Fault processing method and device and electronic equipment
By matching natural language keywords and fault feature combinations in the fault event library, the problem of poor timeliness of the fault event library in the existing technology is solved, and automatic updates and more efficient fault handling effects are achieved.
Patent Information
- Application Number
- CN202510021748.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-06
- Publication Date
- 2025-05-27
AI Technical Summary
In the prior art, professional maintenance personnel are required to maintain keywords in the fault event database, resulting in poor timeliness of the fault event database, which in turn affects the effect of fault handling.
By obtaining natural language keywords from the user fault description statement, match the associated fault events in the fault event library, if a unique fault event is not matched, use the fault feature combination to perform a quadratic match, and add unexistent keywords to the candidate keywords of the fault event in the matching result to achieve automatic update.
It improves the success rate and effectiveness of fault handling, reduces dependence on professional maintenance personnel, and enhances the timeliness and adaptability of the fault event database.
Smart Images
Figure CN120045757A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of automobile technology, and in particular to a fault handling method, device and electronic equipment. Background Art
[0002] With the development of vehicle intelligence, vehicle diagnosis is also gradually becoming intelligent; for example, a maintenance analysis model can be used to analyze a faulty vehicle, and then a maintenance plan can be determined based on the analysis results.
[0003] In the related art, the reported keywords about vehicle failures can be used to retrieve fault events matching the keywords in the fault event library, and then the maintenance analysis model suitable for the fault event can be called to perform fault analysis; however, for keywords that are not maintained in the fault event library, professional maintenance personnel are required to maintain the keywords in the fault event library before retrieval and matching can be performed, resulting in poor timeliness of the fault event library, and in turn resulting in poor results in fault handling based on the story event library. Summary of the invention
[0004] In view of this, the embodiments of the present application propose a fault handling method, device and electronic device to solve the problem that the timeliness of the fault event library is poor due to the need for professional maintenance personnel to maintain keywords in the fault event library.
[0005] The embodiment of the present application is implemented by adopting the following technical solutions:
[0006] In a first aspect, an embodiment of the present application provides a fault handling method, the method comprising: obtaining a target natural language keyword from a user fault description statement reported for a first vehicle; matching a fault event associated with the target natural language keyword in a fault event library to obtain a first matching result, wherein the fault event library contains multiple fault events and candidate natural language keywords associated with each of the fault events and a fault feature combination associated with each of the fault events; if the first matching result indicates that a unique first fault event is not matched in the fault event library, obtaining a target fault feature combination of a fault alarm signal reported by the first vehicle; the first fault event refers to a fault event whose associated candidate natural language keywords include the target natural language keyword; matching a fault event associated with the target fault feature combination in the fault event library to obtain a second matching result; if the second matching result indicates that a second fault event associated with the target fault feature combination is matched in the fault event library, and the target natural language keyword does not exist in the candidate natural language keywords associated with the second fault event, adding the target natural language keyword to the candidate natural language keywords associated with the second fault event; and pushing a maintenance analysis model applicable to the second fault event to the user.
[0007] In a second aspect, an embodiment of the present application provides a fault handling device, the device comprising: a first acquisition module, used to acquire target natural language keywords from a user fault description statement reported for a first vehicle; a first matching module, used to match fault events associated with the target natural language keywords in a fault event library to obtain a first matching result, wherein the fault event library contains multiple fault events and candidate natural language keywords associated with each of the fault events and a fault feature combination associated with each of the fault events; a second acquisition module, used to acquire a target fault feature combination of a fault alarm signal reported by the first vehicle if the first matching result indicates that no unique first fault event is matched in the fault event library; the first A fault event refers to a fault event whose associated candidate natural language keywords include the target natural language keyword; a second matching module is used to match the fault event associated with the target fault feature combination in the fault event library to obtain a second matching result; a self-update module is used to add the target natural language keyword to the candidate natural language keywords associated with the second fault event if the second matching result indicates that the second fault event associated with the target fault feature combination is matched in the fault event library and the target natural language keyword does not exist in the candidate natural language keywords associated with the second fault event; a push module is used to push a maintenance analysis model applicable to the second fault event to the user.
[0008] In a third aspect, an embodiment of the present application provides an electronic device, which includes a processor; a memory, wherein computer-readable instructions are stored in the memory, and when the computer-readable instructions are executed by the processor, the above method is implemented.
[0009] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having computer-readable instructions stored thereon. When the computer-readable instructions are executed by a processor, the above-mentioned method is implemented.
[0010] A fault handling method, device and electronic device provided in an embodiment of the present application, the method comprising: obtaining target natural language keywords from a user fault description statement reported for a first vehicle; matching fault events associated with the target natural language keywords in a fault event library to obtain a first matching result, wherein the fault event library contains multiple fault events and candidate natural language keywords associated with each fault event and a fault feature combination associated with each fault event; if the first matching result indicates that a unique first fault event is not matched in the fault event library, it means that there is no or multiple first fault events in the fault event library. Obviously, in this case, it is impossible to uniquely determine a maintenance analysis model suitable for the first fault event for fault handling, wherein the first fault event refers to a fault event whose associated candidate natural language keywords include the target natural language keywords. Therefore, when the first matching result indicates that the unique first fault event is not matched in the fault event library, the target fault feature combination of the fault alarm signal reported by the first vehicle is obtained; and the fault event associated with the target fault feature combination is matched in the fault event library to obtain a second matching result. If the second matching result indicates that the second fault event associated with the target fault feature combination is matched in the fault event library, a maintenance analysis model suitable for the second fault event is pushed to the user, thereby ensuring the adaptability of the maintenance analysis model suitable for the second fault event pushed to the user to solving the current fault of the first vehicle.
[0011] At the same time, if the target natural language keyword does not exist in the candidate natural language keywords associated with the second fault event, the target natural language keyword is added to the candidate natural language keywords associated with the second fault event; thereby achieving automatic updating of the fault event library. It can be understood that after the fault event library is updated, if the same target natural language keyword is obtained again, the unique first fault event can be directly matched through the target natural language keyword, thereby improving the effect of fault handling based on the fault event library.
[0012] These and other aspects of the present application will become more apparent from the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.
[0014] Figure 1 A first flow chart of the fault handling method provided in an embodiment of the present application is shown.
[0015] Figure 2A first flow chart of the fault handling method provided in an embodiment of the present application is shown.
[0016] Figure 3 A third flow chart of the fault handling method provided in an embodiment of the present application is shown.
[0017] Figure 4 A fourth flow chart of the fault handling method provided in an embodiment of the present application is shown.
[0018] Figure 5 The embodiment of the present application provides Figure 4 Schematic diagram of the process of step S420.
[0019] Figure 6 A fifth flow chart of the fault handling method provided in an embodiment of the present application is shown.
[0020] Figure 7 A schematic diagram of an application scenario involved in an embodiment of the present application is shown.
[0021] Figure 8 A schematic diagram of maintaining and updating candidate natural language keywords involved in an embodiment of the present application is provided.
[0022] Fig. 9 A schematic diagram of a fault handling device provided in an embodiment of the present application is shown.
[0023] Fig.10 A schematic diagram of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION
[0024] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be understood as limiting the present application.
[0025] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. According to the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present application.
[0026] In the following description, the terms "first\second" and the like are only used to distinguish similar objects and do not represent a specific order for the objects. It is understandable that "first\second" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present application described here can be implemented in an order other than that illustrated or described here. In the following description, "some embodiments or some embodiment modes" are involved, which describe a subset of all possible embodiments, but it is understandable that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict.
[0027] The term "multiple" as used herein refers to two or more than two. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the related objects are in an "or" relationship.
[0028] With the development of vehicle intelligence, vehicle diagnosis is also gradually becoming intelligent; for example, a maintenance analysis model can be used to analyze a faulty vehicle, and then a maintenance plan can be determined based on the analysis results.
[0029] In the related art, the reported keywords about vehicle failures can be used to retrieve fault events matching the keywords in the fault event library, and then the maintenance analysis model suitable for the fault event can be called to perform fault analysis; however, for keywords that are not maintained in the fault event library, professional maintenance personnel are required to maintain the keywords in the fault event library before retrieval and matching can be performed, resulting in poor timeliness of the fault event library, and in turn resulting in poor results in fault handling based on the story event library.
[0030] The embodiments provided in this application will be described below with reference to the accompanying drawings.
[0031] See also Figure 1 , Figure 1 A first flow chart of a fault handling method provided in an embodiment of the present application is provided. The fault handling method includes steps S110-S160:
[0032] S110: Acquire target natural language keywords from a user fault description sentence reported for the first vehicle.
[0033] User fault description sentences refer to sentences describing vehicle faults in natural language. Target natural language keywords refer to words describing faults extracted from user fault description sentences. It is worth mentioning that the description of vehicle faults or accurate location of vehicle faults requires high vehicle-related professional accumulation. For users, the user fault description sentences they express describe the vehicle faults based on the results caused by the vehicle faults. The user fault description sentences may not accurately express the specific faults of the vehicle and the causes of the faults.
[0034] In some embodiments, the user fault description sentence reported for the first vehicle can be obtained by performing voice recognition on the user's voice; on this basis, a wake-up word can be pre-set, and after the wake-up word is recognized, voice recognition is performed on the user's voice after the wake-up word; or, a specific sentence format can be set, and after recognizing that the user's voice has a specific sentence format, voice recognition is performed on the user's voice.
[0035] In some embodiments, the user fault description statement reported for the first vehicle may be a description statement with a specific statement format. It is understandable that for a description statement with a specific statement format, a sentence fragment at a specified position in the description statement may be obtained; for example, the specific statement format may be "I encountered XXX fault", and under this statement format, the sentence fragment between "I encountered" and "fault" may be obtained.
[0036] In other implementations, the user fault description statement may be segmented to obtain multiple segmented words, and then the semantics of each segmented word may be identified, and the target natural language keyword may be determined from the multiple segmented words based on the semantics of each segmented word.
[0037] It is understandable that the target natural language keyword obtained from the user fault description sentence reported for the first vehicle may be one or more.
[0038] S120. Match the fault event associated with the target natural language keyword in the fault event library to obtain a first matching result. The fault event library contains multiple fault events and candidate natural language keywords associated with each fault event and a fault feature combination associated with each fault event.
[0039] Among them, a fault event corresponds to a type of vehicle fault; a candidate natural language keyword associated with a fault event refers to a search keyword used to retrieve the fault event, and a candidate natural language keyword associated with a fault event can be a keyword extracted from a natural language description sentence used to describe the fault event collected from the user's history, so as to ensure that the candidate natural language keyword associated with a fault event is a natural language keyword with a high possibility of the user using the candidate natural language keyword to describe the vehicle fault represented by the fault event, so as to ensure the effectiveness of using the candidate natural language keyword associated with the fault event to retrieve the fault event. Among them, the number of candidate natural language keywords associated with a fault event is not limited, and can be one or more.
[0040] The fault feature combination associated with a fault event refers to the combination of features obtained by extracting features from the fault warning signal reported by the vehicle when the vehicle fault corresponding to the fault event occurs. The fault warning signal reported by the vehicle is reported by the sensor on the vehicle. There are a large number of sensors installed on the vehicle. Moreover, since the multiple components on the vehicle work in coordination with each other, when a vehicle fault represented by a fault event occurs in the vehicle, the fault warning signal actually received includes the warning signals reported by multiple sensors on multiple components on the vehicle. Therefore, the warning signals from multiple sensors in the fault warning signal are respectively feature extracted to obtain the fault features of each warning signal, and then the fault features of multiple warning signals are combined as the fault feature combination associated with the fault event. Among them, one fault event is associated with one fault feature combination. The fault feature of the warning signal can be a vectorized representation of the warning signal.
[0041] When there is only one target natural language keyword, the fault event associated with the target natural language keyword refers to a fault event in which the associated candidate natural language keywords include the target natural language keyword.
[0042] In the case where there are multiple target natural language keywords, the fault event associated with the target natural language keyword refers to a fault event in which the associated candidate natural language keywords include at least one target natural language keyword.
[0043] The fault event library is pre-established, and the contents stored in the fault event library may be as shown in Table 1 below:
[0044] Fault events Candidate natural language keywords Fault feature combination A KW1, KW2 E1, E2, E3 B KW2, KW3 E2, E3 C KW1, KW4 E2, E4
[0045] Table 1
[0046] It can be seen from Table 1 that there can be multiple candidate natural language keywords associated with a fault event, and the same candidate natural language keywords can exist in multiple candidate natural language keywords associated with different fault events; at the same time, the fault features associated with different fault events can be the same, but the fault feature combinations associated with different fault events are different, that is, the associated fault event can be uniquely determined by the fault feature combination.
[0047] As mentioned above, since the fault feature combination associated with a fault event refers to a combination of features corresponding to the fault alarm signal reported by the vehicle when the vehicle fault corresponding to the fault event occurs; if the fault feature combinations of two fault events are exactly the same, it means that the fault alarm signals reported by the vehicle for the two fault events are exactly the same, that is, the two fault events are actually the same fault event. Therefore, the fault feature combinations associated with different fault events are different; and the natural language descriptions of the two fault events may have some identical descriptions. Therefore, the same candidate natural language keywords may exist among multiple candidate natural language keywords associated with different fault events.
[0048] S130. If the first matching result indicates that no unique first fault event is matched in the fault event library, obtain a target fault feature combination of the fault alarm signal reported by the first vehicle; the first fault event refers to a fault event whose associated candidate natural language keywords include the target natural language keywords.
[0049] When the first matching result indicates that no unique first fault event is matched in the fault event library, it means that the first fault event does not exist in the current fault event library, or there are multiple first fault events. Obviously, in both cases, it is impossible to uniquely determine the maintenance analysis model applicable to the first fault event for push. Therefore, it is necessary to obtain the target fault feature combination of the fault alarm signal reported by the first vehicle for secondary matching.
[0050] In some implementations, the VIN code (Vehicle Identification Number) of the first vehicle may be obtained first, and then the target fault feature combination of the fault warning signal associated with the VIN code may be queried in a database through the VIN code of the first vehicle.
[0051] In some embodiments, when the first matching result indicates that a unique first fault event is matched in the fault event library, it means that a maintenance analysis model applicable to the first fault event can be uniquely determined, and on this basis, the maintenance analysis model applicable to the first fault event can be pushed to the user.
[0052] S140: Match the fault event associated with the target fault feature combination in the fault event library to obtain a second matching result.
[0053] If the fault features in the fault feature combination associated with a fault event in the fault event library are the same as the fault features in the target fault feature combination, the fault event is determined to be a fault event associated with the target fault feature combination.
[0054] The second matching result is used to indicate whether a fault event associated with the target fault feature combination is matched in the fault event library.
[0055] S150. If the second matching result indicates that a second fault event associated with the target fault feature combination is matched in the fault event library, and the target natural language keyword does not exist in the candidate natural language keywords associated with the second fault event, add the target natural language keyword to the candidate natural language keywords associated with the second fault event.
[0056] It should be noted that the second fault event matched to the associated target fault feature combination in the fault event library refers to a fault event whose associated fault feature combination in the fault event library is exactly the same as the target fault feature combination.
[0057] In theory, the target natural language keyword and the target fault feature combination should be associated with the same fault event; however, when the second matching result indicates that a second fault event associated with the target fault feature combination is matched in the fault event library, and the target natural language keyword does not exist in the candidate natural language keywords associated with the second fault event, the possible reason is that in the process of constructing the fault event library, the candidate natural language keywords associated with the second fault event are incomplete.
[0058] Therefore, if the target natural language keyword does not exist in the candidate natural language keywords associated with the second fault event, the target natural language keyword can be added to the candidate natural language keywords associated with the second fault event, thereby updating the candidate natural language keywords in the fault event library.
[0059] It can be understood that since the fault feature combination associated with a fault event is unique, if a second fault event is matched, the second fault event must be uniquely determined. On this basis, the maintenance analysis model applicable to the second fault event can be uniquely determined. Therefore, based on the second matching result indicating that the second fault event associated with the target fault feature combination is matched in the fault event library, the maintenance analysis model applicable to the second fault event can be pushed to the user.
[0060] S160: Pushing a maintenance analysis model suitable for the second fault event to the user.
[0061] In some implementations, after the maintenance analysis model applicable to the second fault event is pushed to the user, an execution option may also be presented to the user. If the user chooses to execute, the maintenance analysis is performed using the maintenance analysis model; if the user chooses not to execute, the maintenance process ends.
[0062] Through the method provided by the present application, after obtaining the target natural language keyword from the user fault description statement reported for the first vehicle, the first fault event associated with the target natural language keyword is matched in the fault event library; if the unique first fault event is not matched, it means that there is no or multiple first fault events in the fault event library. Obviously, in this case, it is impossible to uniquely determine the maintenance analysis model applicable to the first fault event for fault processing. Therefore, when the first matching result indicates that the unique first fault event is not matched in the fault event library, the target fault feature combination of the fault alarm signal reported by the first vehicle is obtained; and the second fault event associated with the target fault feature combination is matched in the fault event library. If the second fault event associated with the target fault feature combination is matched in the fault event library, since the second fault event associated with the target fault feature combination is uniquely determined, the maintenance analysis model applicable to the second fault event can be pushed to the user, thereby improving the success rate of fault processing through two matches.
[0063] At the same time, if the target natural language keyword does not exist among the candidate natural language keywords associated with the second fault event, the target natural language keyword is added to the candidate natural language keywords associated with the second fault event; thereby achieving self-maintenance and automatic updating of the candidate natural language keywords in the fault event library. After the fault event library is updated, if the same target natural language keyword is obtained again, the unique first fault event can be directly matched through the target natural language keyword, thereby improving the effect of fault handling based on the fault event library.
[0064] In the related art, a method for matching fault events using keywords in user description sentences is provided. However, this method is seriously affected by the user description sentences. If the words in the description sentences used by the user are greatly different from the conventional description words, it will lead to the inability to determine the fault events based on the keyword search in the user description sentences. The method provided by the present application not only supports the retrieval of fault events by natural language keywords, but also supports the retrieval of fault events using the vehicle's fault feature combination when the unique fault event cannot be retrieved by natural language keywords. This can solve the problem of being unable to retrieve fault events due to the large difference between the keywords in the user description sentences and the candidate natural language keywords associated with the fault events in the fault event library. In addition, the present application can also realize the self-maintenance and automatic update of the candidate natural voice keywords in the fault event library, so that the candidate natural voice keywords in the fault event library adapt to the different description habits of users.
[0065] In addition, in the related art, fault events are generally located based on input keywords through a built-in large language model. However, for new keywords, the large language model needs to be trained before it can be used to locate fault events through new keywords. In the present application, the large language model is not used to locate fault events based on keywords, but the corresponding relationship between fault events and candidate natural language keywords is pre-set, and when the fault event is subsequently located through a combination of fault features, the new target natural language keyword is added to the candidate natural language keyword associated with the fault event. In this way, the problem in the related art that the large language model needs to be trained before it can be used to locate fault events through new keywords can be solved.
[0066] In some embodiments, see Figure 2 , Figure 2 A second flow chart of the fault handling method provided in the embodiment of the present application is provided. The candidate natural language keywords associated with each fault event in the fault event library also correspond to a search weight. The search weight corresponding to the candidate natural language keyword associated with a fault event is determined according to the number of matches between the candidate natural language keyword and the fault event. On this basis, the fault handling method further includes steps S210-S230:
[0067] S210. Obtain the retrieval weights corresponding to the candidate natural language keywords associated with each fault event in N consecutive self-test cycles, wherein the retrieval weight corresponding to a candidate natural language keyword associated with a fault event in a self-test cycle is determined based on the number of matches in which the fault event is located through the candidate natural language keyword in the self-test cycle; N is a positive integer greater than 1.
[0068] Exemplarily, for fault event A, its associated candidate natural language keywords include an, and the matching times of each candidate natural language keyword in a self-check cycle are φ(a)-φ(n) respectively; for one of the candidate natural language keywords b, its retrieval weight ω(b) in a self-check cycle can be calculated by the following calculation formula:
[0069]
[0070] It can be seen from the above calculation formula that the higher the number of matches of locating a fault event through a candidate natural language keyword, the higher the retrieval weight of the candidate natural language keyword.
[0071] It is worth mentioning that since the retrieval weight of the candidate natural language keyword is determined by the number of matches of the candidate natural language keyword, in this case, the fault event library also includes the number of matches of the candidate natural language keyword associated with each fault event; for example, it can be stored in the form of key-value pairs, such as "candidate natural language keyword-number of matches".
[0072] S220. Determine, according to the retrieval weights corresponding to the candidate natural language keywords associated with each fault event in N consecutive self-check cycles, a first candidate natural language keyword whose retrieval weight for the associated third fault event in N consecutive self-check cycles is less than a weight threshold.
[0073] It can be understood that, for a fault event, if the retrieval weight of a candidate natural language keyword in N consecutive self-test cycles is less than the weight threshold, it means that the number of times the fault event is retrieved through the candidate natural language keyword in N consecutive self-test cycles is small, which may indicate that the number of times the user uses the candidate natural language keyword to describe the associated fault event is small, which corresponds to the fact that it is not appropriate to use the candidate natural language keyword as the retrieval keyword for the fault event. Therefore, the candidate natural language keyword can be deleted from the candidate natural language keywords associated with the fault event.
[0074] S230: Delete the first candidate natural language keyword from the candidate natural language keywords associated with the third fault event.
[0075] Through the embodiments provided by the present application, in N consecutive self-test cycles, the first candidate natural language keyword whose retrieval weight is less than the weight threshold is deleted from the candidate natural language keywords associated with the third fault event, thereby realizing automatic cleaning of the fault event library, avoiding the candidate natural language keywords associated with each fault event in the fault event being too bloated, thereby reducing the retrieval efficiency, and making the candidate natural language keywords associated with each fault event more accurate, thereby improving the effectiveness of the retrieval; at the same time, the retrieval weight of each candidate natural language keyword in N consecutive self-test cycles can more accurately represent the retrieval ability of the candidate natural language keyword for the fault event than the retrieval weight in one self-test cycle. Therefore, whether deletion is needed is determined according to the retrieval weight of each candidate natural language keyword in N consecutive self-test cycles, thereby reducing the occurrence of accidental deletion.
[0076] In some embodiments, see Figure 3 , Figure 3 A third flow chart of the fault handling method provided in the embodiment of the present application is provided. On the basis that the fault event library includes the number of matches of the candidate natural language keywords, after step S120, the fault handling method may further include step S310:
[0077] S310. If the first matching result indicates that a unique first fault event is matched in the fault event library, the number of matches of the target natural language keyword for the first fault event is accumulated and added by 1 to update the retrieval weight of the target natural language keyword for the first fault event; and a maintenance analysis model applicable to the first fault event is pushed to the user.
[0078] It can be understood that when the first matching result indicates that a unique first fault event is matched in the fault event library, the maintenance analysis model applicable to the first fault event can be uniquely determined, and therefore, the maintenance analysis model applicable to the first fault event can be directly pushed to the user.
[0079] Obviously, when the first matching result indicates that a unique first fault event is matched in the fault event library, there is no need to perform a subsequent second matching.
[0080] In some embodiments, please refer to Figure 3 After step S150, the fault handling method may further include step S320:
[0081] S320: Initialize the number of matches of the target natural language keyword for the second fault event to 1, so as to update the retrieval weight of the target natural language keyword for the second fault event.
[0082] It should be noted that if the target natural language keyword exists among the candidate natural language keywords associated with the second fault event, it proves that the target natural language keyword is valid for the second fault event. Therefore, the number of matches of the target natural language keyword for the second fault event is accumulated and added by 1.
[0083] In some implementations, after step S140, the fault handling method may further include:
[0084] If the second matching result indicates that the second fault event associated with the target fault feature combination is not matched in the fault event library, a target fault event is created in the fault event library, the target fault feature combination is associated with the target fault event, the target natural language keyword is associated with the target fault event, and the number of matches of the target natural language keyword for the target fault event is initialized to 1 to update the retrieval weight of the target natural language keyword for the target fault event.
[0085] It is understandable that when the first vehicle encounters the same fault again, even if the target fault event may not be located through the target keyword, the target fault event can be located through the target fault feature combination, thereby ensuring the detection rate of the target fault event; at the same time, through the newly created target fault event, the fault events in the fault event library are automatically updated. As the fault processing proceeds, the fault event library will gradually be enriched, thereby improving the effect of fault processing based on the fault event library.
[0086] In some embodiments, if the second matching result indicates that the second fault event associated with the target fault feature combination is not matched in the fault event library, it means that there is no fault event in the current fault event library that can be uniquely determined to match the fault of the first vehicle. In this case, a prompt message that a recommendation cannot be made can be directly fed back to the user; or the fault event that is most similar to the fault of the first vehicle can be determined from the current fault event library. Since similar fault events can most likely share the same maintenance analysis model, the maintenance analysis model applicable to the fault event can be recommended to the user.
[0087] In some embodiments, see Figure 4 , Figure 4 A fourth flow chart of the fault handling method provided in the embodiment of the present application is provided. After step S140, the fault handling method may further include steps S410-S430:
[0088] S410: If the second matching result indicates that the second fault event associated with the target fault feature combination is not matched in the fault event library, calculate the feature similarity between the target fault feature combination and each fault feature combination in the fault event library.
[0089] In some embodiments, each fault feature in the target fault feature combination may be vectorized to obtain a feature vector of each fault feature, and then the feature vectors of all fault features in the target fault feature combination may be combined to obtain a feature vector sequence of the target fault feature combination. Each fault feature combination in the fault event library may be processed in a similar manner to obtain a feature vector sequence of each fault feature combination, and then the feature similarity between the target fault feature combination and the feature vector sequence of the fault feature combination is determined by calculating the similarity (e.g., Euclidean distance, cosine similarity, etc.) between the feature vector sequence of the target fault feature combination and the feature vector sequence of the fault feature combination.
[0090] In some embodiments, considering that a fault feature combination may contain multiple fault features, the feature similarity between the target fault feature combination and each fault feature combination in the fault event library can be determined based on the number of common fault features between the fault feature combination of each fault event and the target fault feature combination.
[0091] For example, see Figure 5 , Figure 5 The embodiment of the present application provides Figure 4 Flow chart of step S420, the fault feature combination includes one or more fault features, and step S420 includes steps S411-S412:
[0092] S411 , counting the number of common fault features between each fault feature combination and the target fault feature combination.
[0093] It can be understood that, the greater the number of common fault features between the fault feature combination and the target fault feature combination, the more similar the fault event associated with the fault feature combination is to the fault of the first vehicle.
[0094] S412: taking the ratio of the number of common fault features between each fault feature combination and the target fault feature combination to the number of reference fault features as the feature similarity between the corresponding fault feature combination and the target fault feature combination.
[0095] Among them, the number of reference fault features can be a fixed value or the number of fault features included in the target fault feature combination; obviously, when the number of reference fault features is the number of fault features included in the target fault feature combination, the feature similarity between the fault feature combination and the target fault feature combination is the proportion of the common fault features in the target fault feature combination.
[0096] S420: Determine in the fault event library a candidate fault feature combination whose feature similarity with the target fault feature combination exceeds a similarity threshold.
[0097] It can be understood that if the feature similarity between the candidate fault feature combination and the target fault feature combination exceeds the similarity threshold, it means that the fault event associated with the candidate fault feature combination is sufficiently similar to the fault of the first vehicle. Therefore, the maintenance analysis model suitable for the fault event can be recommended to the user.
[0098] S430: Pushing to the user a maintenance analysis model applicable to the fourth fault event associated with the candidate fault feature combination.
[0099] Among them, the determined candidate fault feature combination can be one or more. If there are multiple candidate fault feature combinations, a maintenance analysis model applicable to the fourth fault event associated with the multiple candidate fault feature combinations can be pushed to the user; or one candidate fault feature combination can be selected from the multiple candidate fault feature combinations (for example, one is selected randomly, or the one with the highest feature similarity with the target fault feature combination is selected), and then the maintenance analysis model applicable to the fourth fault event associated with the selected candidate fault feature combination is pushed to the user.
[0100] In some implementations, when the maintenance analysis model of the fourth fault event is pushed to the user, relevant prompt information may also be pushed, where the prompt information is used to explain to the user that the maintenance analysis model is not completely applicable to the current fault.
[0101] In some embodiments, see Figure 6 , Figure 6 A fifth flow chart of the fault handling method provided in an embodiment of the present application is provided. The candidate natural language keywords associated with each fault event in the fault event library also correspond to a search weight. The search weight corresponding to the candidate natural language keywords associated with a fault event is determined according to the number of matches of locating the fault event through the candidate natural language keywords; it is determined in the fault event library that there are multiple candidate fault feature combinations whose feature similarity with the target fault feature combination exceeds the similarity threshold, that is, there are multiple fourth fault events. In this case, before step S440, the fault handling method further includes steps S510-S520:
[0102] S510: Determine a push score for each fourth fault event according to at least one of the number of times the maintenance analysis model applicable to each fourth fault event is pushed and the search weight of the target natural language keyword for each fourth fault event.
[0103] In some embodiments, the push score of each fourth fault event can be determined only based on the number of times the maintenance analysis model of each fourth fault event is pushed. For example, a functional relationship between the push score and the number of pushes can be pre-set, so that the higher the number of pushes, the higher the push score. The push score of each fourth fault event can also be determined only based on the retrieval weight of the target natural language keyword for each fourth fault event. Similarly, a functional relationship between the push score and the retrieval weight can be pre-set, so that the higher the retrieval weight, the higher the push score.
[0104] In other implementations, a weighted sum may be taken based on the number of times the maintenance analysis model of the fourth fault event is pushed and the retrieval weight of the target natural language keyword for the fourth fault event, and the result of the weighted sum may be used as the push score of the fourth fault event.
[0105] In other embodiments, the push score of the fourth fault event may also include a first score and a second score, that is, the first score is determined based on the number of times the maintenance analysis model of each fourth fault event is pushed, and the second score is determined based on the retrieval weight of the target natural language keyword for each fourth fault event.
[0106] S520: Determine a fourth fault event with the highest push score among multiple fourth fault events.
[0107] In other embodiments, on the basis that the push score of the fourth fault event may include the first score and the second score, priorities of the first push score and the second score may be set, and scores with higher priorities are given priority to determine the fourth fault event with the highest push score.
[0108] Exemplarily, if the priority of the first score is higher than the second score, the fourth fault event with the highest first score is determined as the fourth fault event with the highest push score; if there are multiple fourth fault events with the highest first score, the fourth fault event with the highest second score is selected from them and determined as the fourth fault event with the highest push score.
[0109] On the basis of steps S510-S520, step S440 includes:
[0110] The maintenance analysis model applicable to the fourth fault event with the highest push score is pushed to the user.
[0111] Of course, if there are still multiple fourth fault events with the highest push scores that are finally determined, the relevant information of the maintenance analysis models applicable to the multiple fourth fault events can be pushed to the user, and the user selects one of the maintenance analysis models according to the relevant information of the maintenance analysis models.
[0112] For easier understanding, see Figure 7 , Figure 7 A schematic diagram of an application scenario involved in an embodiment of the present application is provided, including processes 1001-1010:
[0113] 1001, configure a fault event database. The configured fault event database includes multiple fault events and candidate natural language keywords associated with each fault event and a combination of fault features associated with each fault event, and may also include the number of matches of the candidate natural language keywords. It should be noted that the configuration of the fault database may be performed in advance.
[0114] 1002, obtaining a user fault description statement.
[0115] 1003, determine a target natural language keyword from the user's fault description statement, wherein the determined target natural language keyword may be one or more.
[0116] 1004 , judging whether a unique first fault event is matched in the fault event library according to the first matching result, if a unique first fault event is matched, executing step 1007 , if no unique first fault event is matched, executing step 1005 .
[0117] 1005, obtaining a target fault feature combination, wherein the target fault feature combination is a fault feature combination of the fault warning signal reported by the first vehicle.
[0118] 1006. Determine whether the second fault event is matched in the fault event library according to the second matching result. If the second fault event is matched, execute step 1007. If the second fault event is not matched, execute steps 1010 and 1009.
[0119] 1007, determine a reference fault event, wherein the reference fault event may be a first fault event uniquely determined according to the first matching result, or may be a second fault event uniquely determined according to the second matching result.
[0120] 1008 , pushing a maintenance analysis model applicable to the reference failure event to the user.
[0121] See also Figure 8 , Figure 8 A schematic diagram of maintaining and updating the candidate natural language keywords involved in the embodiment of the present application is provided, including adding a target natural language keyword, updating the number of matches of the target natural language keyword, and deleting the first candidate natural language keyword, as follows:
[0122] For newly added target natural language keywords, there are two situations: one is to add a target natural language keyword for the created target fault event, and the other is to add a target natural language keyword for the second fault event in which the target natural language keyword does not exist in the associated candidate natural language keywords.
[0123] After adding a target natural language keyword, it is also necessary to initialize the number of matches of the target natural language keyword to 1 in order to update the search weight of each candidate natural language keyword.
[0124] Regarding updating the number of matches of the target natural language keyword, updating the number of matches of the target natural language keyword also includes two situations: one is to increase the number of matches of the target natural language keyword associated with the only first fault event by 1; the other is to increase the number of matches of the target natural language keyword in the second fault event by 1 when the target natural language keyword exists in the candidate natural language keywords associated with the second fault event.
[0125] Similarly, by updating the number of matches of the target natural language keyword, it is convenient to update the search weight of each candidate natural language keyword.
[0126] As for deleting the first candidate natural language keyword, it is achieved through self-checking of the fault event library. The retrieval weight of the candidate natural language keyword associated with the fault event in N self-checking cycles of the fault event library is calculated. If the retrieval weight of the first candidate natural language keyword in the third fault event in N self-checking cycles is less than the weight threshold, the first candidate natural language keyword is deleted from the candidate natural language keywords associated with the third fault event.
[0127] In the process of using the fault event library, through the above-mentioned maintenance and update of candidate natural language keywords and creation of target fault events, automatic maintenance of the fault event library is achieved, making the retrieval based on the fault event library more efficient and accurate, and improving the effect of fault handling based on the fault event library.
[0128] In some embodiments, see Fig. 9 , Fig. 9 A schematic diagram of a fault handling device provided in an embodiment of the present application is provided. The fault handling device 500 includes:
[0129] The first acquisition module 510 is configured to acquire target natural language keywords from a user fault description sentence reported for the first vehicle.
[0130] The first matching module 520 is used to match the fault event associated with the target natural language keyword in the fault event library to obtain a first matching result. The fault event library contains multiple fault events and candidate natural language keywords associated with each fault event and a fault feature combination associated with each fault event.
[0131] The second acquisition module 530 is used to obtain the target fault feature combination of the fault alarm signal reported by the first vehicle if the first matching result indicates that no unique first fault event is matched in the fault event library; the first fault event refers to a fault event whose associated candidate natural language keywords include the target natural language keywords.
[0132] The second matching module 540 is used to match the fault event associated with the target fault feature combination in the fault event library to obtain a second matching result.
[0133] The self-update module 550 is used to add the target natural language keyword to the candidate natural language keywords associated with the second fault event if the second matching result indicates that a second fault event associated with the target fault feature combination is matched in the fault event library, and the target natural language keyword does not exist in the candidate natural language keywords associated with the second fault event.
[0134] The push module 560 is used to push the maintenance analysis model suitable for the second fault event to the user.
[0135] In some embodiments, the candidate natural language keywords associated with each fault event in the fault event library also correspond to a retrieval weight, and the retrieval weight corresponding to a candidate natural language keyword associated with a fault event is determined according to the number of matches in which the fault event is located through the candidate natural language keyword; the fault handling device 500 also includes a third acquisition module, which is used to obtain the retrieval weights corresponding to the candidate natural language keywords associated with each fault event in N consecutive self-test cycles, wherein the retrieval weight corresponding to the candidate natural language keyword associated with a fault event in a self-test cycle is determined according to the number of matches in which the fault event is located through the candidate natural language keyword in the self-test cycle; N is a positive integer greater than 1; a judgment module, which is used to determine the first candidate natural language keyword whose retrieval weight for the associated third fault event in N consecutive self-test cycles is less than the weight threshold according to the retrieval weights corresponding to the candidate natural language keywords associated with each fault event in N consecutive self-test cycles; a deletion module, which is used to delete the first candidate natural language keyword from the candidate natural language keywords associated with the third fault event.
[0136] In some embodiments, the self-update module 550 is also used to accumulate and add 1 to the number of matches of the target natural language keyword for the first fault event if the first matching result indicates that a unique first fault event is matched in the fault event library, so as to update the retrieval weight of the target natural language keyword for the first fault event; and the push module 560 is also used to push a maintenance analysis model applicable to the first fault event to the user.
[0137] In some implementations, the self-update module 550 is further configured to initialize the number of matches of the target natural language keyword for the second fault event to 1, so as to update the retrieval weight of the target natural language keyword for the second fault event.
[0138] In some embodiments, the fault handling device 500 also includes a creation module for creating a target fault event in the fault event library, associating the target fault feature combination with the target fault event, associating the target natural language keyword with the target fault event, and initializing the number of matches of the target natural language keyword for the target fault event to 1, so as to update the retrieval weight of the target natural language keyword for the target fault event if the second matching result indicates that the second fault event associated with the target fault feature combination is not matched in the fault event library.
[0139] In some embodiments, the fault handling device 500 also includes a similarity calculation module, which is used to calculate the feature similarity between the target fault feature combination and each fault feature combination in the fault event library if the second matching result indicates that the second fault event associated with the target fault feature combination is not matched in the fault event library; a similarity screening module, which is used to determine in the fault event library a candidate fault feature combination whose feature similarity with the target fault feature combination exceeds a similarity threshold; the push module 560 is also used to push to the user a maintenance analysis model applicable to the fourth fault event associated with the candidate fault feature combination.
[0140] In some embodiments, the candidate natural language keywords associated with each fault event in the fault event library also correspond to a retrieval weight, and the retrieval weight corresponding to the candidate natural language keywords associated with a fault event is determined according to the number of matches of locating the fault event through the candidate natural language keywords; there are multiple candidate fault feature combinations; there are multiple fourth fault events; the fault handling device 500 also includes a scoring module, which is used to determine the push score of each fourth fault event based on the number of times the maintenance analysis model applicable to each fourth fault event is pushed, and at least one of the retrieval weights of the target natural language keyword for each fourth fault event; a scoring screening module, which is used to determine the fourth fault event with the highest push score among multiple fourth fault events; the push module 560 is also used to push the maintenance analysis model applicable to the fourth fault event with the highest push score to the user.
[0141] In some embodiments, the fault feature combination includes one or more fault features; the similarity calculation module is also used to count the number of common fault features between each fault feature combination and the target fault feature combination; the ratio of the number of common fault features between each fault feature combination and the target fault feature combination to the number of reference fault features is used as the feature similarity between the corresponding fault feature combination and the target fault feature combination.
[0142] In some implementations, according to the fault handling method provided in the above embodiment, the embodiment of the present application also provides an electronic device, such as Fig.10 , Fig.10 A structural block diagram of an electronic device provided in an embodiment of the present application is given, where the electronic device 600 includes a processor 610; a memory 620; and computer-readable instructions are stored on the memory 620. When the computer-readable instructions are executed by the processor 610, the above method is implemented.
[0143] The electronic device 600 may be a terminal device, and the terminal device may be a vehicle-mounted terminal, a maintenance terminal, or the like.
[0144] The processor 610 may include one or more processing cores. The processor 610 uses various interfaces and lines to connect the various parts of the entire wearable device, and executes various functions of the wearable device and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 620, and calling data stored in the memory 620. Optionally, the processor 610 can be implemented in at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor 610 can integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing the content to be displayed; and the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor, but may be implemented separately through a communication chip.
[0145] The memory 620 may include a random access memory (RAM) or a read-only memory (ROM). The memory 620 may be used to store instructions, programs, codes, code sets or instruction sets. The memory 620 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the following various method embodiments, etc. The data storage area may also store data created by the electronic device during use.
[0146] In some embodiments, the present application further provides a computer-readable storage medium having computer-readable instructions stored thereon, and when the computer-readable instructions are executed by the processor 610, the above method is implemented.
[0147] The computer-readable storage medium may be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read-only memory), an EPROM, a hard disk, or a ROM. Optionally, the computer-readable storage medium includes a non-transitory computer-readable storage medium. The computer-readable storage medium has storage space for computer-readable instructions for executing any method step of the above method. These computer-readable instructions may be read from or written to one or more computer program products. The computer-readable instructions may be compressed in an appropriate form.
[0148] In particular, according to an embodiment of the present application, the process described above can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer instruction. When the computer instruction is executed by a central processing unit (CPU), various functions defined in the system of the present application are executed.
[0149] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program with a predetermined function, and works together with other related parts to achieve a predetermined goal, and can be implemented in whole or in part by using software, hardware (such as processing circuits or memories) or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be a part of the overall module or unit of the module or unit function.
[0150] The above are only preferred embodiments of the present application, and are not intended to limit the present application in any form. Although the present application has been disclosed as above with preferred embodiments, it is not intended to limit the present application. Any technical personnel in the field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present application. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present application without departing from the content of the technical solution of the present application are still within the scope of the technical solution of the present application.
Claims
1. A fault handling method, characterized in that: include: Acquire target natural language keywords from a user fault description sentence reported for the first vehicle; Matching a fault event associated with the target natural language keyword in a fault event library to obtain a first matching result, wherein the fault event library contains a plurality of fault events and candidate natural language keywords associated with each of the fault events and a combination of fault features associated with each of the fault events; If the first matching result indicates that no unique first fault event is matched in the fault event library, obtaining a target fault feature combination of the fault warning signal reported by the first vehicle; The first fault event refers to a fault event in which the associated candidate natural language keywords include the target natural language keyword; Matching the fault event associated with the target fault feature combination in the fault event library to obtain a second matching result; If the second matching result indicates that a second fault event associated with the target fault feature combination is matched in the fault event library, and the target natural language keyword does not exist in the candidate natural language keywords associated with the second fault event, adding the target natural language keyword to the candidate natural language keywords associated with the second fault event; The maintenance analysis model applicable to the second fault event is pushed to the user.
2. The method according to claim 1, characterized in that The candidate natural language keywords associated with each fault event in the fault event database also correspond to a search weight, and the search weight corresponding to a candidate natural language keyword associated with a fault event is determined according to the number of matches between the candidate natural language keyword and the fault event; The method further comprises: Obtaining the retrieval weights corresponding to the candidate natural language keywords associated with each of the fault events in N consecutive self-check cycles, wherein the retrieval weight corresponding to a candidate natural language keyword associated with a fault event in a self-check cycle is determined according to the number of matches in which the fault event is located by the candidate natural language keyword in the self-check cycle; N is a positive integer greater than 1; Determine, according to the retrieval weights respectively corresponding to the candidate natural language keywords associated with each of the fault events in N consecutive self-check cycles, a first candidate natural language keyword whose retrieval weight for the associated third fault event in N consecutive self-check cycles is less than a weight threshold; The first candidate natural language keyword is deleted from candidate natural language keywords associated with the third fault event.
3. The method according to claim 2, characterized in that After matching the fault event associated with the target natural language keyword in the fault event library to obtain a first matching result, the method further includes: If the first matching result indicates that a unique first fault event is matched in the fault event library, the number of matches of the target natural language keyword for the first fault event is accumulated and added by 1 to update the retrieval weight of the target natural language keyword for the first fault event; and a maintenance analysis model applicable to the first fault event is pushed to the user.
4. The method according to claim 2, characterized in that: After adding the target natural language keyword to the candidate natural language keywords associated with the second fault event, the method further includes: The number of matches of the target natural language keyword for the second fault event is initialized to 1 to update the retrieval weight of the target natural language keyword for the second fault event.
5. The method according to any one of claims 2 to 4, characterized in that: After matching the fault event associated with the target fault feature combination in the fault event library to obtain a second matching result, the method further includes: If the second matching result indicates that the second fault event associated with the target fault feature combination is not matched in the fault event library, a target fault event is created in the fault event library, the target fault feature combination is associated with the target fault event, the target natural language keyword is associated with the target fault event, and the number of matches of the target natural language keyword for the target fault event is initialized to 1 to update the retrieval weight of the target natural language keyword for the target fault event.
6. The method according to claim 1, characterized in that After matching the fault event associated with the target fault feature combination in the fault event library to obtain a second matching result, the method further includes: If the second matching result indicates that the second fault event associated with the target fault feature combination is not matched in the fault event library, calculating the feature similarity between the target fault feature combination and each fault feature combination in the fault event library; Determine in the fault event library a candidate fault feature combination whose feature similarity with the target fault feature combination exceeds a similarity threshold; A maintenance analysis model suitable for the fourth fault event associated with the candidate fault feature combination is pushed to the user.
7. The method according to claim 6, characterized in that The candidate natural language keywords associated with each of the fault events in the fault event library also correspond to a search weight, and the search weight corresponding to a candidate natural language keyword associated with a fault event is determined according to the number of matches of locating the fault event through the candidate natural language keyword; the candidate fault feature combination is multiple; the fourth fault event is multiple; Before pushing the maintenance analysis model applicable to the fourth fault event associated with the candidate fault feature combination to the user, the method further includes: Determine a push score for each of the fourth fault events according to the number of times the maintenance analysis model applicable to each of the fourth fault events is pushed and at least one of the search weights of the target natural language keyword for each of the fourth fault events; Determine a fourth fault event with the highest push score among the plurality of fourth fault events; The pushing to the user a maintenance analysis model applicable to the fourth fault event associated with the candidate fault feature combination includes: The maintenance analysis model applicable to the fourth fault event with the highest push score is pushed to the user.
8. The method according to claim 6, characterized in that The fault feature combination includes one or more fault features; The calculating the feature similarity between the target fault feature combination and each fault feature combination in the fault event library includes: Counting the number of common fault features between each of the fault feature combinations and the target fault feature combination; The ratio of the number of common fault features between each of the fault feature combinations and the target fault feature combination to the number of reference fault features is used as the feature similarity between the corresponding fault feature combination and the target fault feature combination.
9. A fault handling device, characterized in that: include: A first acquisition module, configured to acquire target natural language keywords from a user fault description statement reported for the first vehicle; A first matching module, configured to match a fault event associated with the target natural language keyword in a fault event library to obtain a first matching result, wherein the fault event library contains a plurality of fault events and candidate natural language keywords associated with each of the fault events and a combination of fault features associated with each of the fault events; a second acquisition module, configured to acquire a target fault feature combination of the fault warning signal reported by the first vehicle if the first matching result indicates that no unique first fault event is matched in the fault event library; The first fault event refers to a fault event in which the associated candidate natural language keywords include the target natural language keyword; A second matching module, used for matching the fault event associated with the target fault feature combination in the fault event library to obtain a second matching result; A self-update module, configured to add the target natural language keyword to the candidate natural language keywords associated with the second fault event if the second matching result indicates that a second fault event associated with the target fault feature combination is matched in the fault event library, and the target natural language keyword does not exist in the candidate natural language keywords associated with the second fault event; A push module is used to push the maintenance analysis model suitable for the second fault event to the user.
10. An electronic device, characterized in that: include: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method according to any one of claims 1 to 8 is implemented.
11. A computer-readable storage medium, characterized in that: Computer-readable instructions are stored thereon, and when the computer-readable instructions are executed by a processor, the method according to any one of claims 1 to 8 is implemented.