Fault analysis method and device, computer equipment and storage medium

Abstract

The invention relates to a fault analysis method and device, computer equipment and a storage medium, and relates to the technical field of vehicle fault analysis, and the method comprises the steps: obtaining fault data of at least one to-be-analyzed vehicle; constructing a fault map network according to the fault data of each to-be-analyzed vehicle; the fault map network is composed of nodes and connecting edges, the nodes represent vehicle parts, and the connecting edges represent potential propagation paths of faults propagating from one component to another component; determining a feature embedding matrix of each node in the fault map network; wherein the feature embedding matrix is determined through a nonlinear propagation formula; and based on a vehicle fault analysis model, performing forward updating and reverse updating on the feature embedding matrix, and determining a fault type corresponding to the to-be-analyzed vehicle. The potential propagation path of the fault can be effectively captured, so that the fault type and the fault part which cause the vehicle fault are accurately positioned, and the effect of improving the positioning accuracy of the fault type is achieved.

Description

Technical Field

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

[0002] With the continuous development of automobile technology, the structure of automobiles is becoming more and more complex, the functions are increasing, and the correlation between various systems and components is becoming stronger. Therefore, the analysis of vehicle failures needs to be more professional and refined.

[0003] The traditional method is to troubleshoot problems by reading the vehicle fault description information. However, it is unable to quickly and accurately troubleshoot the fault type of the vehicle that has hardware faults, system faults, and on-board software problems (automatic start-stop function abnormalities, remote control function failure, unstable signals, etc.), and there is a problem of inaccurate fault type positioning. Summary of the invention

[0004] Based on this, it is necessary to provide a fault analysis method, device, computer equipment and storage medium that can improve the accuracy of fault type location in response to the above technical problems.

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

[0006] Acquiring fault data of at least one vehicle to be analyzed;

[0007] According to the fault data of each vehicle to be analyzed, a fault graph network is constructed; the fault graph network is composed of nodes and connecting edges, where nodes represent vehicle components and connecting edges represent potential propagation paths of faults from one component to another;

[0008] Determine the feature embedding matrix of each node in the fault spectrum network; wherein the feature embedding matrix is ​​determined by the nonlinear propagation formula; the nonlinear propagation formula is: ,in, , node j represents the adjacent node with a connecting edge to node i, , N(i) represents the set of adjacent nodes of node i, Indicates The node feature vector of layer node i, represents the inverse function of the polynomial kernel function, represents the node feature vector of node j in layer l, represents the node feature vector of node i in layer l, represents the attention coefficient between node i and node j in layer l, x i represents the initial eigenvector of node i, exp(.) represents the exponential function with e as the base, represents the function variance, ||.|| represents the modulo operation;

[0009] Based on the vehicle fault analysis model, the feature embedding matrix is ​​updated forwardly and backward to determine the fault type corresponding to the vehicle to be analyzed.

[0010] In one embodiment, based on a vehicle fault analysis model, a feature embedding matrix is ​​forward updated and reverse updated to determine a fault type corresponding to the vehicle to be analyzed, including: based on the vehicle fault analysis model, dimensionality reduction processing is performed on the embedded feature matrix after the forward update to obtain a first processing result; based on the vehicle fault analysis model, dimensionality reduction processing is performed on the embedded feature matrix after the reverse update to obtain a second processing result; based on the first processing result and the second processing result, the fault type corresponding to the vehicle to be analyzed is determined.

[0011] In one of the embodiments, determining the fault type corresponding to the vehicle to be analyzed based on the first processing result and the second processing result includes: fusing the first processing result and the second processing result to obtain a target processing result; and determining the fault type corresponding to the vehicle to be analyzed based on the target processing result.

[0012] In one of the embodiments, the fault type corresponding to the vehicle to be analyzed is determined based on the target processing result, including: transforming the target processing result through a normalized exponential function to determine the probability value of each node in the fault map network belonging to different node types; and taking the fault type corresponding to the node type with the largest probability value as the fault type corresponding to the vehicle to be analyzed.

[0013] In one embodiment, determining a feature embedding matrix of each node in a fault graph network includes: determining a target node attribute of each node in the fault graph network; constructing an adjacency matrix of the fault graph network according to a connection relationship between each node in the fault graph network; and determining a feature embedding matrix of each node in the fault graph network according to the adjacency matrix and the target node attribute based on a nonlinear propagation formula.

[0014] In one of the embodiments, the target node attributes of each node in the fault map network are determined, including: obtaining the original node attributes of each node in the fault map network; analyzing the original node attributes through an attention mechanism to determine the impact value of each original node attribute on the node type; and selecting a preset number of original node attributes as target node attributes according to the size of the impact value.

[0015] In a second aspect, the present application also provides a fault analysis device. The device comprises:

[0016] A data acquisition module, used to acquire fault data of at least one vehicle to be analyzed;

[0017] A network construction module is used to construct a fault map network based on the fault data of each vehicle to be analyzed; the fault map network is composed of nodes and connecting edges, where nodes represent vehicle components and connecting edges represent potential propagation paths of faults from one component to another;

[0018] The feature matrix determination module is used to determine the feature embedding matrix of each node in the fault spectrum network; wherein the feature embedding matrix is ​​determined by a nonlinear propagation formula; the nonlinear propagation formula is: ,in, , node j represents the adjacent node with a connecting edge to node i, , N(i) represents the set of adjacent nodes of node i, Indicates The node feature vector of layer node i, represents the inverse function of the polynomial kernel function, represents the node feature vector of node j in layer l, represents the node feature vector of node i in layer l, represents the attention coefficient between node i and node j in layer l, x i represents the initial eigenvector of node i, exp(.) represents the exponential function with e as the base, represents the function variance, ||.|| represents the modulo operation;

[0019] The fault type determination module is used to perform forward and reverse updates on the feature embedding matrix based on the vehicle fault analysis model to determine the fault type corresponding to the vehicle to be analyzed.

[0020] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0021] Acquiring fault data of at least one vehicle to be analyzed;

[0022] According to the fault data of each vehicle to be analyzed, a fault graph network is constructed; the fault graph network is composed of nodes and connecting edges, where nodes represent vehicle components and connecting edges represent potential propagation paths of faults from one component to another;

[0023] Determine the feature embedding matrix of each node in the fault spectrum network; wherein the feature embedding matrix is ​​determined by the nonlinear propagation formula; the nonlinear propagation formula is: ,in, , node j represents the adjacent node with a connecting edge to node i, , N(i) represents the set of adjacent nodes of node i; Indicates The node feature vector of layer node i, represents the inverse function of the polynomial kernel function, represents the node feature vector of node j in layer l, represents the node feature vector of node i in layer l, represents the attention coefficient between node i and node j in layer l, x i represents the initial eigenvector of node i, exp(.) represents the exponential function with e as the base, represents the function variance, ||.|| represents the modulo operation;

[0024] Based on the vehicle fault analysis model, the feature embedding matrix is ​​updated forwardly and backward to determine the fault type corresponding to the vehicle to be analyzed.

[0025] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0026] Acquiring fault data of at least one vehicle to be analyzed;

[0027] According to the fault data of each vehicle to be analyzed, a fault graph network is constructed; the fault graph network is composed of nodes and connecting edges, where nodes represent vehicle components and connecting edges represent potential propagation paths of faults from one component to another;

[0028] Determine the feature embedding matrix of each node in the fault spectrum network; wherein the feature embedding matrix is ​​determined by the nonlinear propagation formula; the nonlinear propagation formula is: ,in, , node j represents the adjacent node with a connecting edge to node i, , N(i) represents the set of adjacent nodes of node i; Indicates The node feature vector of layer node i, represents the inverse function of the polynomial kernel function, represents the node feature vector of node j in layer l, represents the node feature vector of node i in layer l, represents the attention coefficient between node i and node j in layer l, x i represents the initial eigenvector of node i, exp(.) represents the exponential function with e as the base, represents the function variance, ||.|| represents the modulo operation;

[0029] Based on the vehicle fault analysis model, the feature embedding matrix is ​​updated forwardly and backward to determine the fault type corresponding to the vehicle to be analyzed.

[0030] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0031] Acquiring fault data of at least one vehicle to be analyzed;

[0032] According to the fault data of each vehicle to be analyzed, a fault graph network is constructed; the fault graph network is composed of nodes and connecting edges, where nodes represent vehicle components and connecting edges represent potential propagation paths of faults from one component to another;

[0033] Determine the feature embedding matrix of each node in the fault spectrum network; wherein the feature embedding matrix is ​​determined by the nonlinear propagation formula; the nonlinear propagation formula is: ,in, , node j represents the adjacent node with a connecting edge to node i, , N(i) represents the set of adjacent nodes of node i; Indicates The node feature vector of layer node i, represents the inverse function of the polynomial kernel function, represents the node feature vector of node j in layer l, represents the node feature vector of node i in layer l, represents the attention coefficient between node i and node j in layer l, x i represents the initial eigenvector of node i, exp(.) represents the exponential function with e as the base, represents the function variance, ||.|| represents the modulo operation;

[0034] Based on the vehicle fault analysis model, the feature embedding matrix is ​​updated forwardly and backward to determine the fault type corresponding to the vehicle to be analyzed.

[0035] The above-mentioned fault analysis method, device, computer equipment, storage medium and product obtain the fault data of at least one vehicle to be analyzed; construct a fault map network according to the fault data of each vehicle to be analyzed; the fault map network is composed of nodes and connecting edges, wherein nodes represent vehicle components and connecting edges represent potential propagation paths of faults from one component to another; determine the feature embedding matrix of each node in the fault map network; wherein the feature embedding matrix is ​​determined by a nonlinear propagation formula; based on the vehicle fault analysis model, the feature embedding matrix is ​​forwardly updated and reversely updated to determine the fault type corresponding to the vehicle to be analyzed. In the above process, each node in the fault map network of the present application obtains a feature embedding matrix through nonlinear propagation on the connecting edges, which effectively captures the high-order nonlinear relationship between different nodes, and ensures that the obtained feature embedding matrix can accurately reflect the own properties of each node in the fault map network and the associated properties between nodes. Furthermore, based on the vehicle fault analysis model, the feature embedding matrix is ​​forward updated so that the updated feature embedding matrix is ​​more consistent with the actual fault conditions of the vehicle to be analyzed; the feature embedding matrix is ​​reversely updated to continuously optimize the parameters of the vehicle fault analysis model, thereby improving the accuracy of the fault type corresponding to the vehicle to be analyzed and the efficiency of fault type location. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A flowchart of a fault analysis method provided in an embodiment of the present application;

[0037] Figure 2 A flowchart of the steps for determining the fault type provided in an embodiment of the present application;

[0038] Figure 3 A flowchart of the steps of determining a feature embedding matrix provided in an embodiment of the present application;

[0039] Figure 4 A flowchart of another fault analysis method provided in an embodiment of the present application;

[0040] Figure 5 A structural block diagram of a fault analysis device provided in an embodiment of the present application;

[0041] Figure 6 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0043] It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. In the description of the present application, the description of the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representation of the above terms is not necessarily for the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples without contradicting each other.

[0044] In one embodiment, Figure 1 A flowchart of a fault analysis method provided in an embodiment of the present application is provided. This embodiment uses the method applied to a vehicle-mounted terminal as an example. It is understandable that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. Among them, the terminal can be, but is not limited to, various personal computers, laptops, smart phones, and tablet computers. The server can be implemented as an independent server or a server cluster consisting of multiple servers. In this embodiment, the method includes the following steps:

[0045] In an exemplary embodiment, Figure 1 The fault analysis method shown includes:

[0046] S101, acquiring fault data of at least one vehicle to be analyzed.

[0047] Among them, fault data includes vehicle parameter data and fault parameter data. Vehicle parameter data is used to describe a series of information such as automobile performance, design, functional characteristics, and the vehicle's own configuration to be analyzed. Fault parameter data is used to describe the body status information of the vehicle to be analyzed when a fault occurs in the vehicle to be analyzed.

[0048] It should be noted that the vehicle parameter data may include at least one of the following: vehicle body size, wheelbase, minimum ground clearance, curb weight, engine type, displacement, maximum horsepower, maximum power, maximum torque, fuel capacity, tire specifications, engine speed, vehicle speed, fuel consumption rate, emission level, sensor data, battery voltage, oil pressure, and coolant temperature. The fault parameter data may include at least one of the following: fault code, actuator status (such as the opening and closing status and working conditions of fuel injectors, ignition coils, throttle positions, etc.), driving behavior data (such as the frequency and intensity of acceleration, braking, steering, and driving mode), time and location of the fault, the vehicle's operating status at the time, maintenance history, and logs before and after the fault occurs.

[0049] As an implementation method, the electronic system of the vehicle can be monitored in real time through the OBD system installed in the vehicle to be analyzed, and fault data can be obtained through physical requests.

[0050] S102, constructing a fault graph network according to the fault data of each vehicle to be analyzed.

[0051] Among them, the fault graph network is composed of nodes and connecting edges. The nodes represent vehicle components, and the connecting edges represent the potential propagation paths of faults from one component to another.

[0052] It should be noted that in the fault graph network, nodes can represent various entities or components in the network, which can be physical or logical. For example, nodes can be physical components of the vehicle such as the engine, transmission, sensor, actuator, etc. Nodes can also represent different systems of the vehicle, such as the power system, braking system, steering system, electrical system, etc. In the fault graph network, connecting edges are used to represent the relationship between nodes, which can be physical, logical, causal or functional. The existence and properties of connecting edges can represent important information about how vehicle systems interact with each other and how faults propagate in the system. For example, connecting edges can represent potential propagation paths for faults to propagate from one component to another; connecting edges can also represent direct connections between physical components, such as wires, pipes, mechanical connections, etc.; connecting edges can also represent the causal relationship that the failure of one component leads to the failure of another component; connecting edges can also indicate that the normal operation of one component depends on the state of another component, such as the normal operation of the engine depends on the supply of the fuel pump.

[0053] In one embodiment of the present application, the construction process of constructing a fault graph network may specifically include the following contents: for any vehicle to be analyzed, entity extraction is performed on the fault data of the vehicle to be analyzed through a word embedding algorithm to obtain at least one fault graph entity, that is, a node in the fault graph network; according to the association relationship between the fault graph entity and the fault graph entity, the connection edges between the nodes in the fault graph network are determined; and each node is connected through the connection edges to construct a fault graph network.

[0054] S103, determining the feature embedding matrix of each node in the fault graph network.

[0055] Among them, the feature embedding matrix is ​​determined by the nonlinear propagation formula, which is:

[0056] (1)

[0057] Among them, node j represents the adjacent node with a connecting edge to node i. , N(i) represents the set of adjacent nodes of node i; Indicates The node feature vector of layer node i, represents the inverse function of the polynomial kernel function, represents the node feature vector of node j in layer l, represents the node feature vector of node i in layer l, represents the attention coefficient between node i and node j in layer l, x i represents the initial feature vector of node i. Further explanation: in formula (1) The expression of is shown in formula (2):

[0058] (2)

[0059] in, represents the node feature vector of node j in layer l, represents the node feature vector of node i in the lth layer, node j represents the adjacent node with a connecting edge to node i, exp(.) represents the exponential function with e as the base, represents the function variance, and ||.|| represents the modulo operation.

[0060] In one embodiment of the present application, the node attributes of each node can be linearly transformed through the principal component analysis method to obtain the initial feature vector of each node in the fault graph network; further, a multi-layer attention network is used to perform attention analysis on the initial feature vector of each node to obtain the attention weight; further, the node attributes of all neighboring nodes of each node are propagated along the adjacency matrix through a nonlinear propagation formula to obtain the node feature vector of each node after propagation; the node feature vectors of each node in the fault graph network are vertically spliced ​​to obtain the feature embedding matrix of each node in the fault graph network.

[0061] S104, based on the vehicle fault analysis model, the feature embedding matrix is ​​forward updated and reverse updated to determine the fault type corresponding to the vehicle to be analyzed.

[0062] Among them, the fault types may include engine failure, electrical system failure, fuel system failure, exhaust system failure, transmission system failure, steering system failure, braking system failure, suspension system failure, tire and wheel failure, air conditioning system failure, on-board electronic equipment failure, body and interior failure, etc.

[0063] It should be noted that engine failures may include but are not limited to abnormal noise, overheating, difficulty in starting, abnormally increased fuel consumption, lack of power, etc.; electrical system failures may include battery self-discharge, charging system failure to charge, starter motor powerlessness or no movement, etc.; fuel system failures may include poor fuel supply, fuel pump failure, fuel injector blockage, etc.; exhaust system failures may include muffler failure, three-way catalytic converter blockage, exhaust pipe leakage, etc.; transmission system failures may include clutch slippage, transmission abnormal noise, drive shaft abnormal noise or vibration, etc.; steering system failures may include heavy steering, imprecise steering, steering wheel sway or steering wheel sway vibration, etc.; brake system failures may include brake failure, poor braking effect, brake drag, brake skidding or deviation, poor parking brake, etc.; suspension system failures may include shock absorber oil leakage, suspension spring breakage, abnormal noise in the suspension system, etc.; tire and wheel failures may include uneven tire wear, insufficient tire pressure, wheel hub bearing damage, etc.; body and interior failures may include body rust, interior damage, electric window failure, etc.; air-conditioning system failures may include air-conditioning not cooling, air-conditioning system leakage, air-conditioning fan not working, etc.; on-board electronic equipment failures may include navigation system failure, audio system failure, reversing radar not working, etc.

[0064] In one embodiment of the present application, the process of determining the fault type corresponding to the vehicle to be analyzed may specifically include the following contents: based on the vehicle fault analysis model, performing dimensionality reduction processing on the embedded feature matrix after forward update to obtain a first processing result; based on the vehicle fault analysis model, performing dimensionality reduction processing on the embedded feature matrix after reverse update to obtain a second processing result; based on the first processing result and the second processing result, determining the fault type corresponding to the vehicle to be analyzed.

[0065] Among them, the vehicle fault analysis model may include but is not limited to: graph attention network model, support vector machine model, Bayesian model, etc.

[0066] The above fault analysis method obtains the fault data of at least one vehicle to be analyzed; constructs a fault map network according to the fault data of each vehicle to be analyzed; the fault map network is composed of nodes and connecting edges, wherein the nodes represent vehicle components and the connecting edges represent the potential propagation path of the fault from one component to another; determines the feature embedding matrix of each node in the fault map network; wherein the feature embedding matrix is ​​determined by a nonlinear propagation formula; based on the vehicle fault analysis model, the feature embedding matrix is ​​forward updated and reverse updated to determine the fault type corresponding to the vehicle to be analyzed. In the above process, each node in the fault map network of the present application obtains a feature embedding matrix through nonlinear propagation on the connecting edge, effectively captures the high-order nonlinear relationship between different nodes, and ensures that the obtained feature embedding matrix can accurately reflect the own attributes of each node in the fault map network and the associated attributes between nodes. Furthermore, based on the vehicle fault analysis model, the feature embedding matrix is ​​forward updated so that the updated feature embedding matrix is ​​more consistent with the actual fault situation of the vehicle to be analyzed; the feature embedding matrix is ​​reversely updated to continuously optimize the parameters of the vehicle fault analysis model, thereby improving the accuracy of the fault type corresponding to the vehicle to be analyzed and the efficiency of fault type location.

[0067] Based on the technical solutions of the above embodiments, the present application also provides an optional embodiment. In this optional embodiment, the process of determining the fault type is optimized and improved, thereby improving the accuracy of the fault type corresponding to the vehicle to be analyzed and the efficiency of fault type location.

[0068] like Figure 2 The process of determining the fault type corresponding to the vehicle to be analyzed shown includes the following steps:

[0069] S201, based on the vehicle fault analysis model, performing dimensionality reduction processing on the embedded feature matrix after forward update to obtain a first processing result.

[0070] S202, based on the vehicle fault analysis model, performing dimensionality reduction processing on the embedded feature matrix after reverse update to obtain a second processing result.

[0071] Among them, the first processing result is used to represent the characteristic matrix composed of the characteristic vectors of each node in the fault graph network in the low-dimensional space after forward updating; the second processing result is used to represent the characteristic matrix composed of the characteristic vectors of each node in the fault graph network in the low-dimensional space after reverse updating.

[0072] It should be noted that since the embedded feature matrix after forward updating can better conform to the actual fault condition of the vehicle to be analyzed, the accuracy of fault classification is ensured. Furthermore, the embedded feature matrix after forward updating is subjected to dimensionality reduction processing, which saves storage space and improves the efficiency of fault analysis. Furthermore, the feature embedding matrix is ​​reversely updated, the parameters of the vehicle fault analysis model are optimized, and the embedded feature matrix after reverse updating is subjected to dimensionality reduction processing, which ensures the accuracy of the second processing result, thereby ensuring the accuracy of the fault type corresponding to the vehicle to be analyzed.

[0073] S203: Determine the fault type corresponding to the vehicle to be analyzed according to the first processing result and the second processing result.

[0074] In one embodiment of the present application, the process of determining the fault type corresponding to the vehicle to be analyzed may specifically include the following contents: fusing the first processing result and the second processing result to obtain the target processing result; and determining the fault type corresponding to the vehicle to be analyzed based on the target processing result.

[0075] It should be noted that there are many ways of fusion. For example, the target processing result can be obtained by horizontally splicing and fusing the first processing result and the second processing result; or, the target processing result can be obtained by averaging and weighted fusing the first processing result and the second processing result.

[0076] To further illustrate, the process of determining the fault type corresponding to the vehicle to be analyzed may specifically include the following contents: transforming the target processing result through a normalized exponential function to determine the probability value of each node in the fault spectrum network belonging to different node types; taking the fault type corresponding to the node type with the largest probability value as the fault type corresponding to the vehicle to be analyzed.

[0077] The above fault analysis method uses the vehicle fault analysis model to forward update the feature embedding matrix, so that the updated feature embedding matrix is ​​more consistent with the actual fault conditions of the vehicle to be analyzed; then, the feature embedding matrix is ​​reversely updated to continuously optimize the parameters of the vehicle fault analysis model, thereby improving the accuracy of the fault type corresponding to the vehicle to be analyzed and the efficiency of fault type location.

[0078] Based on the technical solutions of the above embodiments, the present application also provides an optional embodiment, in which the step of determining the feature embedding matrix of each node in the fault graph network is optimized and improved to effectively capture the high-order nonlinear relationship between different nodes, ensuring that the obtained feature embedding matrix can accurately reflect the properties of each node in the fault graph network and the effect of the associated properties between nodes. Figure 3 The method shown comprises the following steps:

[0079] S301, determining the target node attribute of each node in the fault graph network.

[0080] The target node attribute is used to represent the attribute that has a greater impact on the node classification result, that is, the attribute that has a greater influence on determining the fault type of the vehicle to be analyzed.

[0081] In one embodiment of the present application, the original node attributes of each node in the fault map network can be obtained first; the original node attributes are analyzed through the attention mechanism to determine the impact value of each original node attribute on the node type; and a preset number of original node attributes are selected as target node attributes according to the size of the impact value.

[0082] For example, if there is a node A in the fault graph network, the original node attributes of node A include: attribute 1, attribute 2, attribute 3, attribute 4, and attribute 5. The impact value of attribute 1 on the node type is 0.5, the impact value of attribute 2 on the node type is 0.66, the impact value of attribute 3 on the node type is 0.6, the impact value of attribute 4 on the node type is 0.75, and the impact value of attribute 5 on the node type is 0.8. The preset number is 3. Since 0.8>0.75>0.66>0.6>0.5, attribute 5, attribute 4, and attribute 2 are used as the target node attributes of each node in the fault graph network.

[0083] S302: construct an adjacency matrix of the fault graph network according to the connection relationship between each node in the fault graph network.

[0084] The connection relationship may be a physical connection relationship, a transmission dependency relationship, a fault causal relationship, etc.

[0085] It should be noted that the number of nodes in the fault graph network is the order of the adjacency matrix. For example, if the fault graph network has 10 nodes, the adjacency matrix is ​​10 rows and 10 columns, that is, the order is 10. To further illustrate, if there is a connecting edge between two nodes, the value of the intersection position of the two nodes in the adjacency matrix is ​​1; if there is no connecting edge between the two nodes, the value of the intersection position of the two nodes in the adjacency matrix is ​​0. The values ​​of the diagonal positions of the adjacency matrix are all set to 1.

[0086] For example, if there are three nodes in the fault graph network, namely, node B, node C, and node D, where there is a connection relationship between node C and node D, and there is a connection relationship between node B and node D, then the adjacency matrix is ​​a symmetric matrix with 3 rows and 3 columns, because there is a connection relationship between node C and node D, and there is a connection relationship between node B and node D. Therefore, the value of the position where node C and node D intersect in the adjacency matrix is ​​1, and the value of the position where node B and node D intersect is 1.

[0087] S303, based on the nonlinear propagation formula, according to the adjacency matrix and the target node attributes, determine the feature embedding matrix of each node in the fault graph network.

[0088] It should be noted that the target node attributes are input into the nonlinear propagation formula so that the target node attributes can be propagated nonlinearly through the adjacency matrix, and the high-order nonlinear relationship between different nodes is obtained to obtain the target feature vector; further, the target feature vector is activated by the softmax activation function to determine the feature embedding matrix of each node in the fault graph network.

[0089] The above fault analysis method nonlinearly propagates the target node attributes of each node in the fault graph network on the adjacency matrix, effectively capturing the high-order nonlinear relationship between different nodes, ensuring that the obtained feature embedding matrix can accurately reflect the properties of each node in the fault graph network and the effect of the associated properties between nodes, thereby improving the accuracy and comprehensiveness of node representation.

[0090] Based on the technical solutions of the above embodiments, the present application also provides an optional embodiment, in which the fault analysis process is described in detail. Figure 4 The fault analysis method shown includes:

[0091] S401, acquiring fault data of at least one vehicle to be analyzed.

[0092] S402: construct a fault graph network based on the fault data of each vehicle to be analyzed.

[0093] S403, obtaining the original node attributes of each node in the fault graph network.

[0094] S404, analyzing the original node attributes through the attention mechanism to determine the impact value of each original node attribute on the node type.

[0095] S405 , selecting a preset number of original node attributes as target node attributes according to the magnitude of the influence values.

[0096] S406: construct an adjacency matrix of the fault graph network according to the connection relationship between each node in the fault graph network.

[0097] S407, based on the nonlinear propagation formula, according to the adjacency matrix and the target node attributes, determine the feature embedding matrix of each node in the fault graph network.

[0098] S408, based on the vehicle fault analysis model, perform dimensionality reduction processing on the embedded feature matrix after forward update to obtain a first processing result.

[0099] S409, based on the vehicle fault analysis model, perform dimensionality reduction processing on the embedded feature matrix after reverse update to obtain a second processing result.

[0100] S410, fusing the first processing result and the second processing result to obtain a target processing result.

[0101] S411, transforming the target processing result through a normalized exponential function to determine the probability value of each node in the fault spectrum network belonging to a different node type.

[0102] S412, taking the fault type corresponding to the node type with the largest probability value as the fault type corresponding to the vehicle to be analyzed.

[0103] The above fault analysis method obtains the fault data of at least one vehicle to be analyzed; constructs a fault map network according to the fault data of each vehicle to be analyzed; the fault map network is composed of nodes and connecting edges, wherein the nodes represent vehicle components and the connecting edges represent the potential propagation path of the fault from one component to another; determines the feature embedding matrix of each node in the fault map network; wherein the feature embedding matrix is ​​determined by a nonlinear propagation formula; based on the vehicle fault analysis model, the feature embedding matrix is ​​forward updated and reverse updated to determine the fault type corresponding to the vehicle to be analyzed. In the above process, each node in the fault map network of the present application obtains a feature embedding matrix through nonlinear propagation on the connecting edge, effectively captures the high-order nonlinear relationship between different nodes, and ensures that the obtained feature embedding matrix can accurately reflect the own attributes of each node in the fault map network and the associated attributes between nodes. Furthermore, based on the vehicle fault analysis model, the feature embedding matrix is ​​forward updated so that the updated feature embedding matrix is ​​more consistent with the actual fault situation of the vehicle to be analyzed; the feature embedding matrix is ​​reversely updated to continuously optimize the parameters of the vehicle fault analysis model, thereby improving the accuracy of the fault type corresponding to the vehicle to be analyzed and the efficiency of fault type location.

[0104] It should be understood that, although the steps in the flowcharts involved in the above embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0105] Based on the same inventive concept, the embodiment of the present application also provides a fault analysis device for implementing the fault analysis method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more fault analysis device embodiments provided below can refer to the limitations on the fault analysis method above, and will not be repeated here.

[0106] In one embodiment, Figure 5 As shown, a fault analysis device is provided, comprising: a data acquisition module 10, a network construction module 20, a feature matrix determination module 30 and a fault type determination module 40, wherein:

[0107] A data acquisition module 10, used to acquire fault data of at least one vehicle to be analyzed;

[0108] A network construction module 20 is used to construct a fault map network according to the fault data of each vehicle to be analyzed; the fault map network is composed of nodes and connecting edges, wherein the nodes represent vehicle components and the connecting edges represent potential propagation paths of the fault from one component to another component;

[0109] The feature matrix determination module 30 is used to determine the feature embedding matrix of each node in the fault spectrum network; wherein the feature embedding matrix is ​​determined by a nonlinear propagation formula; the nonlinear propagation formula is: ,in, , node j represents the adjacent node with a connecting edge to node i, , N(i) represents the set of adjacent nodes of node i, Indicates The node feature vector of layer node i, represents the inverse function of the polynomial kernel function, represents the node feature vector of node j in layer l, represents the node feature vector of node i in layer l, represents the attention coefficient between node i and node j in layer l, x i represents the initial eigenvector of node i, exp(.) represents the exponential function with e as the base, represents the function variance, ||.|| represents the modulo operation;

[0110] The fault type determination module 40 is used to perform forward and reverse updates on the feature embedding matrix based on the vehicle fault analysis model to determine the fault type corresponding to the vehicle to be analyzed.

[0111] The above-mentioned fault analysis device obtains the fault data of at least one vehicle to be analyzed; constructs a fault map network according to the fault data of each vehicle to be analyzed; the fault map network is composed of nodes and connecting edges, wherein the nodes represent vehicle components and the connecting edges represent the potential propagation path of the fault from one component to another; determines the feature embedding matrix of each node in the fault map network; wherein the feature embedding matrix is ​​determined by a nonlinear propagation formula; based on the vehicle fault analysis model, the feature embedding matrix is ​​forwardly updated and reversely updated to determine the fault type corresponding to the vehicle to be analyzed. In the above process, each node in the fault map network of the present application obtains a feature embedding matrix through nonlinear propagation on the connecting edge, effectively captures the high-order nonlinear relationship between different nodes and the potential propagation path of the fault, and ensures that the obtained feature embedding matrix can accurately reflect the own attributes of each node in the fault map network and the associated attributes between nodes. Furthermore, based on the vehicle fault analysis model, the feature embedding matrix is ​​forwardly updated so that the updated feature embedding matrix is ​​more consistent with the actual fault situation of the vehicle to be analyzed; the feature embedding matrix is ​​reversely updated to continuously optimize the parameters of the vehicle fault analysis model, thereby improving the accuracy of the fault type corresponding to the vehicle to be analyzed and the efficiency of fault type location.

[0112] Each module in the above-mentioned fault analysis device can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0113] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 6As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be realized through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a fault analysis method is realized. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0114] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0115] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0116] Acquiring fault data of at least one vehicle to be analyzed;

[0117] According to the fault data of each vehicle to be analyzed, a fault graph network is constructed; the fault graph network is composed of nodes and connecting edges, where nodes represent vehicle components and connecting edges represent potential propagation paths of faults from one component to another;

[0118] Determine the feature embedding matrix of each node in the fault spectrum network; wherein the feature embedding matrix is ​​determined by the nonlinear propagation formula; the nonlinear propagation formula is: ,in, , node j represents the adjacent node with a connecting edge to node i, , N(i) represents the set of adjacent nodes of node i; Indicates The node feature vector of layer node i, represents the inverse function of the polynomial kernel function, represents the node feature vector of node j in layer l, represents the node feature vector of node i in layer l, represents the attention coefficient between node i and node j in layer l, x i represents the initial eigenvector of node i, exp(.) represents the exponential function with e as the base, represents the function variance, ||.|| represents the modulo operation;

[0119] Based on the vehicle fault analysis model, the feature embedding matrix is ​​updated forwardly and backward to determine the fault type corresponding to the vehicle to be analyzed.

[0120] In one embodiment, when the processor executes the computer program, the following steps are also implemented: based on the vehicle fault analysis model, the embedded feature matrix after forward update is subjected to dimensionality reduction processing to obtain a first processing result; based on the vehicle fault analysis model, the embedded feature matrix after reverse update is subjected to dimensionality reduction processing to obtain a second processing result; based on the first processing result and the second processing result, the fault type corresponding to the vehicle to be analyzed is determined.

[0121] In one embodiment, when the processor executes the computer program, the following steps are also implemented: fusing the first processing result and the second processing result to obtain a target processing result; and determining the fault type corresponding to the vehicle to be analyzed based on the target processing result.

[0122] In one embodiment, when the processor executes the computer program, the following steps are also implemented: the target processing result is transformed by a normalized exponential function to determine the probability value of each node in the fault map network belonging to different node types; the fault type corresponding to the node type with the largest probability value is used as the fault type corresponding to the vehicle to be analyzed.

[0123] In one embodiment, when the processor executes the computer program, the following steps are also implemented: determining the target node attribute of each node in the fault graph network; constructing an adjacency matrix of the fault graph network based on the connection relationship between each node in the fault graph network; based on the nonlinear propagation formula, determining the feature embedding matrix of each node in the fault graph network according to the adjacency matrix and the target node attribute.

[0124] In one embodiment, when the processor executes the computer program, the following steps are also implemented: obtaining the original node attributes of each node in the fault map network; analyzing the original node attributes through the attention mechanism to determine the impact value of each original node attribute on the node type; and selecting a preset number of original node attributes as target node attributes according to the size of the impact value.

[0125] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0126] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, data block or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The data blocks involved in the various embodiments provided in this application may include at least one of relational data blocks and non-relational data blocks. Non-relational data blocks may include distributed data blocks based on blockchain, etc., but are not limited to this. The processors involved in the various embodiments provided in this application may be general-purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, data processing logic devices based on quantum computing, etc., but are not limited to this.

[0127] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0128] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A fault analysis method, characterized in that: The method comprises: Acquiring fault data of at least one vehicle to be analyzed; According to the fault data of each of the vehicles to be analyzed, a fault graph network is constructed; the fault graph network is composed of nodes and connecting edges, wherein the nodes represent vehicle components and the connecting edges represent potential propagation paths of the fault from one component to another component; Determine the feature embedding matrix of each node in the fault spectrum network; wherein the feature embedding matrix is ​​determined by a nonlinear propagation formula; the nonlinear propagation formula is: ,in, , node j represents the adjacent node with a connecting edge to node i, , N(i) represents the set of adjacent nodes of node i, Indicates The node feature vector of layer node i, represents the inverse function of the polynomial kernel function, represents the node feature vector of node j in layer l, represents the node feature vector of node i in layer l, represents the attention coefficient between node i and node j in layer l, x i represents the initial eigenvector of node i, exp(.) represents the exponential function with e as the base, represents the function variance, ||.|| represents the modulo operation; Based on the vehicle fault analysis model, the feature embedding matrix is ​​forward updated and reverse updated to determine the fault type corresponding to the vehicle to be analyzed.

2. The method according to claim 1, characterized in that The method of performing forward and reverse updating on the feature embedding matrix based on the vehicle fault analysis model to determine the fault type corresponding to the vehicle to be analyzed includes: Based on the vehicle fault analysis model, a dimensionality reduction process is performed on the embedded feature matrix after the forward update to obtain a first processing result; Based on the vehicle fault analysis model, a dimensionality reduction process is performed on the embedded feature matrix after reverse update to obtain a second processing result; The fault type corresponding to the vehicle to be analyzed is determined according to the first processing result and the second processing result.

3. The method according to claim 2, characterized in that The determining, according to the first processing result and the second processing result, a fault type corresponding to the vehicle to be analyzed includes: fusing the first processing result and the second processing result to obtain a target processing result; According to the target processing result, the fault type corresponding to the vehicle to be analyzed is determined.

4. The method according to claim 3, characterized in that Determining the fault type corresponding to the vehicle to be analyzed according to the target processing result includes: The target processing result is transformed by a normalized exponential function to determine the probability value of each node in the fault spectrum network belonging to different node types; The fault type corresponding to the node type with the largest probability value is used as the fault type corresponding to the vehicle to be analyzed.

5. The method according to any one of claims 1 to 4, characterized in that The step of determining a feature embedding matrix of each node in the fault graph network comprises: Determining target node attributes of each node in the fault map network; Constructing an adjacency matrix of the fault graph network according to the connection relationship between each node in the fault graph network; Based on the nonlinear propagation formula, a feature embedding matrix of each node in the fault graph network is determined according to the adjacency matrix and the target node attributes.

6. The method according to claim 1, characterized in that The determining of the target node attribute of each node in the fault graph network includes: Obtaining original node attributes of each node in the fault graph network; Analyze the original node attributes through an attention mechanism to determine the impact value of each original node attribute on the node type; A preset number of the original node attributes are selected as the target node attributes according to the magnitude of the influence value.

7. A fault analysis device, characterized in that: The device comprises: A data acquisition module, used to acquire fault data of at least one vehicle to be analyzed; A network construction module, used to construct a fault graph network according to the fault data of each of the vehicles to be analyzed; the fault graph network is composed of nodes and connecting edges, wherein the nodes represent vehicle components and the connecting edges represent potential propagation paths of the fault from one component to another component; The feature matrix determination module is used to determine the feature embedding matrix of each node in the fault spectrum network; wherein the feature embedding matrix is ​​determined by a nonlinear propagation formula; the nonlinear propagation formula is: ,in, , node j represents the adjacent node with a connecting edge to node i, , N(i) represents the set of adjacent nodes of node i, Indicates The node feature vector of layer node i, represents the inverse function of the polynomial kernel function, represents the node feature vector of node j in layer l, represents the node feature vector of node i in layer l, represents the attention coefficient between node i and node j in layer l, x i represents the initial eigenvector of node i, exp(.) represents the exponential function with e as the base, represents the function variance, ||.|| represents the modulo operation; The fault type determination module is used to perform forward and reverse updates on the feature embedding matrix based on the vehicle fault analysis model to determine the fault type corresponding to the vehicle to be analyzed.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images