A vehicle intelligent auxiliary diagnosis method, server, system and storage medium

Abstract

The application discloses a vehicle intelligent auxiliary diagnosis method and system and a storage medium, and belongs to the technical field of intelligent cloud diagnosis. The method mainly comprises the following steps: if the fault information comprises a fault code, then in a pre-constructed fault knowledge graph, a first fault cause set and a first solution set are determined; if the fault information does not comprise a fault code, then according to the belonging relationship between a first fault phenomenon in the fault information and the fault knowledge graph, a fault phenomenon entity weight corresponding to the first fault phenomenon and a second fault cause are determined, and according to the number of a second fault phenomenon corresponding to the second fault cause, the belonging degree weight of the second fault cause is determined; according to the fault phenomenon entity weight and the belonging degree weight, the correlation degree weight of the second fault cause is determined, and finally, the second fault cause set and the corresponding second solution set are determined. The application improves the fault coverage rate and the fault diagnosis efficiency of vehicle fault diagnosis, and improves the situation that the personal maintenance experience is insufficient.

Description

Technical Field

[0001] This application relates to the field of intelligent cloud diagnostic technology, and in particular to a vehicle intelligent auxiliary diagnostic method, server, system and storage medium. Background Technology

[0002] Current vehicle fault diagnosis methods mainly involve repair workers reading fault codes by connecting to the vehicle's OBD interface. Based on the obtained fault codes, repair workers consult fault repair manuals or rely on their personal experience to repair the vehicle.

[0003] Current technological advancements have led to increasingly sophisticated vehicle performance and more complex structures. As vehicle parts are updated, troubleshooting solutions may need to be modified, fault manuals may not be updated in a timely manner, and manual lookup may be inefficient. Fault codes may not even exist. Repair technicians can only diagnose and repair vehicles based on the symptoms, but the experience of repair technicians varies greatly, and the range of faults they are familiar with is limited, making it difficult to guarantee the quality of vehicle repairs. Summary of the Invention

[0004] This application mainly provides a vehicle intelligent auxiliary diagnostic method, server, system, and storage medium to solve the problems of untimely updates to fault manuals, low efficiency in querying manuals, low fault coverage, inaccurate fault cause analysis due to limited repair experience of repair workers, and lack of differentiation in the degree of relevance of the analyzed fault causes when diagnosing vehicle faults based on fault symptoms.

[0005] To achieve the above objectives, one technical solution adopted in this application is: providing a vehicle intelligent assisted diagnosis method, which includes: analyzing received fault information; if the fault information includes fault codes, determining a first set of fault causes and a first set of solutions based on the fault codes in a pre-constructed fault knowledge graph; if the fault information does not include fault codes, determining the fault phenomenon entity weight corresponding to the first fault phenomenon based on the attribution relationship between the first fault phenomenon in the fault information and the fault knowledge graph; determining a second fault cause corresponding to the first fault phenomenon in the fault knowledge graph, and determining the attribution weight of the second fault cause based on the number of second fault phenomena corresponding to the second fault cause; determining the relevance weight of the second fault cause based on the fault phenomenon entity weight and the attribution weight; determining a second set of fault causes based on the relevance weight, and determining a corresponding second set of solutions based on the second set of fault causes.

[0006] Another technical solution adopted in this application is: providing a vehicle intelligent auxiliary diagnostic server, which includes: a fault information analysis module, used to analyze received fault information; if the fault information includes fault codes, then in a pre-constructed fault knowledge graph, determining a first set of fault causes and a first set of solutions based on the fault codes; an entity weight determination module, used to determine the entity weight of the fault phenomenon corresponding to the first fault phenomenon based on the attribution relationship between the first fault phenomenon in the fault information and the fault knowledge graph if the fault information does not include fault codes; an attribution weight determination module, used to determine the second fault cause corresponding to the first fault phenomenon in the fault knowledge graph, and determine the attribution weight of the second fault cause based on the number of second fault phenomena corresponding to the second fault cause; a relevance weight determination module, used to determine the relevance weight of the second fault cause based on the entity weight of the fault phenomenon and the attribution weight; and a solution determination module, used to determine the second set of fault causes based on the relevance weight, and determine the corresponding second set of solutions based on the second set of fault causes.

[0007] Another technical solution adopted in this application is: providing a vehicle intelligent auxiliary diagnostic system, which includes: a client for inputting fault information, the fault information including a first fault phenomenon and / or fault code; and a server for executing the vehicle intelligent auxiliary diagnostic method in Solution 1, and returning the set of fault causes, the set of solutions, and / or the set of fault phenomena obtained by analyzing the fault information according to the fault knowledge graph to the client.

[0008] Another technical solution adopted in this application is to provide a computer-readable storage medium storing computer instructions that are operated to execute the vehicle intelligent auxiliary diagnosis method in Solution 1.

[0009] The beneficial effects of the technical solution in this application are as follows: This application designs a vehicle intelligent auxiliary diagnostic method, server, system, and storage medium. This method, through the construction of a fault knowledge graph, uniformly organizes fault manuals and historical repair records, summarizes a large amount of experience and knowledge from repair workers, strengthens the correlation between data, provides more comprehensive fault coverage, improves fault diagnosis efficiency, reduces error rates, and addresses the issue of insufficient individual repair experience. This method combines the attribution of fault phenomena to fault causes with fault causes that are highly correlated with the fault phenomena, facilitating faster location of vehicle faults. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a schematic flowchart of a vehicle intelligent assisted diagnosis method provided in Embodiment 1 of this application;

[0012] Figure 2 This is a flowchart of the method for constructing a fault knowledge graph provided in Embodiment 2 of this application;

[0013] Figure 3 This is a schematic diagram of the topology of the fault manual knowledge graph provided in Embodiment 2 of this application;

[0014] Figure 4 This is a schematic diagram of the topology of the fault repair record knowledge graph provided in Embodiment 2 of this application;

[0015] Figure 5 This is a schematic diagram of the topology of the fault knowledge graph provided in Embodiment 2 of this application;

[0016] Figure 6 This is a schematic diagram of the structure of a vehicle intelligent auxiliary diagnostic server provided in Embodiment 3 of this application;

[0017] Figure 7 This is a schematic diagram of the structure of a vehicle intelligent auxiliary diagnostic system provided in Embodiment 4 of this application.

[0018] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0019] The preferred embodiments of this application will now be described in detail with reference to the accompanying drawings, so that the advantages and features of this application can be more easily understood by those skilled in the art, thereby providing a clearer and more definite definition of the scope of protection of this application.

[0020] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0021] In existing technologies, vehicle fault diagnosis can be performed using an automotive diagnostic tool to test the vehicle, with repair data displayed synchronously for easy retrieval. However, this method directly returns historical repair data based on fault codes, and cannot handle cases where fault codes are not present. Vehicle fault diagnosis can also be performed by extracting predetermined features related to the fault and training a fault classification model. The fault classification model is then used to process the fault information to be diagnosed and the predetermined features to generate a fault diagnosis result for the fault to be diagnosed. However, this method requires pre-training a ranking model. If it is a multi-classification model, there will be data imbalance, causing the model's performance to be biased towards common faults. If a model is trained for each fault, there will be too many models, making maintenance difficult. At the same time, uncommon faults may not be able to be trained due to insufficient data.

[0022] This application proposes a vehicle intelligent assisted diagnostic method to solve the aforementioned technical problems. It calculates the relative relevance of each fault phenomenon, achieving simultaneous querying and updating without model training when data is updated. Specifically, the method includes: analyzing received fault information; if the fault information includes a fault code, determining a first set of fault causes and a first set of solutions based on the fault code in a pre-constructed fault knowledge graph; firstly, analyzing the received fault information, which can be a specific fault code and / or fault phenomenon. When the fault information is a specific fault code, directly querying the first set of fault causes and the first set of solutions related to the input fault code in the pre-constructed fault knowledge graph based on the set of relationships between the fault code and fault causes, and the set of relationships between the fault code and solutions. For example, a repair worker can read the fault code of a faulty vehicle through the OBD interface and obtain the corresponding fault cause and solution by consulting the fault knowledge graph. If the fault information does not include a fault code, the entity weight of the fault phenomenon corresponding to the first fault phenomenon is determined based on its association with the fault knowledge graph. When the fault information is a definite fault phenomenon (i.e., it does not include a fault code), the number of first fault phenomena belonging to the fault knowledge graph is determined by whether the input first fault phenomenon corresponds to existing fault phenomena in the fault knowledge graph, thereby determining the entity weight of the fault phenomenon corresponding to the first fault phenomenon in the fault knowledge graph. In the fault knowledge graph, the second fault cause corresponding to the first fault phenomenon is determined, and the association weight of the second fault cause is determined based on the number of second fault phenomena corresponding to the second fault cause. Specifically, the corresponding second fault cause is found based on the first fault phenomenon belonging to the fault knowledge graph, and then the number of corresponding second fault phenomena in the fault knowledge graph is found using the second fault cause, thereby determining the association weight of the second fault cause. The relevance weight of the second fault cause is determined based on the entity weight and association weight; the obtained entity weight and association weight are then substituted into the relevance calculation formula to obtain the relevance weight of the second fault cause. Based on relevance weights, a second set of fault causes is determined, and a corresponding second set of solutions is then identified. For example, the relevance weights can be sorted, and fault causes and solutions with higher relevance weights are prioritized for display to repair personnel. By combining the degree of attribution of fault causes to fault phenomena, the relevance of vehicle faults can be calculated, facilitating faster location of vehicle malfunctions.

[0023] The vehicle intelligent assisted diagnosis method provided in this application is applicable to the following scenarios: This application uses a constructed fault knowledge graph and combines the fault cause to calculate the relevance of the fault phenomenon, thereby assisting maintenance workers in diagnosing and repairing vehicle faults.

[0024] The inventive concept of this application is as follows: This application unifies the fault manual and historical maintenance records by constructing a fault knowledge graph, and summarizes a large amount of experience and knowledge of maintenance workers, so as to strengthen the correlation between data and make the fault coverage more comprehensive; This application calculates the correlation of fault phenomena by combining the fault cause with the degree of attribution, which makes it easier to locate vehicle faults more quickly.

[0025] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0026] Example 1

[0027] Figure 1 This is a flowchart illustrating a vehicle intelligent assisted diagnostic method provided in Embodiment 1 of this application, as shown below. Figure 1 As shown, the method in this embodiment may include:

[0028] S101, Analyze the received fault information. If the fault information includes fault codes, determine the first set of fault causes and the first set of solutions based on the fault codes in the pre-built fault knowledge graph.

[0029] In this embodiment, a fault knowledge graph is constructed in advance. When a fault code is available, the system can directly query and output the first set of fault causes and the first set of solutions related to the fault code based on the fault code input by the user, thereby improving the efficiency of fault diagnosis.

[0030] In one optional embodiment of this application, a first set of fault causes and a first set of solutions are determined based on fault codes in a pre-built fault knowledge graph. This includes determining a first set of fault causes corresponding to fault codes based on the entity relationship set between fault codes and fault causes in the pre-built fault knowledge graph; and determining a first set of solutions corresponding to fault codes based on the entity relationship set between fault codes and solutions. This improves fault diagnosis efficiency and reduces the error rate.

[0031] In this optional embodiment, in a pre-built fault knowledge graph, the correspondence between fault codes and fault causes, and between fault codes and solutions, is determined in advance based on historical vehicle repair records and experience. Then, a "fault code-fault cause" entity relationship set is constructed based on the "fault code" entity and the "fault cause" entity; a "fault code-solution" entity relationship set is constructed based on the "fault code" entity and the "solution" entity; and a "fault code-fault phenomenon" entity relationship set is constructed based on the "fault code" entity and the "fault phenomenon" entity. It should be noted that the constructed entity relationship sets can be updated based on actual vehicle fault diagnosis data. In the "fault code-fault cause" entity relationship set, based on the input fault code, a related first fault cause is retrieved, resulting in a first fault cause set, which contains one or more first fault cause entities corresponding to the input fault code. In the "fault code-solution" entity relationship set, based on the input fault code, a related first solution is retrieved, resulting in a first solution set, which contains one or more solution entities corresponding to the input fault code. In addition, the fault knowledge graph can output one or more fault phenomena that correspond to the input fault code in the "fault code-fault phenomenon" entity relationship set, which greatly facilitates the work of maintenance workers.

[0032] Specifically, when using fault code lookup, upon receiving the fault code, the system outputs a pre-built fault knowledge graph containing a set of fault phenomena S, a set of first fault causes R, and a set of first solutions M, with the following expressions:

[0033] S = {s|<code, s>∈R} cs}

[0034] R = {r | <code, r> ∈ R} cr}

[0035] M = {m | <code, m> ∈ R} cm}

[0036] Here, 'code' represents the input fault code. In the fault knowledge graph, 's' represents the fault phenomenon entity, 'r' represents the fault cause entity, and 'm' represents the solution entity.<code,s> This represents the entity relationship between the input fault code and the fault phenomenon entity.<code,r> This represents the entity relationship between the entered fault code and the entity indicating the cause of the fault.<code,m> This indicates the relationship between the input fault code and the solution entity. Based on the input fault code, it queries and returns the fault phenomenon entity node s, the fault cause entity node r, and the solution entity node m that are related to the fault code node.

[0037] S102, if the fault information does not include a fault code, then determine the weight of the fault phenomenon entity corresponding to the first fault phenomenon based on the attribution relationship between the first fault phenomenon in the fault information and the fault knowledge graph.

[0038] In this embodiment, a fault phenomenon entity weight is set for the first fault phenomenon in the first fault phenomenon set, so as to facilitate subsequent inference of related fault causes based on the first fault phenomenon.

[0039] In an optional embodiment of this application, the fault phenomenon entity weight corresponding to the first fault phenomenon is determined based on the attribution relationship between the first fault phenomenon in the fault information and the fault knowledge graph. This includes determining the number of first fault phenomena belonging to the fault knowledge graph in the set of first fault phenomena in the fault information, wherein the set of first fault phenomena contains first fault phenomena; and setting the fault phenomenon entity weight corresponding to the first fault phenomenon based on the number, wherein the more the number, the smaller the fault phenomenon entity weight.

[0040] In this optional embodiment, the first fault phenomenon belonging to the fault knowledge graph is assigned a fault phenomenon entity weight, and the fault phenomenon entity weight of the first fault phenomenon not belonging to the fault knowledge graph is initialized to zero, ensuring that the sum of the fault phenomenon entity weights corresponding to all first fault phenomena in the first fault phenomenon set is 1.

[0041] Specifically, when the fault information includes a first fault phenomenon, the first fault phenomenon set S' is received, and the expression for the weight of the corresponding fault phenomenon entity in the first fault phenomenon set is as follows:

[0042]

[0043] Where s represents the entity of the fault phenomenon, S represents the set of fault phenomena stored in the fault knowledge graph, and W s The initial weight of the fault phenomenon entity is |S∩S'|, which represents the number of fault phenomenon entities in the intersection of the first fault phenomenon set and the fault phenomenon set existing in the fault knowledge graph. That is, the number of first fault phenomena in the first fault phenomenon set belonging to the first fault phenomenon in the fault knowledge graph is denoted as the number of first existing fault phenomena.

[0044] In a specific instance, the intersection of the first set of fault phenomena and the set of fault phenomena stored in the fault knowledge graph is taken, i.e., S∩S'. If some of the first fault phenomenon entities in the first set of fault phenomena belong to the above intersection, then the reciprocal of the number of fault phenomenon entities belonging to the intersection, i.e., the reciprocal of the number of first existing fault phenomena, is set as the weight of the fault phenomenon entity corresponding to the first fault phenomenon in the first set that belongs to the intersection. If some of the first fault phenomenon entities in the first set of fault phenomena do not belong to the above intersection, then the weight of the fault phenomenon entity corresponding to the first fault phenomenon that does not belong to the intersection is set to 0, thereby ensuring that the sum of the weights of all the fault phenomenon entities corresponding to the first fault phenomena is 1.

[0045] In this specific example, if the first fault phenomenon set S' contains five first fault phenomenon entities, and assuming that three of the five first fault phenomenon entities belong to the aforementioned intersection, then the fault phenomenon entity weights of these three first fault phenomenon entities are all set to 1 / 3, i.e., W. s’1 =W s’2 =W s’3 =1 / 3, and the weights of the other two first fault phenomenon entities that do not belong to the above intersection are set to 0, i.e., W s’4 =W s’5 =0. Therefore, for the first set of fault phenomena S', the weights of the fault phenomenon entities in set S' are all equal.

[0046] S103, in the fault knowledge graph, determine the second fault cause corresponding to the first fault phenomenon, and determine the attribution weight of the second fault cause based on the number of second fault phenomena corresponding to the second fault cause.

[0047] In this embodiment, the first fault phenomenon in the first fault phenomenon set is directly queried in the fault knowledge graph. There may be a large number of related second fault causes. If it is necessary to distinguish the importance of each queried second fault cause to the first fault phenomenon set, it is necessary to first calculate the degree weight of the second fault cause to the first fault phenomenon set.

[0048] In one optional embodiment of this application, determining the attribution weight of the second fault cause based on the number of second fault phenomena corresponding to the second fault cause includes: determining the confidence level of the second fault cause to the first fault phenomenon based on the number of second fault phenomena in the fault knowledge graph, wherein the more second fault phenomena there are, the lower the confidence level; and determining the attribution weight of the second fault cause to the first fault phenomenon based on the confidence level, wherein the higher the confidence level, the higher the attribution weight.

[0049] In this optional embodiment, the first fault phenomena in the first fault phenomenon set are directly queried in the fault knowledge graph to find a large number of related second fault causes. For each second fault cause, the fault phenomena are directly queried in the fault knowledge graph to obtain each second fault phenomenon that corresponds to each second fault cause. Based on the number of each second fault phenomenon, the confidence level of each second fault cause to the first fault phenomenon set is determined. The more second fault phenomena that correspond to each second fault cause, the lower the confidence level of each second fault cause to the first fault phenomenon set. The ratio of the confidence level of each second fault cause to the sum of the confidence levels of all second fault causes is used as the belonging weight of each second fault cause to the first fault phenomenon set.

[0050] Specifically, the formula for calculating the weight of the second fault cause in relation to the set of first fault phenomena is as follows:

[0051]

[0052] Where S' represents the set of first fault phenomena, and R represents the second fault cause.<S’,R> R represents the relationship between the first fault phenomenon and the second fault cause in the first set of fault phenomena. sr W represents the set of relationships between fault phenomena, fault causes, fault descriptions, and fault diagnoses in a fault knowledge graph. S’R W represents the weight of the second fault cause R in relation to the set of first fault phenomena S'. RS’ This represents the confidence level of the second fault cause R relative to the set of first fault phenomena S'. The formula for calculating the confidence level of the second fault cause R is as follows:

[0053]

[0054] Where S" represents the second fault phenomenon in the fault knowledge graph that corresponds to the second fault cause, and R represents the second fault cause. sr This represents the set of relationships between fault phenomena, causes, descriptions, and diagnoses within a fault knowledge graph.<S”,R> This represents the entity relationship between the second fault phenomenon S” and the second fault cause R, where {<S”,R>|<S”,R>∈R sr | represents the number of second fault phenomena that are related to the second fault cause R. Confidence level W RS’ The value is the reciprocal of the number of second fault phenomena that are related to the second fault cause R. In other words, the more second fault phenomena associated with the second fault cause R, the lower the confidence level.

[0055] In a specific example, suppose there are two first fault phenomenon entities, S'1 and S'2, in the first fault phenomenon set S'. Suppose that S'1 and S'2 also exist in a fault knowledge graph. Suppose that these two first fault phenomenon entities can directly query four second fault causes in the fault knowledge graph, namely R1, R2, R3, and R4. Suppose that R1, R2, and R3 can be found from the first fault phenomenon entity S'1, and R1, R3, and R4 can be found from the first fault phenomenon entity S'2. For the second fault cause R1, suppose the number of related second fault phenomena found by the second fault cause R1 in the fault knowledge graph is 5. Then, what is the confidence level W of the second fault cause R1 towards the first fault phenomenon entity S'1? R1S’1 The confidence level W of the second fault cause R1 to the first fault phenomenon entity S'2 R1S’2 Both are 1 / 5, meaning that the confidence level of the second fault cause R1 for the first fault phenomenon set S' is only one, W R1S’ =1 / 5. Similarly, for the second fault cause R3, although the second fault cause R3 corresponds to two first fault phenomenon entities S'1 and S'2 of the first fault phenomenon set, the number of related second fault phenomena found by the second fault cause R3 in the fault knowledge graph is fixed, assuming it is 7. Therefore, whether it is the confidence of the second fault cause R3 to the first fault phenomenon entity S'1 or the confidence of the second fault cause R3 to the first fault phenomenon entity S'2, both of these confidences are the confidence of the second fault cause R3 to the first fault phenomenon set S', that is, its confidence is equal to 1 / 7.

[0056] In this specific example, for the first set of fault phenomena S', a second fault cause corresponds to a confidence level. Similarly, when calculating the attribution weight of a second fault cause, a second fault cause also corresponds to an attribution weight. To calculate the attribution weight of the second fault cause R1 to the first fault phenomenon entity S'1, the total confidence is obtained by summing the confidence levels of the second fault cause R1, the second fault cause R2, the second fault cause R3, and the second fault cause R4. The attribution weight of the second fault cause R1 to the first fault phenomenon entity S'1 can be obtained by calculating the ratio of the confidence level of the second fault cause R1 to the first fault phenomenon entity S'1 to the total confidence level. Since the confidence level of the second fault cause R1 to the first fault phenomenon entity S'1 is equal to the confidence level of the second fault cause R1 to the first fault phenomenon entity S'2, both the attribution weights of the second fault cause R1 to the first fault phenomenon entity S'1 and the attribution weights of the second fault cause R1 to the first fault phenomenon entity S'2 are the attribution weights of the second fault cause R1 to the first fault phenomenon set S'.

[0057] S104. Determine the relevance weight of the second fault cause based on the entity weight of the fault phenomenon and the attribution weight.

[0058] In this embodiment, the second fault cause is combined with the weighted inference of the degree of belonging to the first fault phenomenon set and the second fault cause with a high degree of correlation with the first fault phenomenon set. The correlation weight between the second fault cause and the first fault phenomenon set is first determined so as to obtain the second fault cause with a high degree of correlation with the first fault phenomenon set by using the correlation weight.

[0059] Specifically, the formula for calculating the relevance weight of the second cause of failure is as follows:

[0060]

[0061] Where S' represents the set of first fault phenomena, and R represents the second fault cause.<S’,R> R represents the entity relationship between the first fault phenomenon and the second fault cause in the first set of fault phenomena. sr W represents the set of entity relationships in a fault knowledge graph after the fusion of fault phenomena, fault causes, fault descriptions, and fault diagnoses. R W is the relevance weight for the second cause of failure. S’ W represents the weight of the fault phenomenon entities in the first set of fault phenomena. S’R This represents the weight of the second fault cause R in relation to the first fault phenomenon set S'.

[0062] In a specific example, under the assumptions of the above-mentioned instance of fault phenomenon entity weights, that is, the first fault phenomenon set S' contains five first fault phenomenon entities, namely S'1, S'2, S'3, S'4, and S'5, where the corresponding fault phenomenon entity weights are W respectively. s’1 =W s’2 =W s’3 =1 / 3, W s’4 =W s’5 =0. Assuming that the three first fault phenomenon entities existing in the above intersection can directly find six related second fault causes in the fault knowledge graph, namely R1, R2, R3, R4, R5, and R6, based on the confidence analysis of these six second fault causes, the corresponding attribution weights of the six second fault causes can be obtained, i.e., W. S’R1 W S’R2 W S’R3 W S’R4 W S’R5 and W S’R6When calculating the relevance weight of each second fault cause, first determine the first fault phenomenon entity corresponding to the current second fault cause in the first fault information set S'. There may be one, two, or three corresponding first fault phenomenon entities. Multiply the first fault phenomenon entity corresponding to each second fault cause with the attribution weight of the second fault cause and then add them together to obtain the relevance weight of each second fault cause.

[0063] For example, if the second fault cause R1 corresponds to two first fault phenomenon entities, namely S'1 and S'2, then the relevance weight of the second fault cause R1 to the set of first fault phenomena S' is equal to the product of the attribution weights of the first fault phenomenon entity S'1 and the second fault cause R1 plus the product of the attribution weights of the first fault phenomenon entity S'2 and the second fault cause R1, that is, the relevance weight W of the second fault cause R1. R1 =W S’1 *W S’R1 +W S’2 *W S’R1 If the second fault cause R5 corresponds to three first fault phenomenon entities, namely S'1, S'2, and S'3, then the relevance weight of the second fault cause R5 to the set of first fault phenomena S' is equal to the product of the attribution weights of the first fault phenomenon entity S'1 and the second fault cause R5, plus the product of the attribution weights of the first fault phenomenon entity S'2 and the second fault cause R5, plus the product of the attribution weights of the first fault phenomenon entity S'3 and the second fault cause R5, which is the relevance weight W of the second fault cause R1. R5 =W S’1 *W S’R5 +W S’2 *W S’R5 +W S’3 *W S’R5 .

[0064] S105, determine the second set of fault causes based on the relevance weights, and determine the corresponding second set of solutions based on the second set of fault causes.

[0065] In this embodiment, multiple second fault causes most relevant to the first fault phenomenon set are obtained by weighting the relevance. Based on these multiple second fault causes, the most likely solution to the fault is found in the fault knowledge graph, which helps maintenance workers to locate and resolve the fault more quickly.

[0066] In one optional embodiment of this application, a second set of fault causes is determined based on the relevance weights, and a corresponding set of second solutions is determined based on the second set of fault causes. This includes sorting the relevance weights by size, selecting a certain number of second fault causes based on the sorting results, and obtaining a second set of fault causes; and determining a corresponding set of second solutions in the fault knowledge graph based on the second set of fault causes.

[0067] In this optional embodiment, by sorting the relevance weights of the second fault causes, not only are multiple second fault causes with high relevance to the first fault phenomenon set obtained, but also the degree of relevance of each second fault cause to the first fault phenomenon set is determined. This facilitates repair personnel in diagnosing vehicle faults based on the most probable fault causes. The corresponding second solution set is then retrieved from the fault knowledge graph based on the most probable second fault cause set, and the vehicle fault is repaired according to the most probable second solution set, thus improving the situation where repair workers lack comprehensive repair experience.

[0068] Specifically, the relevance weights are sorted in descending order. The top N weights with the closest values ​​are selected as a certain number. For example, if there are 10 relevance weights, they are sorted in descending order. If the first four relevance weights have relatively large values ​​and are not significantly different from each other, while the last six relevance weights have relatively small values ​​and are significantly different from the first four, then four of these are selected. Based on this experience, a certain number can be selected, or a certain number can be selected according to user requirements. Alternatively, a relevance threshold can be set, and each relevance weight can be compared; weights greater than the threshold are selected as a certain number.

[0069] In a specific instance, according to the relevance weight W R Sort the second fault causes in descending order, select the top N causes, and generate a set E' of the second fault causes. R According to the second set of fault causes E' R Search for the relevant set of secondary solutions E' in the fault knowledge graph. M :

[0070] E' M ={m|<r,m>∈R rm ,r∈E' R}

[0071] Where r represents the fault cause entity in the fault knowledge graph, and m represents the solution entity in the fault knowledge graph.<r,m> This represents the relationship between the fault cause entity r and the solution entity m in the fault knowledge graph. It queries the solution entity nodes that are related to and connected to the fault cause entity.

[0072] In this specific example, a diagnostic result E' of this method is represented as follows:

[0073] E′=E′ R ∪E′ M

[0074] Among them, E' R For the second set of causes of failure, E' M This is the second set of solutions.

[0075] In the vehicle intelligent assisted diagnosis method of this application, existing technologies perform fault diagnosis by training models. This typically requires pre-training a ranking model. If it is a multi-classification model, data imbalance can occur, causing the model's performance to favor common faults. If a model is trained for each fault, there will be too many models, making maintenance difficult. Furthermore, uncommon faults may lack sufficient data for training. Existing technologies only support one diagnostic method: fault description diagnosis. This application avoids all of the above situations and does not rely on any model. It not only supports fault code diagnosis but also supports diagnosis based on fault phenomena when fault codes are absent. It calculates the relevance of each fault's path, achieving simultaneous query and update effects without needing to train a model when data is updated.

[0076] Example 2

[0077] Figure 2 This is a flowchart illustrating a vehicle intelligent assisted diagnostic method provided in Embodiment 2 of this application, as shown below. Figure 2 As shown, the method flow in this embodiment may further include:

[0078] S201. Based on the fault manual and the vehicle's fault history repair records, construct corresponding fault manual knowledge graphs and fault repair record knowledge graphs respectively.

[0079] In this embodiment, a fault manual knowledge graph is compiled based on the fault manual, and a fault repair record knowledge graph is compiled based on the vehicle fault history repair records. These are used to comprehensively summarize the experience of repairing faults and prevent omissions.

[0080] S202, the fault manual knowledge graph and the fault repair record knowledge graph are merged to obtain the fault knowledge graph.

[0081] In this embodiment, the separately constructed fault manual graph and fault repair record knowledge graph are merged into a single fault knowledge graph. This allows the fault knowledge graph to uniformly organize fault manuals and historical repair experience, summarize a large amount of experience and knowledge from repair workers, strengthen the correlation between data, and improve fault coverage.

[0082] In an optional embodiment of this application, the process of constructing corresponding fault manual knowledge graphs and fault repair record knowledge graphs based on fault manuals and vehicle fault history repair records includes: extracting first-category entities from the fault manual, analyzing first-category entity relationships between first-category entities, wherein the first-category entities include fault code entities, fault phenomenon entities, fault cause entities, and solution entities; constructing the fault manual knowledge graph based on the extracted first-category entities and the analyzed first-category entity relationships; and extracting second-category entities from the vehicle fault history repair records, analyzing second-category entity relationships between second-category entities, wherein the second-category entities include fault description entities, fault diagnosis entities, and repair item entities; and constructing the fault repair record knowledge graph based on the extracted second-category entities and the analyzed second-category entity relationships. The first-category entities and their relationships stored in the fault manual knowledge graph facilitate repair workers' querying of vehicle faults using fault codes; the fault repair record knowledge graph summarizes a large amount of repair experience, improving the situation where repair workers lack comprehensive repair experience.

[0083] In this optional embodiment, a first category of entities is extracted from the fault manual, namely, fault code entity, fault phenomenon entity, fault cause entity, and solution entity. A first entity relationship is established between the first category of entities, that is, the fault phenomenon entity is queried based on the fault code entity, the fault cause entity is queried based on the fault code entity, the solution entity is queried based on the fault code entity, the fault cause entity is queried based on the fault phenomenon entity, and the solution entity is queried based on the fault cause entity. A fault manual knowledge graph is constructed based on the first category of entities and the first entity relationship, making it more convenient and efficient to query using the input fault code.

[0084] It should be noted that the fault code entities extracted from the fault manual are used to construct the fault manual knowledge graph. When using the fault knowledge graph, the fault codes in the received fault information are read through the OBD interface. The essence of the fault codes in the fault information is a series of numbers composed of letters, numbers, and symbols, which is consistent with the essence of the fault code entities, but the two have different origins.

[0085] Specifically, extract four types of first-category entities, namely "fault code", "fault symptom", "fault cause", and "solution" from the fault manual, and obtain the first-category entity set E1; construct five first-entity relationships, namely "fault code - fault symptom", "fault code - fault cause", "fault code - solution", "fault symptom - fault cause", and "fault cause - solution" based on these four first-category entities, and obtain the first-entity relationship set R1. Construct the fault manual knowledge graph G1(E1, R1) based on these four first-category entities and these five first-entity relationships. The expressions of the first-category entity set E1 and the first-entity relationship set R1 are as follows:

[0086] E1 = {e | e: fault code} ∪ {e | e: fault symptom} ∪ {e | e: fault cause} ∪ {e | e: solution}

[0087] R1 = {<e1, e2> | e1: fault code, e2: fault symptom} ∪ {<e1, e2> | e1: fault code, e2: fault cause}

[0088] } ∪ {<e1, e2> | e1: fault code, e2: solution} ∪ {<e1, e2〉 | e1: fault symptom, e2: fault cause}

[0089] ∪ {<e1, e2〉 | e1: fault cause, e2: solution}

[0090] Among them, e represents an entity, e: fault code means the category of entity e is "fault code", and {e | e: fault code} represents the set of all entities of the fault code category. e1 and e2 represent two entities, and <e1, e2> represents the entity relationship between entity e1 and entity e2. In the fault manual knowledge graph, the information queried using the fault code is saved, that is, the fault symptom entity, the fault cause entity, and the fault solution entity can be queried through the fault code entity; at the same time, the information of querying the fault cause entity through the fault symptom entity and querying the solution entity through the fault cause entity is also saved. The topological structure of the fault manual knowledge graph is as Figure 3 shown.

[0091] In this optional embodiment, extract second-category entities from the vehicle fault historical maintenance records, namely fault description entities, fault diagnosis entities, and maintenance item entities, establish second-entity relationships between the second-category entities, that is, query the fault diagnosis entity according to the fault description entity, query the maintenance item entity according to the fault diagnosis entity, and construct the fault maintenance record knowledge graph according to the second-category entities and the second-entity relationships. The fault maintenance record knowledge graph summarizes a large amount of maintenance experience and improves the situation where the maintenance experience of maintenance workers is not comprehensive.

[0092] Specifically, three types of second-category entities, namely "fault description", "fault diagnosis", and "repair item", are extracted from historical maintenance records to obtain the second-category entity set E2. Based on these three types of second-category entities, two types of second-entity relationships, namely "fault description - fault diagnosis" and "fault diagnosis - repair item", are constructed to obtain the second-entity relationship set R2. A fault maintenance record knowledge graph G2(E2, R2) is constructed based on these three types of second-category entities and these two types of second-entity relationships. The expressions of the second-category entity set E2 and the second-entity relationship set R2 are as follows:

[0093] E2 = {e|e: fault description} ∪ {e|e: fault diagnosis} ∪ {e|e: repair item}

[0094] R2 = {<e1,e2〉|e1: fault description, e2: fault diagnosis} ∪ {<e1,e2〉|e1: fault diagnosis, e2: repair item}

[0095] Among them, e represents an entity, e: fault description means the category of entity e is "fault description", and {e|e: fault description} represents the set of all entities of the fault description category. e1 and e2 represent two entities, and <e1, e2> represents the entity relationship between entity e1 and entity e2. In the fault maintenance record knowledge graph, information on querying the fault diagnosis entity based on the fault description entity and querying the repair item entity based on the fault diagnosis entity is saved. The topological structure of the fault maintenance record knowledge graph is as Figure 4 shown.

[0096] In an optional embodiment of the present application, the process of fusing the fault manual knowledge graph and the fault maintenance record knowledge graph to obtain the fault knowledge graph includes merging the first entity relationships in the fault manual knowledge graph with the corresponding second entity relationships in the fault maintenance record knowledge graph respectively; aligning the fault phenomenon entity, fault cause entity, and solution entity in the fault manual knowledge graph with the fault description entity, fault diagnosis entity, and repair item entity in the fault maintenance record knowledge graph one by one, thereby obtaining the fault knowledge graph. The fault knowledge graph comprehensively sorts out the fault manual and historical maintenance experience, summarizes the experience knowledge of a large number of maintenance workers, improves the situation of incomplete personal maintenance experience, strengthens the correlation between data, and increases the fault coverage rate.

[0097] In this optional embodiment, the "fault phenomenon" in the fault manual knowledge graph is aligned with the "fault description" in the fault repair record knowledge graph; the "fault cause" in the fault manual knowledge graph is aligned with the "fault diagnosis" in the fault repair record knowledge graph; the "solution" in the fault manual knowledge graph is aligned with the "repair item" in the fault repair record knowledge graph; the "fault phenomenon-fault cause" entity relationship in the fault manual knowledge graph is merged with the "fault description-fault diagnosis" entity relationship in the fault repair record knowledge graph; and the "fault cause-solution" entity relationship in the fault manual knowledge graph is merged with the "fault diagnosis-repair item" entity relationship in the fault repair record knowledge graph, thereby obtaining the fused fault knowledge graph.

[0098] Specifically, the fault manual knowledge graph G1 and the fault repair record knowledge graph G2 are merged to generate a fault knowledge graph G(En, Re). The expression for the merged set of category entities En is as follows:

[0099] En = E c ∪E s ∪E r ∪E m

[0100] Where En is the set of category entities obtained by aligning the first category entities in the fault manual knowledge graph with the second category entities in the fault repair record knowledge graph, and E c E s E r and E m The expressions are as follows:

[0101] E C ={e|e: fault code}

[0102] E S ={e|e: fault phenomenon}∪{e|e: fault description}

[0103] E r ={e|e: cause of failure}∪{e|e: fault diagnosis}

[0104] E m ={e|e: Solution}∪{e|e: Repair Items}

[0105] Where e represents an entity, e: fault code indicates that the category of entity e is "fault code", and {e|e: fault code} represents the set of entities of all fault code categories. E c E is a set of fault code entities. s E is the set of entities resulting from the fusion of fault phenomenon entities and fault description entities. rE is the set of entities formed by fusing the entities responsible for the fault cause and the entities responsible for the fault diagnosis. m This is the set of entities resulting from the merging of solution entities and maintenance project entities.

[0106] In the fault knowledge graph, the expression for the fused entity relation set Re is as follows:

[0107] Re = R cs ∪R cr ∪R cm ∪R sr ∪R rm

[0108] Where Re is the set of entity relations obtained by fusing the first entity relation in the fault manual knowledge graph and the second entity relation in the fault repair record knowledge graph, and R cs R cr R cm R sr and R rm The expressions are as follows:

[0109] R cs ={<e1,e2> |e1: Fault code, e2: Fault symptom}

[0110] R cr ={<e1,e2>|e1: fault code,e2: fault cause}

[0111] R cm ={<e1,e2> |e1: Fault code, e2: Solution}

[0112] R sr ={<e1,e2> |e1: Fault symptom, e2: Fault cause}∪{<e1,e2> |e1: Fault description, e2: Fault diagnosis}

[0113] R rm ={<e1,e2> |e1: Cause of the fault, e2: Solution}∪{<e1,e2> |e1: Fault Diagnosis, e2: Repair Items}

[0114] Where e1 and e2 represent two entities,<e1,e2> This represents the entity relationship between entity e1 and entity e2. (R) cs R is a set of entity relations for "fault code - fault phenomenon". cr R is a set of entity relations for "fault code - fault cause". cm For the set of entity relationships of "fault code-solution", R sr R is the set of entity relations after the fusion of "fault phenomenon-fault cause" and "fault description-fault diagnosis". rmThis is the entity relationship set after merging "Fault Cause-Solution" and "Fault Diagnosis-Repair Item". Entity alignment and entity relationship fusion are performed on the fault phenomenon entities, fault cause entities, and solution entities in the fault manual knowledge graph and the fault description entities, fault diagnosis entities, and repair item entities in the fault repair record knowledge graph. The topology of the merged fault knowledge graph is as follows: Figure 5 As shown.

[0115] S203, Analyze the received fault information. If the fault information includes fault codes, determine the first set of fault causes and the first set of solutions based on the fault codes in the pre-built fault knowledge graph.

[0116] S204. If the fault information does not include a fault code, then determine the weight of the fault phenomenon entity corresponding to the first fault phenomenon based on the attribution relationship between the first fault phenomenon in the fault information and the fault knowledge graph.

[0117] S205, in the fault knowledge graph, determine the second fault cause corresponding to the first fault phenomenon, and determine the attribution weight of the second fault cause based on the number of second fault phenomena corresponding to the second fault cause.

[0118] S206. Determine the relevance weight of the second fault cause based on the entity weight of the fault phenomenon and the attribution weight.

[0119] S207. Determine the second set of causes of failure based on the relevance weight, and determine the corresponding second set of solutions based on the second set of causes of failure.

[0120] In this embodiment, please refer to the detailed implementation process and technical principles of steps S203 to S207. Figure 1 The relevant descriptions of steps S101 to S105 in the method shown will not be repeated here.

[0121] In this embodiment, firstly, a fault manual knowledge graph and a fault repair record knowledge graph are constructed based on the fault manual and vehicle fault history repair records, respectively. These two knowledge graphs are then merged into a single fault knowledge graph. This fault knowledge graph comprehensively organizes the fault manual and vehicle fault history repair records, summarizing the extensive experience and knowledge of repair workers, thus strengthening the correlation between data and achieving more comprehensive fault coverage. Then, if a specific fault code is input, the fault code is queried based on the topology of the fault knowledge graph, returning the related fault symptoms, causes, and solutions. If no fault code exists, a specific fault symptom is input. Based on the topology of the fault knowledge graph and the degree of attribution of the fault symptom to the cause, the fault cause with a high degree of relevance to the symptom is inferred. The relevant solutions and causes are then retrieved and returned, reducing the error rate and improving the situation where individual repair experience is incomplete.

[0122] Example 3

[0123] Figure 6 This is a schematic diagram of the structure of a vehicle intelligent auxiliary diagnostic server provided in Embodiment 3 of this application, as shown below. Figure 6 As shown, the server in this embodiment may include:

[0124] Module 601, Fault Information Analysis Module, is used to analyze the received fault information. If the fault information includes fault codes, the first set of fault causes and the first set of solutions are determined based on the fault codes in the pre-built fault knowledge graph.

[0125] Module 602, Entity Weight Determination Module, is used to determine the entity weight of the fault phenomenon corresponding to the first fault phenomenon based on the attribution relationship between the first fault phenomenon in the fault information and the fault knowledge graph if the fault information does not include a fault code.

[0126] Module 603, the attribution weight determination module, is used to determine the second fault cause corresponding to the first fault phenomenon in the fault knowledge graph, and to determine the attribution weight of the second fault cause based on the number of second fault phenomena corresponding to the second fault cause.

[0127] Module 604, the relevance weight determination module, is used to determine the relevance weight of the second fault cause based on the entity weight of the fault phenomenon and the attribution weight.

[0128] Module 605, Solution Determination Module, is used to determine a second set of fault causes based on relevance weights, and to determine a corresponding second set of solutions based on the second set of fault causes.

[0129] In this embodiment, the vehicle intelligent auxiliary diagnostic server provides vehicle fault diagnosis services to the outside world via a REST interface. It analyzes and diagnoses received fault information and returns results. This vehicle intelligent auxiliary diagnostic server can query based on fault codes, and can also query fault symptoms simultaneously when fault codes are not found. This vehicle intelligent auxiliary diagnostic server avoids the problem of data model dependency, and also avoids the problem of data imbalance in the data model causing the model's performance to be biased towards common faults. Modules 602 to 605 in the vehicle intelligent auxiliary diagnostic server ultimately calculate the relevance of each fault symptom. When data is updated, there is no need to retrain the model, enabling the vehicle intelligent auxiliary diagnostic server to achieve the effect of querying and updating simultaneously.

[0130] Example 4

[0131] Figure 7 This is a schematic diagram of the structure of a vehicle intelligent auxiliary diagnostic system provided in Embodiment 3 of this application, as shown below. Figure 7 As shown, the system in this embodiment may include:

[0132] Module 701, the client, is used to input fault information, which includes the first fault symptom and / or fault code;

[0133] Module 702, the server side, is used to execute the vehicle intelligent auxiliary diagnosis method in Scheme 1, and return the set of fault causes, the set of solutions, and, or the set of fault phenomena obtained by analyzing fault information based on the fault knowledge graph to the client.

[0134] In a specific instance of this application, if the received fault information is a first fault symptom, the diagnostic result E' inferred from the above instance is returned to the client. The repair worker then repairs the vehicle fault based on the returned set of first fault causes and the corresponding set of first solutions. If the received fault information is a fault code, the set of fault symptoms S, the set of first fault causes R, and the set of first solutions M directly queried from the above instance are returned to the client. This allows the repair worker to gain more repair experience and improve fault diagnosis efficiency.

[0135] The vehicle intelligent auxiliary diagnostic system provided in this application can be used to execute the vehicle intelligent auxiliary diagnostic method described in any of the above embodiments. Its implementation principle and technical effect are similar, and will not be repeated here.

[0136] In this embodiment, the functional modules of the vehicle intelligent auxiliary diagnostic system of this application can be directly in hardware, in software modules executed by a processor, or in a combination of both.

[0137] Software modules may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disks, removable disks, CD-ROMs, or any other form of storage medium known in this art. An exemplary storage medium is coupled to the processor, enabling the processor to read information from and write information to the storage medium.

[0138] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or any combination thereof. A general-purpose processor can be a microprocessor, but alternatively, it can be any conventional processor, controller, microcontroller, or state machine. The processor can also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors incorporating a DSP core, or any other such configuration. Alternatively, the storage medium can be integrated with the processor. The processor and storage medium can reside in an ASIC. The ASIC can reside in the user terminal. Alternatively, the processor and storage medium can reside as discrete components in the user terminal.

[0139] Furthermore, embodiments of this application also provide a computer-readable storage medium storing computer instructions that are operated to perform the vehicle intelligent assisted diagnostic method in any embodiment.

[0140] Computer-readable media include computer storage media and communication media, wherein communication media include any medium that facilitates the transfer of computer programs from one place to another. Storage media can be any available medium accessible to a general-purpose or special-purpose computer. 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. Alternatively, the ASIC can reside in a user device. Of course, the processor and storage medium can also exist as discrete components in a communication device.

[0141] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units 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 apparatuses or units may be electrical, mechanical, or other forms.

[0142] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0143] The above description is merely an embodiment of this application and does not limit the patent scope of this application. Any equivalent structural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A vehicle intelligent assisted diagnostic method, characterized in that, include: The received fault information is analyzed. If the fault information includes a fault code, then a first set of fault causes and a first set of solutions are determined based on the fault code in a pre-constructed fault knowledge graph. If the fault information does not include the fault code, then the fault phenomenon entity weight corresponding to the first fault phenomenon is determined according to the attribution relationship between the first fault phenomenon in the fault information and the fault knowledge graph. In the fault knowledge graph, the second fault cause corresponding to the first fault phenomenon is determined, and the attribution weight of the second fault cause is determined according to the number of second fault phenomena corresponding to the second fault cause. The relevance weight of the second fault cause is determined based on the entity weight of the fault phenomenon and the attribution weight. Based on the relevance weights, a second set of fault causes is determined, and a corresponding second set of solutions is determined based on the second set of fault causes. Before analyzing the received fault information, the fault knowledge graph is constructed, which includes: constructing corresponding fault manual knowledge graphs and fault repair record knowledge graphs based on the fault manual and vehicle fault history repair records, respectively. The fault knowledge graph is obtained by fusing the fault manual knowledge graph and the fault repair record knowledge graph.

2. The vehicle intelligent assisted diagnostic method as described in claim 1, characterized in that, The step of determining the attribution weight of the second fault cause based on the number of second fault phenomena corresponding to the second fault cause includes: Based on the number of the second fault phenomena in the fault knowledge graph, the confidence level of the second fault cause for the first fault phenomenon is determined, wherein the more the number of the second fault phenomena, the lower the confidence level. Based on the confidence level, the attribution weight of the second fault cause to the first fault phenomenon is determined, wherein the higher the confidence level, the higher the attribution weight.

3. The vehicle intelligent assisted diagnostic method as described in claim 1, characterized in that, The step of determining the entity weight of the fault phenomenon corresponding to the first fault phenomenon based on the attribution relationship between the first fault phenomenon in the fault information and the fault knowledge graph includes: In the first set of fault phenomena in the fault information, the number of the first fault phenomena belonging to the fault knowledge graph is determined, wherein the first set of fault phenomena includes the first fault phenomena. Based on the quantity, the entity weight of the fault phenomenon corresponding to the first fault phenomenon is set, wherein the larger the quantity, the smaller the entity weight of the fault phenomenon.

4. The vehicle intelligent assisted diagnostic method as described in claim 1, characterized in that, The step of determining a second set of fault causes based on the relevance weights, and determining a corresponding second set of solutions based on the second set of fault causes, includes: The relevance weights are sorted by size, and a certain number of the second fault causes are selected based on the sorting results to obtain the second fault cause set; In the fault knowledge graph, the corresponding set of second solutions is determined based on the second set of fault causes.

5. The intelligent auxiliary diagnosis method for vehicle faults according to claim 1, characterized in that, The process involves constructing corresponding fault manual knowledge graphs and fault repair record knowledge graphs based on the fault manual and vehicle fault history repair records, respectively, including: Extract the first category entities from the fault manual and analyze the first entity relationships between the first category entities, wherein the first category entities include fault code entities, fault phenomenon entities, fault cause entities, and solution entities; Based on the extracted first category entities and the analyzed first entity relationships, a fault manual knowledge graph is constructed. Extract the second category entities from the vehicle fault history repair records, and analyze the second entity relationships between the second category entities. The second category entities include fault description entities, fault diagnosis entities, and repair item entities. Based on the extracted second category entities and the analyzed second entity relationships, a fault repair record knowledge graph is constructed.

6. The intelligent auxiliary diagnosis method for vehicle faults according to claim 5, characterized in that, The process of fusing the fault manual knowledge graph and the fault repair record knowledge graph to obtain the fault knowledge graph includes: Merge the first entity relationship in the fault manual knowledge graph with the corresponding second entity relationship in the fault repair record knowledge graph; The fault phenomenon entity, fault cause entity, and solution entity in the fault manual knowledge graph are aligned one-to-one with the fault description entity, fault diagnosis entity, and maintenance item entity in the fault repair record knowledge graph to obtain the fault knowledge graph.

7. A vehicle intelligent auxiliary diagnostic server, characterized in that, include: The fault information analysis module is used to analyze the received fault information. If the fault information includes a fault code, then in the pre-constructed fault knowledge graph, a first set of fault causes and a first set of solutions are determined according to the fault code. An entity weight determination module is used to determine the entity weight of the fault phenomenon corresponding to the first fault phenomenon based on the attribution relationship between the first fault phenomenon in the fault information and the fault knowledge graph if the fault information does not include the fault code. The attribution weight determination module is used to determine the second fault cause corresponding to the first fault phenomenon in the fault knowledge graph, and to determine the attribution weight of the second fault cause based on the number of second fault phenomena corresponding to the second fault cause. The relevance weight determination module is used to determine the relevance weight of the second fault cause based on the entity weight of the fault phenomenon and the attribution weight. as well as The solution determination module is used to determine a second set of fault causes based on the relevance weights, and to determine a corresponding second set of solutions based on the second set of fault causes. Before analyzing the received fault information, the fault knowledge graph is constructed, which includes: constructing corresponding fault manual knowledge graphs and fault repair record knowledge graphs based on the fault manual and vehicle fault history repair records, respectively. The fault knowledge graph is obtained by fusing the fault manual knowledge graph and the fault repair record knowledge graph.

8. A vehicle intelligent auxiliary diagnostic system, characterized in that, include: A client for inputting fault information, which includes a first fault symptom and / or a fault code; The server side is used to execute the vehicle intelligent auxiliary diagnosis method according to any one of claims 1-6, and return the set of fault causes, the set of solutions and / or the set of fault phenomena obtained by analyzing the fault information according to the fault knowledge graph to the client.

9. A computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are operated to perform the vehicle intelligent assisted diagnostic method according to any one of claims 1-6.

Images