresponding maintenance information in the second maintenance work order as potential risk information, and pushing the potential risk information.

[0052] In this embodiment, the present invention can also achieve early prediction of fault trends based on mileage. The difference between the second mileage and the first mileage can be greater than or equal to 10,000 km and less than or equal to 30,000 km, for example, 20,000 km. In this way, based on the current mileage of the vehicle to be repaired, the possible fault risk after 20,000 km can be predicted in advance.

[0053] In some technical solutions, optionally, the maintenance information includes one or a combination of the following: maintenance man-hours, maintenance tools, maintenance part name, maintenance part model, and maintenance steps.

[0054] In this embodiment, based on the pushed maintenance information, the maintenance technician can intuitively understand the maintenance tools, maintenance parts names, maintenance parts models and other information, and there will be no problems such as taking the wrong parts or not replacing the parts that need to be replaced. This improves the maintenance efficiency and overcomes the problem of users returning to the factory multiple times.

[0055] In some technical solutions, optionally, when the maintenance information includes maintenance steps, the step of pushing the maintenance information corresponding to the similarity vector specifically includes: displaying the maintenance steps corresponding to the similarity vector by means of video.

[0056] In this embodiment, the maintenance steps are displayed by means of video, so that the car owner can easily perform the maintenance, thereby improving the maintenance efficiency.

[0057] In some technical solutions, optionally, the vehicle maintenance information push device can exchange information with the accessories purchasing device. When the maintenance information includes the maintenance part name and the maintenance part model, the vehicle maintenance information push method also includes: after receiving the purchase instruction, sending the maintenance part name and the maintenance part model to the accessories purchasing device.

[0058] In this embodiment, after receiving the purchase instruction, the repair part name and the repair part model can be directly sent to the parts purchasing device, thereby realizing automatic purchase and improving maintenance efficiency.

[0059] Another embodiment of the present invention provides a method for pushing vehicle maintenance information, aiming to solve the following problems: 1. Efficient use of massive data: With the increase in the number of cars and the diversification of car models, the maintenance data generated by vehicles at different stages of use and mileage intervals has increased exponentially. Existing systems cannot effectively use these complex, multi-dimensional data for in-depth analysis, and it is difficult to extract valuable maintenance and failure modes from them, thus affecting the accuracy and timeliness of maintenance decisions.

[0060] 2. Limitations of traditional push systems: Most current maintenance push systems are based on predefined rules or simple statistical methods, which perform poorly when processing unstructured and diverse maintenance data, and can easily lead to inaccurate push results. Traditional systems cannot dynamically adapt to changes in different brands, models, and usage environments, resulting in a lack of pertinence and effectiveness in maintenance push content.

[0061] 3. Lack of intelligent reasoning capabilities: Traditional push systems lack advanced reasoning and learning capabilities and are unable to intelligently reason on large amounts of maintenance data, identify potential causes of failures, and make accurate matches. With the development of AI technology, the market needs a new system that can deeply learn and intelligently analyze data to provide more insightful maintenance recommendations and fault predictions.

[0062] 4. It is difficult to unify data processing for different brands and models: Vehicles of different brands and models have significant differences in design, usage habits, maintenance specifications, etc., resulting in high heterogeneity of maintenance data. Existing systems are difficult to uniformly process and analyze these heterogeneous data, which affects the comprehensiveness and accuracy of maintenance content push.

[0063] The present invention aims to solve the above problems by introducing large models and vectorized matching technology to perform intelligent reasoning and analysis on a large amount of maintenance data, thereby achieving accurate matching of vehicle faults and intelligent push of maintenance content, thereby improving the overall level of vehicle after-sales maintenance services.

[0064] like Figure 2 As shown, this embodiment provides a vehicle maintenance information push method, including: S202: Data collection and preprocessing; Specifically, the system collects data from a large number of vehicle maintenance work orders, including information such as vehicle brand, model, mileage, maintenance reason, maintenance part, and maintenance content. By cleaning, normalizing, and structuring the collected data, data noise and redundancy are eliminated, ensuring data accuracy and consistency, laying the foundation for subsequent analysis and reasoning.

[0065] S204: Large model reasoning; Specifically, a pre-trained large language model is used to perform reasoning analysis on the pre-processed maintenance data. The large model uses deep learning of various maintenance data to infer the main maintenance reasons, maintenance parts, and required maintenance content for different brands and models in each mileage range. It further extracts and summarizes common maintenance modes and failure trends to provide data support for maintenance push and failure prediction.

[0066] S206: vectorized processing and storage; Specifically, the maintenance reasons, maintenance parts and other information obtained by large model reasoning are vectorized using a vector model, and each maintenance reason, part and other information is converted into a high-dimensional vector. These vectorized data are stored in a vector database for subsequent fast query and efficient matching. The use of a vector database can significantly improve the speed of data retrieval and the accuracy of matching.

[0067] S208: Intelligent matching and push.

[0068] Specifically, when a vehicle enters the store for repair, the system automatically obtains information such as the vehicle's brand, model, mileage, and preliminary repair reason. The system first uses a large model to pre-process this information, parse out the specific repair reason, and then uses a vector database to match the most relevant content in the historical repair records using similarity calculations, including repair plans, common fault locations, and pushed repair resources. Based on the matching results, the system extracts the corresponding repair resource push from the relational database, including the required repair time and accessories and detailed repair suggestions.

[0069] The push method of the present invention not only sorts the push results according to the matching accuracy, but also combines the use frequency of historical maintenance resources to ensure the accuracy and practicality of the push content. Through this intelligent push process, users can intuitively view relevant maintenance suggestions, so as to quickly and efficiently solve vehicle maintenance problems. This method not only improves the maintenance efficiency and accuracy, reduces the diagnosis and decision-making time, but also makes the maintenance service more convenient and efficient, significantly improves customer satisfaction and the scientific nature of maintenance decisions, and provides strong support for vehicle after-sales maintenance. The present invention combines large model reasoning with vectorized matching technology to intelligently process and analyze vehicle maintenance data, thereby realizing the accurate push of vehicle maintenance content. This technical solution overcomes the limitations of traditional maintenance systems that rely on rules and simple statistical analysis, significantly improves data utilization and push accuracy, and is widely applicable to automobile maintenance services, vehicle management, after-sales support for automobile companies and other fields, providing innovative solutions for intelligent after-sales services.

[0070] In actual operation, Figure 3 As shown, the user first inputs the fault cause of the new work order, and then the corresponding fault cause can be matched through the AI ​​model, and then the historical fault that is most similar to the current fault can be matched from the vector database through the vector processing model, and finally the recommended items and repair parts are displayed through the relational database merger and deduplication. After this maintenance, a new work order can be generated, and the fault cause in the work order can be analyzed, and the fault cause of this time can be stored in the vector database. At the same time, other information, such as maintenance time, maintenance parts, etc., can be stored in the relational database, thereby realizing the update of the vector database and the relational database. It can be seen that the method of the present invention can input the repair data of this time into the system after each use, thereby realizing the automation and improvement of the system.

[0071] In addition, in order to verify the effectiveness and reliability of the technical solution of the present invention, the following tests must be performed: (1) Data processing accuracy test criteria: Data cleaning and preprocessing accuracy: After cleaning, normalizing and structuring the actual vehicle maintenance work order data, check whether the data meets the expected format and quality standards. Use data quality indicators (such as missing value ratio, data consistency, and outlier detection) to evaluate the data processing effect to ensure that the data is complete, accurate, and structured.

[0072] Vectorization accuracy: Verify the training effect of the vectorization model to ensure that the maintenance reasons and fault information can be accurately converted into high-dimensional vectors. Evaluate the consistency between the vector representation and the actual maintenance record by calculating the similarity between the vectors. Perform sample comparison to check whether the vectorization results are consistent with the actual maintenance records.

[0073] (2) Large model reasoning effect test: Accuracy of inference results: Verify the reasoning results of the large language model used. Evaluate the accuracy and reliability of reasoning by comparing the maintenance reasons, maintenance parts, and maintenance contents inferred by the model with the actual maintenance records. Calculate the accuracy, precision, and recall of the reasoning results to ensure that the model's reasoning ability meets the requirements of actual applications.

[0074] (3) Intelligent matching and push effect inspection rules: Accuracy of push content: Verify through real vehicle maintenance cases, compare the maintenance content pushed by the system with the actual maintenance records, and evaluate the relevance and accuracy of the push. Calculate the accuracy, precision and recall rate of the push to evaluate the reliability and practicality of the push results.

[0075] User experience evaluation: Collect user feedback on the push system, especially the relevance, practicality and usability of the push results. Quantify user satisfaction and identify improvement needs through user satisfaction surveys and feedback analysis tools.

[0076] In order to verify the practical effect of the technical solution of the present invention, the following experiments were conducted: (1) Data processing experiment: Experimental design: Collect sample data from multiple vehicle maintenance data sources, perform data cleaning, normalization and structure processing, and select a certain amount of maintenance data for verification.

[0077] Experimental steps: Evaluate the quality of processed data and calculate data consistency and accuracy indicators. Compare the original data with the processed data to verify the effect of data preprocessing.

[0078] (2) Large model reasoning experiment: Experimental design: Use the large model to perform inference analysis on the processed maintenance data. Select sample data for inference and verify the accuracy of the inference results.

[0079] Experimental steps: Compare the model inference results with the actual maintenance records, and calculate the inference accuracy, precision, and recall. Verify the reliability of the inference results through manual review and data comparison.

[0080] (3) Intelligent matching and push experiment: Experimental design: simulate the actual scenario of vehicles entering the store for repair, test the system's processing and push effects on different vehicle information and repair reasons, and ensure that the system can provide accurate and practical repair suggestions. Use real vehicle repair work order data, including common faults and repair reasons of various brands, models and different mileage ranges, covering common and complex repair scenarios. The experiment is conducted in a simulated real maintenance environment to ensure the consistency of experimental conditions with actual application scenarios.

[0081] Experimental steps: 1. The system automatically obtains the vehicle brand, model, mileage and preliminary maintenance reasons, simulating the user's operation in actual maintenance. 2. The system uses a large model to pre-process the input information and parse out the specific maintenance reasons. 3. Call the vector database for similarity calculation to match the most relevant content in the historical maintenance records, such as maintenance plans, fault locations and push resources. Based on the matching results, extract the corresponding maintenance resource push from the relational database, including the required accessories, working hours and detailed maintenance suggestions. Sort the push content according to the matching accuracy and the frequency of use of historical maintenance resources to ensure the accuracy and practicality of the push results. 4. Collect feedback from maintenance technicians or users participating in the experiment to evaluate the applicability of the push content in actual maintenance. Use questionnaires and direct feedback to collect user satisfaction with the system's ease of use and push results. 5. Calculate the accuracy, precision and recall of the push content and analyze the push effect of the system. Statistics and analysis of user feedback are performed to evaluate the user experience and satisfaction of the system in actual applications.

[0082] Through the above tests and experimental verification, the technical solution of the present invention can ensure the expected effects of improving the accuracy of maintenance push, the accuracy of fault prediction, optimizing data utilization efficiency and improving user after-sales service experience in practical applications.

[0083] like Figure 4As shown, the second aspect of the present invention provides a vehicle maintenance information push device 1, comprising: a data acquisition unit 12, used to acquire a maintenance record database, and acquire multiple historical maintenance work orders from the maintenance record database, wherein each historical maintenance work order includes a vehicle model, mileage, fault type and maintenance information; a vector conversion unit 14, used to vectorize the vehicle model, mileage, and fault type for each historical maintenance work order to obtain a first vector, and establish a corresponding relationship between the first vector and the maintenance information, so that for multiple historical maintenance work orders, multiple first vectors and multiple corresponding relationships can be obtained; the data acquisition unit 12 is also used to obtain Fault description information of the vehicle to be repaired is obtained; a determination unit 16 is used to analyze the fault description information and determine the fault type of the vehicle to be repaired; the data acquisition unit 12 is also used to obtain the model and first mileage of the vehicle to be repaired; the determination unit 16 is also used to vectorize the model, first mileage and fault type of the vehicle to be repaired to obtain a second vector; similarity is calculated for the second vector and each first vector respectively; multiple first vectors are sorted in descending order of similarity, and the first preset number of first vectors are recorded as similar vectors; a push unit 18 is used to push maintenance information corresponding to the similar vectors according to the corresponding relationship.

[0084] The vehicle maintenance information push device 1 provided by the present invention can implement the vehicle maintenance information push method of any technical solution of the first aspect of the present invention, and therefore has all the beneficial effects of the vehicle maintenance information push method.

[0085] like Figure 5 As shown, the third aspect of the present invention provides an electronic device 2, including: a memory 22 and a processor 24, the memory 22 stores a program, and when the processor 24 executes the program, it implements the steps of the vehicle maintenance information push method as the technical solution of the first aspect of the present application.

[0086] The electronic device 2 provided by the present invention can implement the vehicle maintenance information push method of any technical solution of the first aspect of the present invention, and therefore has all the beneficial effects of the vehicle maintenance information push method.

[0087] The fourth aspect of the present invention provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the vehicle maintenance information push method as described in the technical solution of the first aspect of the present application are implemented.

[0088] The readable storage medium provided in this embodiment can implement the vehicle maintenance information push method of any technical solution of the first aspect of the present invention, and therefore has all the beneficial effects of the vehicle maintenance information push method.

[0089] The present invention relates to the field of data analysis and artificial intelligence, with a particular focus on the intelligent processing and push system of vehicle maintenance data. This technology uses vector matching and AI models to conduct in-depth analysis of historical vehicle fault data, maintenance records, mileage and other information, and realizes the intelligent push of vehicle maintenance resources.

[0090] The present invention is applicable to multiple scenarios such as automobile maintenance services, vehicle management, and after-sales support for automobile manufacturers, and can provide efficient decision support tools for vehicle maintenance service providers, fleet managers, and after-sales support departments of automobile manufacturers. By improving the accuracy of fault diagnosis and the efficiency of pushing maintenance solutions, the system helps to reduce maintenance costs, optimize maintenance processes, and improve the overall maintenance level of vehicles and user satisfaction. This technology is suitable for various enterprises and organizations that need to improve the level of intelligence of after-sales services and data-driven decision-making capabilities.

[0091] In the embodiments according to the present invention, the terms "first", "second", and "third" are only used for description and cannot be understood as indicating or implying relative importance; the term "plurality" refers to two or more, unless otherwise clearly defined. The terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense. For example, "connected" can be a fixed connection, a detachable connection, or an integral connection; "connected" can be a direct connection or an indirect connection through an intermediate medium. For those of ordinary skill in the art, the specific meanings of the above terms in the embodiments according to the present invention can be understood according to specific circumstances.

[0092] In addition, although each operation is described in a specific order, this should be understood as requiring such operation to be performed in a specific order or in a sequential order, or requiring that all illustrated operations should be performed to obtain desired results. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present invention. Some features described in the context of a separate embodiment can also be implemented in a single implementation in combination. On the contrary, the various features described in the context of a single implementation can also be implemented in multiple implementations individually or in any suitable sub-combination mode.

[0093] Although the subject matter has been described in the language of specific structural features and / or method logical actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing the claims.

[0094] The above are only preferred embodiments according to the embodiments of the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, various modifications and changes may be made according to the embodiments of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the embodiments of the present invention shall be included in the protection scope of the embodiments of the present invention.

Claims

1. A method for pushing vehicle maintenance information, characterized in that: include: Obtaining a maintenance record database, and obtaining a plurality of historical maintenance work orders from the maintenance record database, wherein each of the historical maintenance work orders includes vehicle model, mileage, fault type and maintenance information; For each of the historical maintenance work orders, vector processing is performed on the vehicle model, the mileage, and the fault type to obtain a first vector, and a corresponding relationship between the first vector and the maintenance information is established, so that for multiple historical maintenance work orders, multiple first vectors and multiple corresponding relationships can be obtained; Obtain fault description information of the vehicle to be repaired; Analyzing the fault description information and determining the fault type of the vehicle to be repaired; Obtaining the model and first mileage of the vehicle to be repaired; Performing vector processing on the model of the vehicle to be repaired, the first mileage, and the fault type to obtain a second vector; respectively calculating similarity between the second vector and each of the first vectors; Sort the first vectors in descending order of similarity, and record the first vectors of a preset number as similar vectors; According to the corresponding relationship, the maintenance information corresponding to the similarity vector is pushed.

2. The vehicle maintenance information push method according to claim 1, characterized in that: Also includes: After obtaining input of first maintenance information for the vehicle to be repaired; forming a first maintenance work order based on the first maintenance information, the model of the vehicle to be repaired, the first mileage, and the fault type; The first maintenance work order is input into the maintenance record database to update the maintenance record database.

3. The vehicle maintenance information push method according to claim 1, characterized in that: Also includes: Obtaining an evaluation score for the pushed maintenance information; The evaluation score is associated with the corresponding similarity vector.

4. The vehicle maintenance information push method according to claim 1, characterized in that: Also includes: Based on the model of the vehicle to be repaired and the first mileage, extracting a second repair work order of the same model as the vehicle to be repaired and having a second mileage from the plurality of historical repair work orders, wherein the second mileage is greater than the first mileage; Extracting the fault type and corresponding maintenance information in the second maintenance work order; The fault type and corresponding maintenance information in the second maintenance work order are used as potential risk information, and the potential risk information is pushed.

5. The vehicle maintenance information push method according to any one of claims 1 to 4, characterized in that: The maintenance information includes one or a combination of the following: Maintenance hours, maintenance tools, maintenance parts names, maintenance parts models, and maintenance steps.

6. The vehicle maintenance information push method according to claim 5, characterized in that: When the maintenance information includes the maintenance step, the step of pushing the maintenance information corresponding to the similarity vector specifically includes: The maintenance steps corresponding to the similarity vectors are displayed by means of video.

7. The vehicle maintenance information push method according to claim 5, characterized in that: The vehicle maintenance information push device can exchange information with the parts purchasing device. When the maintenance information includes the name of the maintenance part and the model of the maintenance part, the vehicle maintenance information push method further includes: After receiving the purchase instruction, the repair part name and the repair part model are sent to the parts purchasing device.

8. A vehicle maintenance information push device, characterized in that: include: A data acquisition unit, used to acquire a maintenance record database, and acquire a plurality of historical maintenance work orders from the maintenance record database, wherein each of the historical maintenance work orders includes a vehicle model, mileage, fault type and maintenance information; a vector conversion unit, configured to perform vector processing on the vehicle model, the mileage, and the fault type for each of the historical maintenance work orders to obtain a first vector, and establish a corresponding relationship between the first vector and the maintenance information, so that for multiple historical maintenance work orders, multiple first vectors and multiple corresponding relationships can be obtained; The data acquisition unit is also used to acquire fault description information of the vehicle to be repaired; A determination unit, configured to analyze the fault description information and determine the fault type of the vehicle to be repaired; The data acquisition unit is also used to acquire the model and first mileage of the vehicle to be repaired; The determining unit is further used to perform vector processing on the model of the vehicle to be repaired, the first mileage and the fault type to obtain a second vector; respectively calculating similarity between the second vector and each of the first vectors; Sort the first vectors in descending order of similarity, and record the first vectors of a preset number as similar vectors; A pushing unit is used to push the maintenance information corresponding to the similarity vector according to the corresponding relationship.

9. An electronic device, characterized in that: include: A memory and a processor, wherein the memory stores a program, and when the processor executes the program, the steps of the vehicle maintenance information push method as described in any one of claims 1 to 7 are implemented.

10. A readable storage medium, characterized in that: The readable storage medium stores a program or an instruction, and when the program or the instruction is executed by the processor, the steps of the vehicle maintenance information pushing method as described in any one of claims 1 to 7 are implemented.

Images