Vehicle fault early warning method and device, electronic equipment and storage medium

Abstract

This application discloses a vehicle fault early warning method, device, electronic device, and storage medium, applied to vehicle diagnostic equipment. The vehicle diagnostic equipment is equipped with a preset local AI diagnostic model, and a preset cloud server is equipped with a preset cloud AI diagnostic model. The method includes: acquiring target operating data of a target vehicle; processing the target operating data using the preset local AI diagnostic model to obtain a first diagnostic result; acquiring a first network status of the vehicle diagnostic equipment; uploading the target operating data to the preset cloud server when the first network status includes a normal network; processing the target operating data using the preset cloud AI diagnostic model to obtain a second diagnostic result; determining the target fault level based on the first and second diagnostic results; and executing a corresponding target early warning operation based on the target fault level. Using this application improves the reliability of vehicle fault early warning.

Description

Technical Field

[0001] This application relates to the field of vehicle fault warning technology, and in particular to a vehicle fault warning method, device, electronic device and storage medium. Background Technology

[0002] With the rapid development of automotive intelligence, the vehicle's onboard data collection capabilities have been continuously improved. Vehicle fault early warning technology has become an important means to ensure driving safety and improve the level of vehicle operation and maintenance intelligence. It can identify potential driving faults in advance and effectively avoid the safety risks caused by sudden vehicle failures.

[0003] Currently, traditional vehicle fault warning methods mostly adopt a single diagnostic architecture, relying solely on a simple on-board local model or a single cloud model to complete fault identification. This fails to combine the dual advantages of local real-time inference and cloud-based high-precision diagnosis, resulting in low diagnostic accuracy and poor real-time performance. Consequently, it is prone to misjudgment and missed judgment in fault warnings, leading to poor overall reliability of the warnings.

[0004] Therefore, improving the reliability of vehicle fault warnings has become an urgent problem to be solved. Summary of the Invention

[0005] This application provides a vehicle fault early warning method, device, electronic device, and storage medium, which improves the reliability of vehicle fault early warning.

[0006] In a first aspect, embodiments of this application provide a vehicle fault early warning method, characterized in that it is applied to a vehicle diagnostic device, wherein the vehicle diagnostic device is equipped with a preset local AI diagnostic model, the vehicle diagnostic device is connected to a preset cloud server, and the preset cloud server is equipped with a preset cloud AI diagnostic model; the method includes: Obtain the target vehicle's operational data; The target running data is processed by the preset local AI diagnostic model to obtain a first diagnostic result; Obtain the first network status of the vehicle diagnostic device; the first network status includes any one of the following: network normal, network abnormal; When the first network state includes a normal network, the target operation data is uploaded to the preset cloud server; the target operation data is processed by the preset cloud AI diagnostic model to obtain a second diagnostic result; Based on the first diagnostic result and the second diagnostic result, the target fault level is determined; Execute corresponding target warning operations according to the target fault level; the target alarm operations include at least one of the following: pop-up alarm, audible and visual alarm, and SMS alarm.

[0007] Secondly, embodiments of this application provide a vehicle fault early warning device applied to a vehicle diagnostic device. The vehicle diagnostic device includes a preset local AI diagnostic model and is connected to a preset cloud server, which in turn includes a preset cloud AI diagnostic model. The device comprises: an acquisition module, a first diagnostic module, a second diagnostic module, and an alarm module, wherein: The acquisition module is used to acquire the target operating data of the target vehicle; The first diagnostic module is used to process the target running data through the preset local AI diagnostic model to obtain a first diagnostic result; The acquisition module is further configured to acquire the first network status of the vehicle diagnostic device; the first network status includes any one of the following: network normal, network abnormal; The second diagnostic module is used to upload the target running data to the preset cloud server when the first network status includes a normal network; and to process the target running data through the preset cloud AI diagnostic model to obtain a second diagnostic result. The first diagnostic module is further configured to determine the target fault level based on the first diagnostic result and the second diagnostic result; The alarm module is used to perform corresponding target early warning operations according to the target fault level; the target alarm operations include at least one of the following: pop-up alarm, audible and visual alarm, and SMS alarm.

[0008] Thirdly, embodiments of this application provide an electronic device, including: a processor, a memory, a communication interface, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the programs include instructions for performing the steps in the first aspect of embodiments of this application.

[0009] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program for electronic data interchange, wherein the computer program causes a computer to perform some or all of the steps described in the first aspect of embodiments of this application.

[0010] Fifthly, embodiments of this application provide a computer program product, wherein the computer program product includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps described in the first aspect of embodiments of this application. The computer program product may be a software installation package.

[0011] Implementing this application will have the following beneficial effects: As can be seen, the vehicle fault warning method, device, electronic equipment, and storage medium described in this application complete real-time data analysis locally based on a local model (i.e., a preset local AI diagnostic model), quickly outputting the first diagnostic result to ensure the real-time nature of fault identification. Then, through network status determination logic, vehicle operation data is uploaded to the cloud when the network is normal. A second deep diagnosis is completed using a high-precision model in the cloud (i.e., a preset cloud AI diagnostic model), generating a second diagnostic result to compensate for the accuracy shortcomings of the local model. Based on this, the local and cloud-based dual-dimensional diagnostic results are integrated for mutual verification and correction, jointly determining the fault type and fault level. This effectively reduces the misjudgment and omission problems of single-model diagnosis, significantly lowering the probability of warning errors. Simultaneously, differentiated pop-up windows, sounds, and SMS-based graded warning methods are matched according to the fault type and level, avoiding excessive reminders for minor faults and providing strong warnings for high-risk faults. Furthermore, when the network is abnormal, diagnosis and warning can be completed independently using only the local model, ensuring continuous and stable service operation, thereby comprehensively improving the reliability of vehicle fault warnings. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of this application or the background art, the accompanying drawings used in the embodiments of this application or the background art will be described below.

[0013] Figure 1 This is a schematic diagram of the structure of a vehicle diagnostic device provided in an embodiment of this application; Figure 2 This is a schematic diagram of a vehicle diagnostic device provided in an embodiment of this application; Figure 3 This is a flowchart of a vehicle fault early warning method provided in an embodiment of this application; Figure 4 This is a flowchart of a method for determining a target fault level provided in an embodiment of this application; Figure 5 This is a flowchart illustrating a method for determining store push information provided in an embodiment of this application; Figure 6 This is a flowchart of a model training method provided in an embodiment of this application; Figure 7 This is a schematic diagram of the structure of a vehicle fault warning device provided in an embodiment of this application; Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0015] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0016] It should be understood that the term "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this document indicates that the preceding and following related objects are in an "or" relationship. In the embodiments of this application, "multiple" refers to two or more.

[0017] In the embodiments of this application, "at least one item" or its similar expression refers to any combination of these items, including any combination of a single item or a plurality of items. "One or more" means one or more, while "multiple" means two or more. For example, "at least one item" of a, b, or c can represent the following seven cases: a, b, c; a and b; a and c; b and c; a, b, and c. Each of a, b, and c can be an element or a set containing one or more elements.

[0018] In this application, the term "connection" refers to various connection methods, such as direct connection or indirect connection, to achieve communication between devices. This application does not impose any limitations on this.

[0019] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0020] The electronic devices described in this application embodiment may include smartphones (such as Android phones, iOS phones, Windows Phones, etc.), tablet computers, PDAs, laptops, video matrices, monitoring platforms, mobile internet devices (MIDs), or wearable devices, etc. The above are merely examples and not exhaustive, and include but are not limited to the above devices.

[0021] Of course, the aforementioned electronic devices can also be servers, such as cloud servers.

[0022] The following describes the relevant content, concepts, meanings, technical issues, technical solutions, and beneficial effects involved in the embodiments of this application.

[0023] First, let me explain some of the technical terms or phrases used in this application: Vehicle diagnostic equipment: refers to the hardware devices and supporting software systems used to detect and analyze the operating status of vehicle power systems, electronic control systems, chassis and other components, and to identify fault types and causes.

[0024] Confidence level: refers to the degree of credibility and probability of the model's output fault diagnosis results. It is used to characterize the reliability of the fault judgment results. The higher the value, the higher the accuracy of the corresponding fault judgment results. It can be used as an important basis for fault level classification and early warning triggering.

[0025] Knowledge distillation: a lightweight model optimization technique that uses a large, complex, and highly accurate model as a teacher model, and transfers the data features and fault identification knowledge learned by the teacher model to a small student model with a simple structure and few parameters. This significantly reduces the model size and computational overhead while maximizing the retention of the high-precision model's recognition capabilities, thus achieving lightweight model deployment.

[0026] Item-based collaborative filtering algorithm: an intelligent recommendation algorithm that uses the characteristics of the item itself as the matching basis. By calculating the similarity between the characteristics of the target demand and the characteristics of various items, it selects the target object with a higher feature matching degree.

[0027] Please see Figure 1 , Figure 1 This is a schematic diagram of the structure of a vehicle diagnostic device provided in an embodiment of this application. As can be seen, the vehicle diagnostic device may include: a data acquisition module, a control module, and a communication module. The control module may be equipped with a preset local AI diagnostic model. The data acquisition module is responsible for collecting various vehicle operation data such as driving conditions and operating parameters in real time, completing the acquisition and preliminary processing of raw data, and providing a basic data source for fault diagnosis.

[0028] The control module, as the core processing unit of the equipment, is equipped with a preset local AI diagnostic model. It performs calculations and analysis on the received vehicle operation data, completes local fault diagnosis and judgment, and coordinates the work of other modules to issue various control commands.

[0029] The communication module is responsible for establishing a data transmission channel between the vehicle diagnostic equipment and the preset cloud server. When the network is working properly, it uploads vehicle operation data and receives diagnostic data sent from the cloud, enabling information interaction between the local device and the cloud platform.

[0030] Please see Figure 2 , Figure 2 This is a schematic diagram of a vehicle diagnostic device provided in an embodiment of this application. As can be seen, the vehicle diagnostic device communicates bidirectionally with a preset cloud server and the target vehicle. The target vehicle transmits vehicle operation data to the vehicle diagnostic device. On the one hand, the vehicle diagnostic device analyzes the data in real time through a built-in preset local AI diagnostic model. On the other hand, when the network is normal, the device uploads the data to the preset cloud server. The preset cloud server completes in-depth diagnosis through the preset cloud AI diagnostic model and sends back the results. Finally, the vehicle diagnostic device integrates the local and cloud diagnostic results to achieve fault warning and graded reminder for the target vehicle.

[0031] Please see Figure 3 , Figure 3 This is a flowchart of a vehicle fault early warning method provided in an embodiment of this application; it is used in a vehicle diagnostic device, the vehicle diagnostic device is equipped with a preset local AI diagnostic model, the vehicle diagnostic device is connected to a preset cloud server, and the preset cloud server is equipped with a preset cloud AI diagnostic model; the method includes the following steps: S301. Obtain the target vehicle's target operating data.

[0032] In this application embodiment, the vehicle diagnostic equipment may include any of the following: in-vehicle intelligent OBD box, in-vehicle intelligent device (e.g., vehicle central control unit), driving recorder, vehicle diagnostic tool, in-vehicle T-BOX (remote communication terminal), etc., without limitation.

[0033] It should be explained that the vehicle diagnostic equipment can be an external device independent of the target vehicle (e.g., a vehicle diagnostic tool) or a device integrated into the target vehicle (e.g., a central control unit). This application does not limit the scope of the application.

[0034] In a specific embodiment, the vehicle diagnostic device can establish a physical and data link connection with the target vehicle, complete a communication protocol handshake (e.g., ISO 15765 CAN bus protocol), and realize a data interaction channel between the device and the vehicle. Then, the vehicle diagnostic device can send a standardized data request command to the vehicle ECU at a preset sampling frequency (e.g., 10Hz) to request the reading of the vehicle's operating data, including parameters such as engine speed, vehicle speed, oil pressure, battery voltage, braking status, and fault codes, which are not limited here. After receiving the raw data fed back by the vehicle ECU, the device performs integrity verification and format parsing on the data, extracts valid operating parameters, and forms the raw operating data sequence of the target vehicle. Then, the raw operating data sequence can be preprocessed for data standardization, for example, using the Min-Max normalization method to normalize the data to the [0, 1] interval and remove outliers to obtain target operating data with a unified format that is compatible with subsequent model inputs for use by the local AI diagnostic model.

[0035] In some embodiments, the vehicle diagnostic equipment can be a vehicle diagnostic tool. The vehicle diagnostic tool can establish a physical and data link connection with the on-board diagnostic system of the target vehicle through the OBD interface, complete the communication protocol handshake, and establish a data interaction channel. Then, the vehicle diagnostic tool sends a standardized data request command to the vehicle ECU at a preset sampling frequency, requesting to read operating parameters such as engine speed, vehicle speed, oil pressure, battery voltage, braking status, and fault codes. After receiving the raw data fed back by the vehicle ECU, the tool performs integrity verification and format parsing on the data, extracts valid operating parameters, and forms the raw operating data sequence of the target vehicle. Subsequently, the raw operating data sequence is preprocessed to obtain the target operating data.

[0036] Among them, the OBD interface is the vehicle automatic diagnostic interface, which is a standard external communication interface reserved by the vehicle. It is mainly used to establish data communication between external devices and the vehicle's electronic control system and ECU. It can read various vehicle condition data such as real-time operating parameters, historical fault codes, and operating status. It is a universal physical interface for external vehicle diagnostic equipment to access the vehicle and collect operating data, and is widely compatible with various vehicle diagnostic peripherals.

[0037] In some embodiments, the vehicle diagnostic equipment can be the central control unit (CCU) of the target vehicle. The CCU can establish a built-in data link connection with the ECU, sensors, and other components of the target vehicle through the vehicle's internal bus (e.g., CAN or LIN bus), without the need for additional physical interfaces, and complete the communication handshake through the vehicle network protocol. Subsequently, the CCU sends data acquisition commands to the vehicle ECU and sensors at a preset sampling frequency to read operating parameters such as engine speed, vehicle speed, oil pressure, battery voltage, braking status, and fault codes. After receiving the raw data fed back by the vehicle ECU and sensors, it performs integrity verification and format parsing on the data, extracts valid operating parameters, and forms a raw operating data sequence. Then, it preprocesses the raw operating data sequence to obtain the target operating data.

[0038] S302. The target running data is processed by the preset local AI diagnostic model to obtain a first diagnostic result.

[0039] In this embodiment of the application, the preset local AI diagnostic model can be preset in advance or defaulted. Specifically, the preset local AI diagnostic model can be a lightweight model, which is constructed using a single LSTM layer architecture and achieves model lightweighting by reducing the number of convolutional layers.

[0040] In a specific embodiment, the target running data can be input into a preset local AI diagnostic model, and a first diagnostic result can be output.

[0041] S303. Obtain the first network status of the vehicle diagnostic device; the first network status includes any one of the following: network normal, network abnormal.

[0042] In this embodiment, the network connectivity status of the vehicle diagnostic equipment is detected in real time, and the data transmission channel between the equipment and the preset cloud server is checked for smoothness. If data interaction can be established normally and network data packets can be successfully sent and received, the first network status is determined to be normal. If network disconnection, weak signal, connection timeout, or inability to establish communication with the cloud server occurs, the first network status is determined to be abnormal.

[0043] S304. When the first network state includes a normal network, the target running data is uploaded to the preset cloud server; the target running data is processed by the preset cloud AI diagnostic model to obtain a second diagnostic result.

[0044] In this embodiment, both the preset cloud server and the preset cloud AI diagnostic model can be preset or defaulted in advance. Specifically, the preset cloud AI diagnostic model can be a high-precision complex early warning model, which is built using a CNN and LSTM fusion architecture. It relies on CNN to mine local features in vehicle operation data and uses LSTM to capture the inherent temporal correlation features of the data. It has powerful feature extraction and fault analysis capabilities. After the model is trained on a large number of vehicle condition samples in advance, it is pre-deployed in the preset cloud server and can be directly called to carry out high-precision fault diagnosis.

[0045] In a specific embodiment, when the first network state includes a normal network, the vehicle diagnostic device can send the target operating data to a preset cloud server; then, the preset cloud server can input the target operating data into a preset cloud AI diagnostic model and output a second diagnostic result.

[0046] When the first network status includes network anomaly, there is no need to upload target operation data to the preset cloud server. Instead, the target fault type and target fault level are determined separately based on the first diagnostic result, and then the corresponding early warning operation is executed according to the target fault type and target fault level.

[0047] S305. Determine the target fault level based on the first diagnostic result and the second diagnostic result.

[0048] In the embodiments of this application, each fault level may include any of the following: emergency fault, serious fault, minor fault, etc., without limitation.

[0049] In a specific embodiment, the first diagnostic result and the second diagnostic result can be analyzed to determine the target fault level.

[0050] In some embodiments, the first diagnostic result includes: a first fault type and a first fault confidence level; the second diagnostic result includes: a second fault type and a second fault confidence level; determining the target fault level based on the first diagnostic result and the second diagnostic result includes: S11. Determine whether the first fault type and the second fault type are consistent; S12. If they match, then determine the target fault type based on the first fault type; determine the first fault score corresponding to the target fault type; determine the comprehensive confidence level based on the first fault confidence level and the second fault confidence level; determine the first adjustment coefficient corresponding to the comprehensive confidence level; determine the vehicle driving state corresponding to the target vehicle based on the target operating data; determine the second adjustment coefficient corresponding to the vehicle driving state; adjust the first fault score based on the first adjustment coefficient and the second adjustment coefficient to obtain a second fault score; determine the target fault level based on the second fault score. S13. If they are inconsistent, determine the larger confidence level between the first fault confidence level and the second fault confidence level, and determine the fault type corresponding to the larger confidence level as the target fault type; determine the target fault level based on the target fault type and the larger confidence level.

[0051] In this embodiment of the application, each fault type may include any of the following: power system fault, electrical circuit fault, transmission system fault, chassis driving fault, etc., without limitation; the fault score is a quantitative value that characterizes the severity of vehicle faults. The higher the fault score, the more serious the fault. Specifically, the fault score can range from 0 to 100.

[0052] In a specific embodiment, the first fault type and the second fault type can be compared item by item to check whether the system to which the fault belongs, the fault manifestation, and the fault judgment direction are the same. If all the judgment contents match, the two are determined to be consistent; if there are different judgment categories or deviations in the fault direction, the two are determined to be inconsistent.

[0053] If the two are consistent, the first fault type can be directly identified as the target fault type. Next, the first fault score corresponding to the target fault type can be determined. Specifically, a pre-stored mapping relationship between fault types and fault scores can be used to determine the first fault score corresponding to the target fault type. Then, the comprehensive confidence level can be determined based on the first fault confidence level and the second fault confidence level. Specifically, the first weight corresponding to the first fault confidence level and the second weight corresponding to the second fault confidence level can be determined, wherein the sum of the first weight and the second weight is 1. The comprehensive confidence level is obtained by performing a weighted operation based on the first fault confidence level, the second fault confidence level, the first weight, and the second weight.

[0054] In some embodiments, the first weight can be 0.3 and the second weight can be 0.7.

[0055] Next, a first adjustment coefficient corresponding to the overall confidence level can be determined. Specifically, a pre-stored mapping relationship between confidence levels and adjustment coefficients can be used to determine the first adjustment coefficient corresponding to the overall confidence level. The higher the overall confidence level, the smaller the first adjustment coefficient. Then, the vehicle driving state corresponding to the target vehicle can be determined based on the target operating data. Specifically, driving parameters such as vehicle speed and vehicle acceleration can be extracted from the target operating data, and the vehicle driving state can be determined based on these parameters. For example, assuming the vehicle speed is stable with small fluctuations and the vehicle acceleration is within a small range, the vehicle driving state is determined to be a constant speed driving state. Or, assuming the vehicle speed increases and the vehicle acceleration is a large positive value, the vehicle driving state is determined to be a rapid acceleration driving state, and so on. The vehicle driving state can include any of the following: constant speed driving state, rapid acceleration driving state, rapid deceleration driving state, high speed driving state, low speed driving state, etc., without limitation.

[0056] Then, a second adjustment coefficient corresponding to the vehicle's driving state can be determined. Specifically, a pre-stored mapping relationship between driving states and adjustment coefficients can be used to determine the second adjustment coefficient corresponding to the vehicle's driving state. The values ​​of the first and second adjustment coefficients can range from -0.2 to 0.2. Next, the first fault score can be adjusted according to the first and second adjustment coefficients, as follows: Second fault score = First fault score × (1 + First adjustment coefficient) × (1 + Second adjustment coefficient); The second fault score can be obtained by calculating according to the above formula, and then the target fault level can be determined based on the second fault score.

[0057] If the two are inconsistent, the higher confidence level is selected from the first and second fault confidence levels to obtain a larger confidence level. The fault type corresponding to this larger confidence level in the first and second fault types is determined as the target fault type. Finally, the target fault level can be determined based on the target fault type and the larger confidence level. Specifically, the first fault score can be determined first based on the target fault type, and then the first fault score can be adjusted based on the larger confidence level and the vehicle driving status to obtain a second fault score. Finally, the target fault level is determined based on the second fault score.

[0058] Thus, when the diagnostic results from both ends are consistent, the fault score is adjusted by combining the overall confidence level and the vehicle driving status, so that the fault level determination is consistent with the actual driving conditions and the accuracy of the determination is improved; when the results are inconsistent, the higher confidence diagnostic conclusion is selected for rapid classification, taking into account the accuracy, rationality and response efficiency of the fault determination, and effectively improving the reliability of vehicle fault diagnosis.

[0059] In some embodiments, determining the target fault level based on the second fault score includes: S21. Obtain the target vehicle model and first location corresponding to the target vehicle; S22. Determine the vehicle repair shop data within a preset distance range around the first location; the vehicle repair shop data includes: the location of a shops and the repairable vehicle models of a shops; a is a positive integer; S23. Based on the repairable vehicle data of the a stores, determine b stores among the a stores that can repair the target vehicle; b is an integer less than or equal to a. S24. Determine the locations of the b stores corresponding to the b stores based on the locations of the a stores; S25. Determine the distance between each of the b store locations and the first location to obtain b distances; S26. Determine the target fault level based on the b stores, the b routes, and the second fault score.

[0060] In this embodiment of the application, the preset distance range can be preset in advance or set by default.

[0061] In a specific embodiment, the vehicle model of the target vehicle can be obtained, and the target vehicle model can be determined based on the vehicle model. Then, the target vehicle may contain a positioning module, and the positioning information in the positioning module can be read to determine the first location. Then, the vehicle repair shop data within a preset distance range around the first location can be determined. Specifically, publicly available map data can be obtained through the Internet. Taking the first location as the center, relevant information data of all vehicle repair shops within the preset distance range can be retrieved from the publicly available map data to obtain the vehicle repair shop data.

[0062] Then, based on the repairable vehicle data of store 'a', we can determine 'b' stores among store 'a' that can repair the target vehicle. Specifically, for each store's repairable vehicle data, we match the repairable vehicles with the target vehicle. If the match is successful, we add the store to store 'b'; if the match fails, we discard the store. In this way, we can obtain store 'b'. Then, we can extract the corresponding store locations of store 'b' from the store locations of store 'a'.

[0063] Furthermore, the distance between each of the b store locations and the first location can be determined, resulting in b distances. Specifically, for each store location, the shortest route from the first location to that store location can be determined based on publicly available map data. Then, the length (i.e., distance) of the shortest route can be determined, thus obtaining b distances. Finally, the target fault level can be determined based on the b stores, the b distances, and the second fault score.

[0064] In this way, by combining vehicle model, real-time location, distribution of nearby repair shops, travel distance, and comprehensive fault score to determine the fault level, the system relies on the severity of the vehicle's fault as the core criterion, while also taking into account the proximity and suitability of offline repair resources. This allows the fault level classification to no longer solely depend on the fault itself, but can be reasonably adjusted based on nearby maintenance conditions. This makes the fault classification more closely aligned with actual travel and maintenance scenarios, facilitating the subsequent accurate delivery of repair suggestions and graded warning prompts, and improving the practicality and rationality of fault handling.

[0065] In some embodiments, please refer to Figure 4 , Figure 4 This is a flowchart of a method for determining a target fault level according to an embodiment of this application. It shows that determining the target fault level based on the b stores, the b routes, and the second fault score includes... Figure 4 The steps shown are as follows: S31. Determine the shortest distance among the b distances; S32. Determine the target store corresponding to the minimum distance among the b stores; S33. Determine the target store rating corresponding to the target store; S34. Determine the target correction coefficient based on the target store score; S35. Correct the second fault score according to the target correction coefficient to obtain the third fault score; S36. Determine the target fault level based on the third fault score.

[0066] In this embodiment, the minimum value among b distances, i.e., the minimum distance, can be found. Then, the target store corresponding to the minimum distance among the b stores can be determined. Specifically, the target store corresponding to the minimum distance can be found among the b stores. Then, the target store rating corresponding to the target store can be determined. Specifically, the store rating of the target store can be obtained from the Internet, or the basic store information of the target store can be obtained, and the target store can be scored according to the basic store information and the preset scoring rules to obtain the target store rating. The preset scoring rules can be preset in advance or defaulted.

[0067] For example, let's assume the preset scoring rules are (the total score is 100 points): Possesses official brand-authorized repair qualifications: 30 points; Main vehicle models repaired include the target model: 25 points; Open 24 / 7: 20 minutes; Online overall user satisfaction rate ≥ 95%: 15 points; Equipped with professional fault diagnosis and repair equipment: 10 points; Basic store information is as follows: The store does not have official brand repair qualifications, specializes in repairing the target model, operates normally during the day and closes in the evening, has an online user satisfaction rate of ≥95%, and is equipped with basic repair equipment.

[0068] The target store is scored according to the preset scoring rules: no points are awarded for lack of repair qualifications, 25 points are awarded for being good at repairing the target model; 10 points are awarded for normal daytime operation and evening closure; 15 points are awarded for an online overall user satisfaction rate of ≥95%; 5 points are awarded for having basic repair equipment; the total score (i.e. the target store rating) is 55 points.

[0069] Furthermore, a target correction coefficient can be determined based on the target store rating. Specifically, a pre-stored mapping relationship between store ratings and correction coefficients can be used to determine the target correction coefficient corresponding to the target store rating. The target correction coefficient can range from -0.3 to 0.3. Then, the second fault rating can be corrected based on the target correction coefficient, as follows: Third fault score = Second fault score × (1 + Target correction coefficient); The third fault score is calculated using the formula above. Finally, the target fault level can be determined based on the third fault score. Specifically, the target fault level can be determined based on the magnitude of the third fault score. For example, a third fault score of 80 or above corresponds to an emergency fault; 40 to 79 corresponds to a major fault; and below 40 corresponds to a minor fault.

[0070] In some embodiments, an emergency fault may directly threaten driving safety and requires immediate stopping and handling. Typical faults include engine misfire and braking system failure. The determination criteria are a fault prediction probability of ≥90% or the identification of a core safety system fault code.

[0071] Significant faults can affect vehicle performance and require timely repair at a service center. Typical faults include transmission jerking and abnormal oil pressure. The criteria for judgment are that the probability of fault prediction is in the range of 70% to 89%, or that a fault code related to vehicle performance is identified.

[0072] Minor faults do not affect the vehicle's core driving functions and can be addressed during routine maintenance. Typical faults include insufficient air conditioning cooling and stuck window operation. The criteria for judgment are that the fault prediction probability is in the range of 60% to 69%, or only non-core system fault codes are identified.

[0073] In this way, by prioritizing the selection of the nearest suitable repair shop, and determining the correction coefficient based on the comprehensive score of the shop, the fault score is corrected to obtain the final score and then the fault level is determined. The severity of the fault is combined with the nearby maintenance conditions and the service capabilities of the shop, so that the fault level determination is no longer limited to the fault data of the vehicle itself, but takes into account the actual harm of the fault and the convenience of offline maintenance, making the classification results more in line with the actual vehicle use scenario, and making it easier to accurately push reasonable fault handling suggestions and maintenance arrangements.

[0074] S306. Perform the corresponding target warning operation according to the target fault level; the target alarm operation includes at least one of the following: pop-up alarm, audible and visual alarm, and SMS alarm.

[0075] In this embodiment, a pre-stored mapping relationship between fault levels and warning operations may be used to determine the target warning operation corresponding to the target fault level.

[0076] In some embodiments, if the target fault level is an emergency fault, then the target warning operation is to simultaneously activate three types of alarms: pop-up alarm, audible and visual alarm, and SMS alarm. Specifically, a dual warning can be enabled, which includes a red background APP pop-up window with warning light and warning sound effects, and SMS push notifications to the vehicle owner's mobile phone number. The pop-up window cannot be manually closed and can only be deactivated after the vehicle owner clicks to confirm and stops the vehicle. If the target fault level is a critical fault, then the target warning operation is to only activate the pop-up alarm. Specifically, a yellow background APP pop-up notification will be pushed. The pop-up can be manually closed, and the reminder will automatically pop up again after one hour. If the target fault level is minor, the target warning operation will not trigger an active pop-up alarm. Specifically, it will only be displayed silently as a gray icon in the APP maintenance reminder section, without actively popping up a reminder, for the car owner to view when they actively check.

[0077] In this way, by matching different fault levels with corresponding warning methods and flexibly selecting various alarm formats such as pop-ups, sounds, and text messages, it is possible to achieve graded control of warning intensity. High-risk faults provide strong reminders to ensure driving safety, while low-risk faults provide lightweight reminders to avoid user confusion. This ensures that dangerous faults reach the vehicle owner in a timely manner, while reducing information interference caused by invalid reminders, thereby improving the accuracy of vehicle fault warnings and the user experience.

[0078] In some embodiments, after performing the corresponding target early warning operation according to the target fault level, the method further includes: S41. Obtain a preset store database; the preset store database includes: c store feature vectors corresponding to c stores, and c store locations; each store corresponds to one store feature vector; c is a positive integer; S42. Determine the target fault feature vector corresponding to the target fault type; S43. Based on a preset matching algorithm, determine the fault fit degree between the target fault feature vector and the c store feature vectors to obtain c fault fit degrees; each fault fit degree corresponds to a store feature vector. S44. Determine the distance between the c store locations and the first location to obtain c distances; S45. Determine c store ratings based on the c routes and the c fault adaptability. S46. Based on the ratings of the c stores and the preset store database, determine the store push information; the store push information is used to prompt the customer to go to the corresponding store for fault repair.

[0079] In this embodiment, the preset store database and preset matching algorithm can be preset in advance or defaulted.

[0080] In a specific embodiment, a preset store database can be obtained; then, the target fault feature vector corresponding to the target fault type can be determined. Specifically, multi-dimensional fault attribute information such as fault category, fault severity, fault system, fault impact range, and compatible maintenance technology type corresponding to the target fault type can be obtained. Then, the multi-dimensional fault attribute information is quantified and structured and integrated according to the preset vector encoding rules to generate a target fault feature vector consistent with the store feature vector format.

[0081] Then, based on a preset matching algorithm, the fault fit between the target fault feature vector and the c store feature vectors can be determined, resulting in c fault fits. Specifically, the preset matching algorithm can be an item-based collaborative filtering algorithm, which calculates the similarity between the target fault feature vector and each store feature vector to obtain c similarities, i.e., c fault fits. Next, the distances between the c store locations and the first location can be determined, resulting in c distances. Specifically, the method for obtaining the c distances can be the same as the method for obtaining the b distances mentioned above, and will not be repeated here.

[0082] Then, based on c routes and c fault adaptability, c store scores can be determined. Specifically, for each route and its corresponding fault adaptability, preset route weights and preset adaptability weights can be obtained. Based on the preset route weights, preset adaptability weights, routes and their corresponding fault adaptability, a weighted calculation is performed to obtain the corresponding store scores. In this way, c store scores can be obtained.

[0083] In some embodiments, the process of obtaining the ratings of c stores is as follows: An item-based collaborative filtering algorithm is used to calculate the similarity between the target fault feature vector and the feature vector of each store, resulting in c similarity scores. Then, the Haversine formula is used to calculate c actual distances between the driver's real-time location and the locations of the c stores, and c distance weights are derived, with closer distances corresponding to higher weight scores. Simultaneously, c user rating scores for each of the c stores are retrieved. The preset weighting is 0.4 for similarity, 0.3 for distance, and 0.3 for rating score. Store ratings are then calculated based on this data, as follows: First store rating = 0.4 × First similarity + 0.3 × First distance weight + 0.3 × First user review score; In this calculation, the first store rating is one of the c store ratings; the first similarity is any one of the c similarities; the first distance weight is the distance weight corresponding to the first similarity among the c distance weights; the first user rating score is the user rating score corresponding to the first similarity among the c user rating scores; and so on, c store ratings can be obtained by calculating c times.

[0084] It should be explained that the Haversine formula, also known as the Haversine formula, is a formula for calculating the straight-line distance between two points on the Earth's surface based on their latitude and longitude. It is suitable for calculating the distance between a car owner's location and a store's location.

[0085] Finally, the store push information can be determined based on the ratings of c stores and the preset store database.

[0086] In this way, by combining fault characteristics to determine repair suitability, and then combining the store's geographical location to calculate the travel distance, the store score is calculated and repair store information is pushed out by comprehensively considering multiple factors. This ensures that the repair store is well-matched with the vehicle's fault, while also taking into account the travel distance. It can accurately select repair stores with high suitability and convenient travel, providing car owners with reasonable and reliable repair location suggestions, and effectively improving the practicality and rationality of fault repair guidance.

[0087] In some embodiments, the process of obtaining the preset store database is as follows: Information collected: Collect multi-dimensional basic information about each repair shop, including shop location, shop qualification type, main vehicle models that can be repaired, types of faults that can be repaired, parts quality level, user rating, repair appointment response time, etc. Feature Quantization Construction: The collected multi-dimensional basic information of stores is structured and normalized. A unique store feature vector is constructed for each repair store according to preset vector coding rules. The vector contains quantitative feature data such as store location coordinates, main vehicle types, fault repair compatibility categories, and user comprehensive ratings, thus completing the database construction. For example, the feature vector of a certain store could be "Location coordinates: (x, y), Main vehicle types: New energy / fuel, Fault compatibility: Engine / transmission, Rating: 4.8 stars".

[0088] In some embodiments, please refer to Figure 5 , Figure 5 This is a flowchart illustrating a method for determining store push information provided in an embodiment of this application. It shows that determining store push information based on the c store ratings and the preset store database includes... Figure 5 The steps shown are as follows: S51. Sort the c stores in descending order of their ratings to obtain the first store order; S52. Filter the first N stores from the first store sequence, and obtain the N store information corresponding to the N stores from the preset store database; N is a preset value; S53. Generate the store push information based on the N store information.

[0089] In this embodiment of the application, N can be preset or defaulted, for example, N can be 3.

[0090] In a specific embodiment, the scores of the c stores are compared pairwise, and the stores are sorted from largest to smallest to form a first order of stores. The higher the score, the higher the priority of the store recommendation. Then, the top N stores in the first order are selected, and the information of the N stores is retrieved from a preset store database. This information includes store location, qualification type, main repair services, user reviews, appointment methods, etc. Finally, store push information is generated based on the N store information. Specifically, the N store information is organized and arranged according to a standardized push format to form an intuitive and clear push information, which can display store ranking, distance, fault compatibility, store details, etc., and is used to push repair store recommendation prompts to car owners.

[0091] In some embodiments, after generating store push information, the method further includes: The system displays store information on the vehicle diagnostic equipment's screen. Simultaneously, it integrates detailed information from the top N stores into the vehicle fault diagnosis report, displaying store name, address, distance from the owner's real-time location, main repair services, user reviews, and an online appointment portal. Owners can use the app to make an appointment, select a planned repair time, and fill in a description of the vehicle's fault. Once completed, the appointment data is synchronized in real-time to the corresponding store's backend management system. After store staff verify the information and confirm the order, the system automatically sends a successful appointment notification to the owner via the app, along with the store contact person's contact information and the optimal route to the store, completing the entire repair appointment service.

[0092] In this way, by sorting stores from high to low according to their comprehensive scores and selecting the best ones, only the top-ranked stores are selected to integrate information and generate push content. This not only prioritizes recommending repair stores with high fault adaptability, proximity, and good reputation, but also simplifies the push content to avoid information overload. It can also quickly provide car owners with high-quality and reliable repair location solutions, greatly improving the accuracy and reference value of repair store recommendations.

[0093] In some embodiments, please refer to Figure 6 , Figure 6 This is a flowchart of a model training method provided in an embodiment of this application. As can be seen, the method further includes... Figure 6 The steps shown are as follows: S61. Utilize big data technology to obtain vehicle operation datasets; S62. Train the first preset AI diagnostic model based on the vehicle operation dataset to obtain the preset cloud AI diagnostic model; S63. Using the preset cloud AI diagnostic model as the teacher model and the second preset AI diagnostic model as the student model, perform knowledge distillation training on the second preset AI diagnostic model to obtain the preset local AI diagnostic model.

[0094] In this embodiment, both the first preset AI diagnostic model and the second preset AI diagnostic model can be preset in advance or defaulted to; wherein, the first preset AI diagnostic model can be a high-precision complex early warning model, which is built using a CNN and LSTM fusion architecture; the second preset AI diagnostic model can be a lightweight model, which is built using a single LSTM layer architecture.

[0095] In a specific embodiment, big data technology is used to collect massive amounts of vehicle operation sequence data covering more than 100 vehicle models and 501 common potential faults. The data types include normal vehicle operation data and early warning time sequence data before the fault occurs, ensuring that the data covers a wide range of vehicle models and fault types. Subsequently, the collected raw data is cleaned, deduplicated, and standardized to remove abnormal and invalid data, ultimately forming a high-quality vehicle operation dataset that can be used for model training. Then, the first preset AI diagnostic model can be trained based on the vehicle operation dataset to obtain a preset cloud-based AI diagnostic model. Specifically, the first preset AI diagnostic model is trained using the vehicle operation dataset as training samples, with the training objectives being the prediction of fault type, fault probability, and fault occurrence time. Gradient descent and other optimization algorithms are used to iteratively update the model parameters, continuously improving the model's prediction accuracy until the model converges and the potential fault prediction accuracy reaches more than 92%. After training, a high-precision preset cloud-based AI diagnostic model is obtained.

[0096] Then, a preset cloud-based AI diagnostic model can be used as the teacher model, while a second preset AI diagnostic model can be used as the student model. The distillation temperature coefficient is set to 8. The soft labels (fault prediction probability distribution) output by the teacher model are combined with the hard labels of real faults. The mean squared error loss function is used to optimize the parameters of the student model to complete the knowledge distillation training. Finally, under the premise that the total number of student model parameters is reduced by 70%, the prediction accuracy is guaranteed to decrease by no more than 7% and the final accuracy is not less than 85%, resulting in a lightweight, vehicle-deployable preset local AI diagnostic model.

[0097] In this way, by training a high-precision cloud-based diagnostic model based on massive vehicle operation data, and then completing the model capability transfer through knowledge distillation, the fault diagnosis knowledge of the cloud-based model is empowered to the lightweight local model. This approach not only relies on the complex fusion model to ensure the overall accuracy of fault diagnosis, but also significantly reduces the number of model parameters and computing power consumption, allowing the lightweight local model to be directly deployed on the vehicle terminal, taking into account both diagnostic accuracy and the practical application requirements of real-time and low-power operation on the vehicle terminal.

[0098] In summary, this method utilizes a local model (i.e., a pre-set local AI diagnostic model) to perform real-time data analysis on-site, quickly outputting the first diagnostic result to ensure the real-time nature of fault identification. Furthermore, through network status determination logic, vehicle operation data is uploaded to the cloud when the network is normal. A second, in-depth diagnosis is then performed using a high-precision model in the cloud (i.e., a pre-set cloud AI diagnostic model), generating a second diagnostic result to compensate for the accuracy limitations of the local model. Based on this, the local and cloud-based dual-dimensional diagnostic results are integrated for mutual verification and correction, jointly determining the fault type and level. This effectively reduces misjudgments and omissions associated with single-model diagnosis, significantly lowering the probability of false warnings. Simultaneously, differentiated pop-ups, audible alerts, and SMS messages are matched according to the fault type and level, avoiding excessive alerts for minor faults while providing strong warnings for high-risk faults. Moreover, in the event of network anomalies, the local model alone can independently complete diagnostic warnings, ensuring continuous and stable service operation, thereby comprehensively improving the reliability of vehicle fault warnings.

[0099] Please see Figure 7 , Figure 7 This is a schematic diagram of a vehicle fault warning device provided in an embodiment of this application. It is applied to a vehicle diagnostic device, which includes a preset local AI diagnostic model and is connected to a preset cloud server. The preset cloud server contains a preset cloud AI diagnostic model. The vehicle fault warning device 700 includes: an acquisition module 701, a first diagnostic module 702, a second diagnostic module 703, and an alarm module 704, wherein: The acquisition module 701 is used to acquire the target operating data of the target vehicle; The first diagnostic module 702 is used to process the target running data through the preset local AI diagnostic model to obtain a first diagnostic result; The acquisition module 701 is further configured to acquire the first network status of the vehicle diagnostic device; the first network status includes any one of the following: network normal, network abnormal; The second diagnostic module 703 is used to upload the target running data to the preset cloud server when the first network status includes a normal network; and to process the target running data through the preset cloud AI diagnostic model to obtain a second diagnostic result. The first diagnostic module 702 is further configured to determine the target fault level based on the first diagnostic result and the second diagnostic result; The alarm module 704 is used to perform corresponding target early warning operations according to the target fault level; the target alarm operations include at least one of the following: pop-up alarm, audible and visual alarm, and SMS alarm.

[0100] In specific implementations, the vehicle fault warning device 700 described in the embodiments of the present invention can also execute other implementations described in the vehicle fault warning method provided in the embodiments of the present invention, which will not be repeated here.

[0101] Please see Figure 8 , Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include a processor, a memory, a communication interface, and one or more programs. The processor, memory, and communication interface can be interconnected via a bus. The one or more programs are stored in the memory and configured to be executed by the processor. In this embodiment, the programs include instructions for performing the following steps: Obtain the target vehicle's operational data; The target running data is processed by the preset local AI diagnostic model to obtain a first diagnostic result; Obtain the first network status of the vehicle diagnostic device; the first network status includes any one of the following: network normal, network abnormal; When the first network state includes a normal network, the target operation data is uploaded to the preset cloud server; the target operation data is processed by the preset cloud AI diagnostic model to obtain a second diagnostic result; Based on the first diagnostic result and the second diagnostic result, the target fault level is determined; Execute corresponding target warning operations according to the target fault level; the target alarm operations include at least one of the following: pop-up alarm, audible and visual alarm, and SMS alarm.

[0102] In specific implementations, the electronic devices described in the embodiments of the present invention may also execute other implementation methods described in any of the above method embodiments, which will not be repeated here.

[0103] The processor can be a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, cells, and circuits described in conjunction with the disclosure of this application. The processor can also be a combination that implements computational functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc. The communication unit can be a communication interface, transceiver, transceiver circuit, etc., and the storage unit can be a memory.

[0104] The memory can be volatile or non-volatile, or a combination of both. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0105] It is understood that electronic devices may include more or fewer structural elements than those shown in the above block diagram, such as power modules, physical buttons, Wi-Fi modules, speakers, Bluetooth modules, sensors, display modules, etc., without limitation.

[0106] This application also provides a computer-readable storage medium storing a computer program for electronic data interchange, which causes a computer to perform some or all of the steps of any of the methods described in the above method embodiments, wherein the computer includes an electronic device.

[0107] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps of any of the methods described in the above method embodiments. The computer program product may be a software installation package, and the computer may include an electronic device.

[0108] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0109] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0110] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical or other forms.

[0111] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

[0112] The steps of the methods or algorithms described in the embodiments of this application can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in RAM, flash memory, ROM, EPROM, electrically erasable programmable read-only memory (EEPROM), registers, hard disk, portable hard disk, read-only optical disk (CD-ROM), or any other form of storage medium well known in the art. An exemplary storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and storage medium can reside in an ASIC. Furthermore, the ASIC can reside in a terminal device or management device. Alternatively, the processor and storage medium can exist as discrete components in the terminal device or management device.

[0113] Those skilled in the art will recognize that, in one or more of the examples above, the functions described in the embodiments of this application can be implemented, in whole or in part, by software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, in the form of a computer program product. This computer program product includes one or more computer instructions. When these computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated.

[0114] The aforementioned computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media.

[0115] The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., digital video discs (DVDs)), or semiconductor media (e.g., solid-state disks (SSDs)).

[0116] The modules / units included in the various devices and products described in the above embodiments can be software modules / units, hardware modules / units, or a combination of both. For example, for devices and products applied to or integrated into a chip, all modules / units can be implemented using hardware methods such as circuits, or at least some modules / units can be implemented using software programs that run on a processor integrated within the chip, while the remaining (if any) modules / units can be implemented using hardware methods such as circuits. For devices and products applied to or integrated into a chip module, all modules / units can be implemented using hardware methods such as circuits. Different modules / units can be located in the same component (e.g., chip, circuit module, etc.) or different components of the chip module, or at least some modules / units can be implemented using hardware methods such as circuits. The implementation is achieved through a software program that runs on the processor integrated within the chip module. The remaining modules / units (if any) can be implemented using hardware methods such as circuits. For various devices and products applied to or integrated into terminal equipment, each of their modules / units can be implemented using hardware methods such as circuits. Different modules / units can be located in the same component (e.g., chip, circuit module, etc.) or different components within the terminal equipment. Alternatively, at least some modules / units can be implemented through a software program that runs on the processor integrated within the terminal equipment, while the remaining modules / units (if any) can be implemented using hardware methods such as circuits.

[0117] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the embodiments of this application. It should be understood that the above descriptions are merely specific embodiments of the embodiments of this application and are not intended to limit the protection scope of the embodiments of this application. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solutions of the embodiments of this application should be included within the protection scope of the embodiments of this application.

Claims

1. A vehicle failure early warning method characterized by, The method is applied to vehicle diagnostic equipment, wherein the vehicle diagnostic equipment is equipped with a preset local AI diagnostic model, and the vehicle diagnostic equipment is connected to a preset cloud server, wherein the preset cloud server is equipped with a preset cloud AI diagnostic model; the method includes: Obtain the target vehicle's target operating data; The target running data is processed by the preset local AI diagnostic model to obtain a first diagnostic result; Obtain the first network status of the vehicle diagnostic device; the first network status includes any one of the following: network normal, network abnormal; When the first network state includes a normal network, the target operation data is uploaded to the preset cloud server; the target operation data is processed by the preset cloud AI diagnostic model to obtain a second diagnostic result; Based on the first diagnostic result and the second diagnostic result, the target fault level is determined; Execute corresponding target warning operations according to the target fault level; the target alarm operations include at least one of the following: pop-up alarm, audible and visual alarm, and SMS alarm.

2. The method of claim 1, wherein, The first diagnostic result includes: a first fault type and a first fault confidence level; the second diagnostic result includes: a second fault type and a second fault confidence level. The step of determining the target fault level based on the first diagnostic result and the second diagnostic result includes: Determine whether the first fault type and the second fault type are consistent; If they match, then determine the target fault type based on the first fault type; determine the first fault score corresponding to the target fault type; determine the comprehensive confidence level based on the first fault confidence level and the second fault confidence level; determine the first adjustment coefficient corresponding to the comprehensive confidence level; determine the vehicle driving state corresponding to the target vehicle based on the target operating data; determine the second adjustment coefficient corresponding to the vehicle driving state; adjust the first fault score based on the first adjustment coefficient and the second adjustment coefficient to obtain a second fault score; determine the target fault level based on the second fault score. If they are inconsistent, the greater confidence level between the first fault confidence level and the second fault confidence level is determined, and the fault type corresponding to the greater confidence level is determined as the target fault type; the target fault level is determined based on the target fault type and the greater confidence level.

3. The method of claim 2, wherein, Determining the target fault level based on the second fault score includes: Obtain the target vehicle model and first location corresponding to the target vehicle; Determine vehicle repair shop data within a preset distance range around the first location; the vehicle repair shop data includes: the locations of *a* shops and the repairable vehicle models for *a* shops; where *a* is a positive integer. Based on the repairable vehicle data of the a stores, determine b stores among the a stores that can repair the target vehicle; b is an integer less than or equal to a. The locations of the b stores corresponding to the a stores are determined based on the store locations of the a stores. Determine the distance between each of the b store locations and the first location to obtain b distances; The target fault level is determined based on the b stores, the b routes, and the second fault score.

4. The method of claim 3, wherein, The step of determining the target fault level based on the b stores, the b routes, and the second fault score includes: Determine the minimum distance among the b distances; Determine the target store corresponding to the minimum distance among the b stores; Determine the target store rating corresponding to the target store; Determine the target correction coefficient based on the target store score; The second fault score is corrected according to the target correction coefficient to obtain the third fault score; The target fault level is determined based on the third fault score.

5. The method of claim 3 or 4, wherein, After performing the corresponding target early warning operation based on the target fault level, the method further includes: Obtain a preset store database; the preset store database includes: c store feature vectors corresponding to c stores, and c store locations; each store corresponds to one store feature vector; c is a positive integer; Determine the target fault feature vector corresponding to the target fault type; Based on a preset matching algorithm, the fault fit degree between the target fault feature vector and the c store feature vectors is determined, resulting in c fault fit degrees; each fault fit degree corresponds to a store feature vector. Determine the distances between the c store locations and the first location to obtain c distances; Based on the c routes and the c fault adaptability, determine c store scores; Based on the ratings of the c stores and the preset store database, store push information is determined; the store push information is used to prompt customers to go to the corresponding store for fault repair.

6. The method of claim 5, wherein, The process of determining store push information based on the c store ratings and the preset store database includes: The c stores are sorted from highest to lowest score to obtain the first store order. The first N stores are selected from the first store sequence, and the information of the N stores corresponding to the N stores is obtained from the preset store database; N is a preset value. The store push information is generated based on the information of the N stores.

7. The method according to any one of claims 1 to 4, wherein The method further includes: Utilize big data technology to obtain vehicle operation datasets; The first preset AI diagnostic model is trained based on the vehicle operation dataset to obtain the preset cloud AI diagnostic model; The preset cloud-based AI diagnostic model is used as the teacher model, and the second preset AI diagnostic model is used as the student model. Knowledge distillation training is performed on the second preset AI diagnostic model to obtain the preset local AI diagnostic model.

8. A vehicle failure early warning device characterized by comprising: An application is made in vehicle diagnostic equipment, wherein the vehicle diagnostic equipment is equipped with a preset local AI diagnostic model, and the vehicle diagnostic equipment is connected to a preset cloud server, the preset cloud server being equipped with a preset cloud AI diagnostic model; the device includes: an acquisition module, a first diagnostic module, a second diagnostic module, and an alarm module, wherein: The acquisition module is used to acquire the target operating data of the target vehicle; The first diagnostic module is used to process the target running data through the preset local AI diagnostic model to obtain a first diagnostic result; The acquisition module is further configured to acquire the first network status of the vehicle diagnostic device; the first network status includes any one of the following: network normal, network abnormal; The second diagnostic module is used to upload the target running data to the preset cloud server when the first network status includes a normal network; and to process the target running data through the preset cloud AI diagnostic model to obtain a second diagnostic result. The first diagnostic module is further configured to determine the target fault level based on the first diagnostic result and the second diagnostic result; The alarm module is used to perform corresponding target early warning operations according to the target fault level; the target alarm operations include at least one of the following: pop-up alarm, audible and visual alarm, and SMS alarm.

9. An electronic device, comprising: include: Processor, memory, communication interface, and one or more programs; The one or more programs are stored in the memory and configured to be executed by the processor, the programs including instructions for performing the steps of the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, A computer program for storing electronic data interchange, wherein the computer program causes a computer to perform the method as described in any one of claims 1-7.

Images