Vehicle fault prediction methods, devices, readable storage media and equipment

Abstract

This invention discloses a vehicle fault prediction method, apparatus, readable storage medium, and device. The method includes: collecting multiple ECU data and manual troubleshooting items from historical vehicle faults, and determining a vehicle status dataset based on the ECU data and manual troubleshooting items; acquiring multiple fault results corresponding to historical vehicle faults, and determining a vehicle fault dataset based on the multiple fault results; using the vehicle status dataset as input and the vehicle fault dataset as output, inputting the vehicle status dataset into a preset neural network for training to obtain a vehicle fault prediction model; when a vehicle fault detection is detected, acquiring the current ECU data and manual troubleshooting items of the vehicle, and inputting the ECU data and manual troubleshooting items into the vehicle fault prediction model to determine the vehicle fault result. This invention solves the problems of low efficiency and low accuracy in fault detection in the prior art.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, and in particular to a vehicle fault prediction method, apparatus, readable storage medium, and device. Background Technology

[0002] With the rapid development of the Internet, intelligent connected vehicles and other fields, more and more components in vehicles have been transformed into electronic components. When a vehicle malfunctions, it can be quickly located by alarms reported by the ECU in the electronic components. When the actual malfunction of the vehicle corresponds to the alarm reported by the ECU, the problem can be quickly located.

[0003] However, the process of locating vehicle faults is very complex, and most non-critical parts in a vehicle do not have ECUs. In this case, it becomes extremely difficult to locate the fault, requiring experienced personnel to troubleshoot one by one. When encountering difficult faults, it is difficult to locate and resolve them quickly.

[0004] When a vehicle malfunctions, information reported by the vehicle's electronic control unit (ECU) can provide a preliminary diagnosis, but further testing and confirmation are needed to ensure accurate fault location. However, many vehicles lack ECUs. When a component without an ECU malfunctions, relying solely on experience with only the fault symptoms and ECU reports, fault diagnosis becomes difficult. Inexperienced diagnostic personnel often struggle to accurately locate the fault quickly. Furthermore, some complex faults can leave diagnostic personnel at a loss. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a vehicle fault prediction method, apparatus, readable storage medium and device, which aims to solve the problems of low efficiency and low accuracy in fault prediction in the prior art.

[0006] The embodiments of the present invention are implemented as follows:

[0007] A vehicle fault prediction method, the method comprising:

[0008] Collect multiple ECU data points and manual troubleshooting items from historical vehicle malfunctions, and determine the vehicle status dataset based on the ECU data and manual troubleshooting items;

[0009] Obtain multiple fault results corresponding to the vehicle's historical faults, and determine the vehicle fault dataset based on the multiple fault results;

[0010] Using the vehicle status dataset as input and the vehicle fault dataset as output, the vehicle status dataset is input into a preset neural network for training to obtain a vehicle fault prediction model.

[0011] When a fault detection is detected in the vehicle, the current ECU data and manual troubleshooting items of the vehicle are obtained, and the ECU data and manual troubleshooting items are input into the vehicle fault prediction model to determine the fault result of the vehicle.

[0012] Furthermore, in the above-mentioned vehicle fault prediction method, before the step of using the vehicle state dataset as input and the vehicle fault dataset as output, and inputting the vehicle state dataset into a preset neural network for training to obtain a vehicle fault prediction model, the method further includes:

[0013] The vehicle components contained in the vehicle status dataset and the vehicle fault dataset are obtained respectively, and the vehicle status data sequence and the vehicle fault status data sequence are determined based on the vehicle components.

[0014] The vehicle status data sequence and the vehicle fault status data sequence are filled according to the status of the corresponding vehicle components to convert the format of the vehicle status dataset and the vehicle fault dataset.

[0015] The value is set to 1 when the vehicle component is in a normal state and to 0 when the vehicle component is in a fault state.

[0016] Furthermore, in the above-mentioned vehicle fault prediction method, the training process of the vehicle fault prediction model includes:

[0017] The vehicle state dataset is input into a preset neural network for training. Once the loss function value of the preset neural network stabilizes, the vehicle fault prediction model is obtained.

[0018] Furthermore, in the above-mentioned vehicle fault prediction method, the method further includes:

[0019] Obtain the neuron nodes of the preset neural network, and determine the corresponding loss function based on the number of neuron nodes;

[0020] When the neuron node value is one, the loss function is:

[0021] ;

[0022] Where x is the actual value and y is the predicted value;

[0023] When there are multiple neuron node values, the loss function is:

[0024] ;

[0025] Where V1, V2, and V3 are neuron nodes, n is the number of all sample data in the training set, and f(x) i ) represents the true value, y i These are predicted values.

[0026] Furthermore, in the above vehicle fault prediction method, the neuron node formula is as follows:

[0027] ;

[0028] As basic neuron nodes, For learning rate, The distance of backpropagation. These are the neuron nodes after backpropagation.

[0029] Furthermore, in the aforementioned vehicle fault prediction method, the preset neural network includes an output layer, two hidden layers, and an output layer. The hidden layers serve as a continuous mapping from one-dimensional input to another one-dimensional input of the input layer. The first hidden layer is the output of the input layer and also the input of the second hidden layer. The second hidden layer is the output of the first hidden layer and also the input of the output layer.

[0030] Furthermore, in the above-mentioned vehicle fault prediction method, the method further includes:

[0031] The vehicle status dataset includes a validation set and a test set. A first proportion of the vehicle status dataset is randomly selected as the training set, and the remaining second proportion of the vehicle status dataset is used as the test set.

[0032] Wherein, the first proportion is greater than the second proportion, the training set is used for model training, and the test set is used for model validation.

[0033] Another object of the present invention is to provide a vehicle fault prediction device, the device comprising:

[0034] The data acquisition module is used to collect multiple ECU data and manual troubleshooting items from historical vehicle faults, and to determine the vehicle status dataset based on the ECU data and manual troubleshooting items.

[0035] The acquisition module is used to acquire multiple fault results corresponding to the vehicle's historical faults, and determine the vehicle fault dataset based on the multiple fault results.

[0036] The training module is used to take the vehicle status dataset as input and the vehicle fault dataset as output, and input the vehicle status dataset into a preset neural network for training to obtain a vehicle fault prediction model.

[0037] The prediction module is used to obtain the current ECU data and manual troubleshooting items of the vehicle when it is detected that the vehicle needs to perform fault detection, and input the ECU data and manual troubleshooting items into the vehicle fault prediction model to determine the fault result of the vehicle.

[0038] Another object of the present invention is to provide a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0039] Another object of the present invention is to provide an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the steps of the method described above.

[0040] This invention collects multiple ECU data points and manual troubleshooting items from historical vehicle faults to obtain the corresponding vehicle fault dataset; trains a vehicle fault prediction model for fault detection, and finds the correlation between ECU data, manually inspected vehicle component states, and actual faults. After obtaining all ECU data and manual troubleshooting items, the possible location of the fault can be directly obtained, improving the efficiency and accuracy of vehicle ECU fault location, and solving the problems of low efficiency and low accuracy in fault location in existing technologies. Attached Figure Description

[0041] Figure 1 This is a flowchart of the vehicle fault prediction method in the first embodiment of the present invention;

[0042] Figure 2 This is a structural block diagram of the vehicle fault prediction device in the fourth embodiment of the present invention.

[0043] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0044] To facilitate understanding of the present invention, the present invention will be described more fully below with reference to the accompanying drawings. The drawings illustrate several embodiments of the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and comprehensive understanding of the present invention.

[0045] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0046] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0047] The following will describe in detail, with reference to specific embodiments and accompanying drawings, how to improve the efficiency and accuracy of vehicle fault prediction.

[0048] Example 1

[0049] Please see Figure 1 The figure shows a vehicle fault prediction method in the first embodiment of the present invention, the method including steps S10 to S13.

[0050] Step S10: Collect multiple ECU data and manual troubleshooting items from historical vehicle faults, and determine the vehicle status dataset based on the ECU data and manual troubleshooting items.

[0051] To improve the efficiency and accuracy of fault detection, this invention trains a fault prediction model to enable the model to learn the detection patterns of faults and thus detect them. Specifically, it is necessary to collect some data from historical vehicle fault occurrences, including fault information reported by the ECU in the vehicle components, as well as vehicle status information, including whether the vehicle is faulty, which is manually identified by vehicle components without ECUs or ECUs that do not report faults. In practice, there are usually multiple ECU data entries and multiple manual investigation items.

[0052] Step S11: Obtain multiple fault results corresponding to the vehicle's historical faults, and determine the vehicle fault dataset based on the multiple fault results.

[0053] The training model works by using a pre-set neural network. All data reported by the ECU and manual troubleshooting items are taken as network input, while the corresponding final component damage results are taken as output. After the neural network is trained and learned, it can master the rules for determining the component damage results. Thus, after obtaining a set of data reported by the ECU and manual troubleshooting items, it can output the corresponding component damage results, i.e., the vehicle's fault status. Correspondingly, after collecting historical ECU data and manual troubleshooting data when the vehicle malfunctions, the fault results corresponding to the ECU data and manual troubleshooting items are obtained, thereby forming the corresponding vehicle fault dataset, i.e., the output of the neural network.

[0054] Step S12: Using the vehicle status dataset as input and the vehicle fault dataset as output, the vehicle status dataset is input into a preset neural network for training to obtain a vehicle fault prediction model.

[0055] Specifically, the corresponding vehicle status dataset and vehicle fault dataset are used as the input and output of the network, respectively. The neural network can be trained to understand the potential logical relationship between the vehicle status dataset and the vehicle fault dataset, and a vehicle fault prediction model that can predict vehicle faults can be obtained.

[0056] In addition, in some optional embodiments of the present invention, the vehicle state dataset includes a validation set and a test set. The training set is used for model training, and the test set is used for model validation. The test set is used to verify the accuracy of the results.

[0057] In practice, a first proportion of the vehicle status dataset is randomly selected as the training set, and the remaining second proportion of the vehicle status dataset is used as the test set. The first proportion is greater than the second proportion. For example, 80% of the dataset is randomly selected as the training set, and the remaining 20% ​​is used as the test set.

[0058] Step S13: When it is detected that the vehicle needs to be fault detected, the current ECU data and manual troubleshooting items of the vehicle are obtained, and the ECU data and manual troubleshooting items are input into the vehicle fault prediction model to determine the fault result of the vehicle.

[0059] Specifically, when a vehicle fault detection is detected, for example, when the ECU of a vehicle component uploads fault information, or when the vehicle needs to undergo regular fault detection, the current ECU information of the vehicle and the manual inspection information (the fault information of the vehicle detected by the manual inspection) are obtained and input into the vehicle fault prediction model to determine the fault result of the vehicle, including the specific fault location of the vehicle.

[0060] In summary, the vehicle fault prediction method in the above embodiments of the present invention collects multiple ECU data and manual inspection items from historical vehicle faults to obtain the corresponding vehicle fault dataset; trains a vehicle fault prediction model for fault detection; finds the correlation between ECU information, manually inspected vehicle component status, and actual faults; and, after obtaining all ECU information and manual inspection information, can directly determine the possible location of the fault, thereby improving the efficiency and accuracy of vehicle ECU fault location and solving the problems of low efficiency and low accuracy in fault location in the prior art.

[0061] Example 2

[0062] The vehicle fault prediction method in this embodiment differs from the vehicle fault prediction method in Embodiment 1 in that:

[0063] The steps preceding step S12 also include:

[0064] The vehicle components contained in the vehicle status dataset and the vehicle fault dataset are obtained respectively, and the vehicle status data sequence and the vehicle fault status data sequence are determined based on the vehicle components.

[0065] The vehicle status data sequence and the vehicle fault status data sequence are filled according to the status of the corresponding vehicle components to convert the format of the vehicle status dataset and the vehicle fault dataset.

[0066] The value is set to 1 when the vehicle component is in a normal state and to 0 when the vehicle component is in a fault state.

[0067] Specifically, the pre-defined neural network includes an output layer, two hidden layers, and an output layer. The hidden layers serve as a continuous mapping from one-dimensional input to another one-dimensional input. The first hidden layer is the output of the input layer and also the input of the second hidden layer. The second hidden layer is the output of the first hidden layer and also the input of the output layer. The input layer consists of a one-dimensional variable, composed of the states reported by all vehicle ECUs and the states of manually inspected components. The structure of the two hidden layers can fit a smooth mapping with a defined accuracy. The output layer consists of a one-dimensional variable, composed of specific fault results.

[0068] More specifically, the data construction of the model input layer needs to format all data into computable data. Specifically, the input format, such as [tire pressure status, engine status, battery charge status, DFP status, etc.], is converted into the computable [1, 1, 1, 1, 1, 1...]. The tire pressure status corresponds to the first position of [1, 1, 1, 1...] in the data, and similarly, the remaining positions correspond one-to-one with the values ​​in the array. A value of 1 indicates that the corresponding position may be in a fault state, and a value of 0 indicates that the corresponding position is normal. A two-layer linear model is used as the hidden layer, mapping the actual fault results, such as [abnormal tire pressure, engine failure, low battery charge, etc.], to [1, 1, 1...] as the output layer. The engine failure corresponds to the first position of the array [1, 1, 1...]. A value of 1 indicates that the corresponding component is faulty, and a value of 0 indicates that the corresponding component is not faulty.

[0069] Furthermore, the training process of the vehicle fault prediction model includes:

[0070] The vehicle state dataset is input into a preset neural network for training. Once the loss function value of the preset neural network stabilizes, the vehicle fault prediction model is obtained.

[0071] Specifically, the training set is input into the prediction model to obtain its prediction results, which are then processed using a loss function. To obtain the degree of loss, x represents the true value and y represents the predicted value. Calculate the gradient of the node weights. The direction of the gradient represents the direction of error amplification, and the inverse represents the direction of error reduction. Invert the gradient of each node and add it proportionally to the weight of each node. Repeat the steps until the deviation between the predicted value and the true value is within the set range.

[0072] More specifically, the loss function is calculated using the mean square method. The loss function is the squared difference between the predicted and actual values. Backpropagation uses gradient descent, which employs the gradient of a differentiable and computable function as the direction of propagation, moving in the direction of decreasing gradient to locally minimize the function's value. (The formula is used here.) The cost function for the above calculation process is defined as follows: where V1, V2, and V3 are neuron nodes, n is the number of samples in the training set, and f(x) is the cost function. i ) represents the true value, y i The predicted value is the distance traveled during each backpropagation. , For function C in The partial derivative in the direction is similar. For function C in Partial derivative in direction. That is, the partial derivative of function C in... The rate of change in direction. In order to propagate backwards, to The distance moved in the direction, The formula for the change of variable v, is the learning rate, and is a very small integer. , where is the distance of backpropagation. These are the neuron nodes after backpropagation. Using the basic neuron nodes, the above process is repeatedly calculated, which is the implementation principle of gradient descent. The parameter values ​​of the hidden layer are adjusted in the manner of gradient descent, and the above process is repeated until the parameters of the hidden layer tend to stabilize.

[0073] In summary, the vehicle fault prediction method in the above embodiments of the present invention collects multiple ECU data and manual inspection items from historical vehicle faults to obtain the corresponding vehicle fault dataset; trains a vehicle fault prediction model for fault detection; finds the correlation between ECU information, manually inspected vehicle component status, and actual faults; and, after obtaining all ECU information and manual inspection information, can directly determine the possible location of the fault, thereby improving the efficiency and accuracy of vehicle ECU fault location and solving the problems of low efficiency and low accuracy in fault location in the prior art.

[0074] Example 3

[0075] Please see Figure 2 The image shows a vehicle fault prediction device proposed in the third embodiment of the present invention, the device comprising:

[0076] The data acquisition module 100 is used to collect multiple ECU data and manual troubleshooting items during historical vehicle faults, and to determine the vehicle status dataset based on the ECU data and manual troubleshooting items.

[0077] The acquisition module 200 is used to acquire multiple fault results corresponding to the vehicle's historical faults, and determine the vehicle fault dataset based on the multiple fault results.

[0078] The training module 300 is used to take the vehicle state dataset as input and the vehicle fault dataset as output, and input the vehicle state dataset into a preset neural network for training to obtain a vehicle fault prediction model.

[0079] The prediction module 400 is used to obtain the current ECU data and manual troubleshooting items of the vehicle when it is detected that the vehicle needs to be fault detected, and input the ECU data and manual troubleshooting items into the vehicle fault prediction model to determine the fault result of the vehicle.

[0080] Furthermore, in some optional embodiments of the present invention, the apparatus further includes:

[0081] The sequence determination module is used to acquire the vehicle components contained in the vehicle status dataset and the vehicle fault dataset respectively, and determine the vehicle status data sequence and the vehicle fault status data sequence based on the vehicle components.

[0082] The format conversion module is used to fill the vehicle status data sequence and the vehicle fault status data sequence according to the status of the corresponding vehicle components to convert the format of the vehicle status dataset and the vehicle fault dataset.

[0083] The value is set to 1 when the vehicle component is in a normal state and to 0 when the vehicle component is in a fault state.

[0084] Furthermore, in some optional embodiments of the present invention, the training process of the vehicle fault prediction model includes:

[0085] The vehicle state dataset is input into a preset neural network for training. Once the loss function value of the preset neural network stabilizes, the vehicle fault prediction model is obtained.

[0086] Furthermore, in some optional embodiments of the present invention, the apparatus further includes:

[0087] The loss function determination module is used to obtain the neuron nodes of the preset neural network and determine the corresponding loss function based on the number of neuron nodes.

[0088] When the neuron node value is one, the loss function is:

[0089] ;

[0090] Where x is the actual value and y is the predicted value;

[0091] When there are multiple neuron node values, the loss function is:

[0092] ;

[0093] Where V1, V2, and V3 are neuron nodes, n is the number of all sample data in the training set, and f(x) i ) represents the true value, y i These are predicted values.

[0094] Furthermore, in some optional embodiments of the present invention, the neuron node formula is as follows:

[0095] ;

[0096] As basic neuron nodes, For learning rate, The distance of backpropagation. These are the neuron nodes after backpropagation.

[0097] Furthermore, in some optional embodiments of the present invention, the preset neural network includes an output layer, two hidden layers, and an output layer. The hidden layers serve as a continuous mapping from one-dimensional input to another one-dimensional input of the input layer. The first hidden layer is the output of the input layer and also the input of the second hidden layer. The second hidden layer is the output of the first hidden layer and also the input of the output layer.

[0098] Furthermore, in some optional embodiments of the present invention, the apparatus further includes:

[0099] The diversity module is used to randomly select a first proportion of the vehicle state dataset as the training set and the remaining second proportion of the vehicle state dataset as the test set, when the vehicle state dataset includes a validation set and a test set.

[0100] Wherein, the first proportion is greater than the second proportion, the training set is used for model training, and the test set is used for model validation.

[0101] The functions or operation steps implemented by the above modules are largely the same as those in the above method embodiments, and will not be repeated here.

[0102] Example 4

[0103] In another aspect, the present invention provides a readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the steps of the method described in any one of the above embodiments one to two.

[0104] Example 5

[0105] In another aspect, the present invention provides an electronic device, the electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any one of the methods described in embodiments one to two above.

[0106] The technical features of the above embodiments can be combined in any way. For the sake of brevity, 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.

[0107] Those skilled in the art will understand that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequential list of executable instructions for implementing logical functions, and can be embodied in any computer-readable storage medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable storage medium" can mean any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0108] More specific examples (a non-exhaustive list) of computer-readable storage media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable storage media can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0109] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0110] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0111] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A vehicle fault prediction method, characterized in that, The method includes: Collect multiple ECU data points and manual troubleshooting items from historical vehicle malfunctions, and determine the vehicle status dataset based on the ECU data and manual troubleshooting items; Obtain multiple fault results corresponding to the vehicle's historical faults, and determine the vehicle fault dataset based on the multiple fault results; Using the vehicle status dataset as input and the vehicle fault dataset as output, the vehicle status dataset is input into a preset neural network for training to obtain a vehicle fault prediction model. When it is detected that the vehicle needs to be fault detected, the current ECU data and manual troubleshooting items of the vehicle are obtained, and the ECU data and manual troubleshooting items are input into the vehicle fault prediction model to determine the fault result of the vehicle. Before the step of using the vehicle status dataset as input and the vehicle fault dataset as output, and inputting the vehicle status dataset into a preset neural network for training to obtain a vehicle fault prediction model, the method further includes: The vehicle components contained in the vehicle status dataset and the vehicle fault dataset are obtained respectively, and the vehicle status data sequence and the vehicle fault status data sequence are determined based on the vehicle components. The vehicle status data sequence and the vehicle fault status data sequence are filled according to the status of the corresponding vehicle components to convert the format of the vehicle status dataset and the vehicle fault dataset. The value is set to 1 when the vehicle component is in a normal state and to 0 when the vehicle component is in a fault state. The training process of the vehicle fault prediction model includes: The vehicle state dataset is input into a preset neural network for training. Once the loss function value of the preset neural network stabilizes, the vehicle fault prediction model is obtained.

2. The vehicle fault prediction method according to claim 1, characterized in that, The method further includes: Obtain the neuron nodes of the preset neural network, and determine the corresponding loss function based on the number of neuron nodes; When the neuron node value is one, the loss function is: ； Where x is the actual value and y is the predicted value; When there are multiple neuron node values, the loss function is: ； Where V1, V2, and V3 are neuron nodes, n is the number of all sample data in the training set, and f(x) i ) represents the true value, y i These are predicted values.

3. The vehicle fault prediction method according to claim 2, characterized in that, The neuron node formula: ； As basic neuron nodes, For learning rate, The distance of backpropagation. These are the neuron nodes after backpropagation.

4. The vehicle fault prediction method according to claim 1, characterized in that, The preset neural network includes an output layer, two hidden layers, and an output layer. The hidden layer serves as a continuous mapping from one-dimensional input to another one-dimensional input of the input layer. The first hidden layer is the output of the input layer and also the input of the second hidden layer. The second hidden layer is the output of the first hidden layer and also the input of the output layer.

5. The vehicle fault prediction method according to any one of claims 1 to 4, characterized in that, The method further includes: The vehicle status dataset includes a validation set and a test set. A first proportion of the vehicle status dataset is randomly selected as the training set, and the remaining second proportion of the vehicle status dataset is used as the test set. Wherein, the first proportion is greater than the second proportion, the training set is used for model training, and the test set is used for model validation.

6. A vehicle fault prediction device, characterized in that, The apparatus for implementing the vehicle fault prediction method according to any one of claims 1 to 5, the apparatus comprising: The data acquisition module is used to collect multiple ECU data and manual troubleshooting items from historical vehicle faults, and to determine the vehicle status dataset based on the ECU data and manual troubleshooting items. The acquisition module is used to acquire multiple fault results corresponding to the vehicle's historical faults, and determine the vehicle fault dataset based on the multiple fault results. The training module is used to take the vehicle status dataset as input and the vehicle fault dataset as output, and input the vehicle status dataset into a preset neural network for training to obtain a vehicle fault prediction model. The prediction module is used to obtain the current ECU data and manual troubleshooting items of the vehicle when it is detected that the vehicle needs to perform fault detection, and input the ECU data and manual troubleshooting items into the vehicle fault prediction model to determine the fault result of the vehicle.

7. A readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method as described in any one of claims 1 to 5.

8. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the steps of the method as described in any one of claims 1 to 5.

Images