Automobile cable loss prediction method, device and storage medium based on neural network

Abstract

The present invention discloses a neural network-based automotive cable loss prediction method, device, and storage medium. The method comprises: obtaining environmental data, first temperature data, and second temperature data of the automotive cable; calculating a current stability coefficient based on the first temperature data and a temperature distribution coefficient based on the second temperature data; multiplying the current stability coefficient and the temperature distribution coefficient to obtain a risk coefficient for the automotive cable; calculating a loss degree sequence for the automotive cable based on the environmental data and the risk coefficient; and inputting the loss degree sequence into a preset loss degree prediction model to obtain the loss degree of the automotive cable at the next moment as output by the loss degree prediction model. The present invention uses a neural network model to predict the loss degree of the automotive cable in real time based on the temporal current stability coefficient and spatial temperature distribution characteristics of the automotive cable, combined with the influence of automotive environmental factors, thereby enabling timely detection of cable problems and risk warnings.

Description

Technical Field

[0001] The present invention relates to the technical field of automobile cables, and in particular to a method, device and storage medium for predicting automobile cable loss based on a neural network. Background Art

[0002] During use, new energy vehicles experience frequent current exchange due to the input and output of battery current, causing significant temperature fluctuations. This in turn leads to fatigue in automotive cables, increasing plastic deformation and shrinkage, and affecting the insulation life and performance of automotive cables. While the sheath material experiences smaller temperature fluctuations than the insulation, fatigue creep can still affect the life of new energy vehicle cables to a certain extent.

[0003] However, existing technologies can only regularly monitor the usage of cables in new energy vehicles. The operation process is relatively cumbersome, and regular monitoring cannot cope with the risks caused by emergencies. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a neural network-based automobile cable loss prediction method, device and storage medium. According to the temporal current stability coefficient and spatial temperature distribution characteristics of the automobile cable, combined with the influence of automobile environmental factors, the loss of the cable during use is determined, and the loss degree of the automobile cable is predicted in real time through a neural network model, so that cable problems can be discovered in time and risk warnings can be issued.

[0005] To achieve the above objectives, an embodiment of the present invention provides a method for predicting automotive cable loss based on a neural network, comprising:

[0006] Acquire environmental data, first temperature data, and second temperature data of the vehicle cable; wherein the first temperature data is temperature variation data over time, and the second temperature data is temperature variation data over space;

[0007] Calculating a current stability coefficient of the automotive cable according to the first temperature data, and calculating a temperature distribution coefficient of the automotive cable according to the second temperature data;

[0008] Multiplying the current stability coefficient and the temperature distribution coefficient to obtain a risk coefficient of the automotive cable;

[0009] Calculate the loss degree sequence of the automotive cable according to the environmental data and the risk coefficient;

[0010] The loss degree sequence is input into a preset loss degree prediction model to obtain the loss degree of the automobile cable at the next moment output by the loss degree prediction model.

[0011] As an improvement to the above solution, the calculation formula of the current stability coefficient is:

[0012]

[0013] Where Q represents the current stability coefficient of the automotive cable, T i represents the first temperature, represents the average value of the first temperature data, n represents the number of the first temperature data, and e represents an exponent.

[0014] As an improvement to the above solution, the calculation formula of the temperature distribution coefficient is:

[0015]

[0016] Where W represents the temperature distribution coefficient of the automotive cable, H i represents the second temperature, represents the average value of the second temperature data, H Max Indicates the maximum value of the second temperature data, H Min represents the minimum value of the second temperature data, m represents the number of the second temperature data, and e represents an exponent.

[0017] As an improvement to the above solution, the environmental data includes environmental temperature data and environmental humidity data, and the loss degree sequence of the automotive cable is calculated based on the environmental data and the risk coefficient, specifically including:

[0018] According to the ambient temperature data, the ambient humidity data and the formula Calculate the environmental impact factors;

[0019] In the formula, S represents the environmental impact factor, P i Indicates the ambient temperature, P Max Indicates the maximum value of the ambient temperature data, P Min Indicates the minimum value of the ambient temperature data, L i Indicates ambient humidity, L Max Indicates the maximum value of the ambient humidity data, L Min Indicates the minimum value of the ambient humidity data;

[0020] Multiplying the environmental impact factor and the risk coefficient to obtain the loss degree of each of the automotive cables;

[0021] The loss levels of all automobile cables in the automobile are summarized to obtain a loss level sequence of the automobile cables.

[0022] As an improvement to the above solution, the training method of the loss degree prediction model specifically includes:

[0023] Inputting the loss degree sequence into a temporal convolutional network to obtain an output value of the temporal convolutional network;

[0024] The output value is used as a label, and the temporal convolutional network is optimized according to a preset confidence level and loss function to obtain a loss degree prediction model.

[0025] As an improvement to the above solution, the method further includes:

[0026] Calculating the similarity between different cables according to the current stability coefficient and the risk coefficient;

[0027] Calculate the sample distance between the cables according to the similarity;

[0028] The automotive cables are grouped according to the sample distance based on a K-Means clustering method, and a corresponding threshold is set for each group.

[0029] As an improvement to the above solution, the method further includes:

[0030] comparing the loss level of the automotive cable with a threshold value of each group;

[0031] When the loss level of the automotive cable reaches a threshold value of one group, a risk warning process is performed.

[0032] An embodiment of the present invention further provides a vehicle cable loss prediction device based on a neural network, comprising:

[0033] A data acquisition module, configured to acquire environmental data, first temperature data, and second temperature data of the vehicle cable; wherein the first temperature data is temperature variation data over time, and the second temperature data is temperature variation data over space;

[0034] a first calculation module, configured to calculate a current stability coefficient of the automotive cable according to the first temperature data, and to calculate a temperature distribution coefficient of the automotive cable according to the second temperature data;

[0035] a second calculation module, configured to multiply the current stability coefficient and the temperature distribution coefficient to obtain a risk coefficient of the automotive cable;

[0036] A third calculation module is configured to calculate a loss degree sequence of the automotive cable according to the environmental data and the risk coefficient;

[0037] The prediction module is used to input the loss degree sequence into a preset loss degree prediction model to obtain the loss degree of the automobile cable at the next moment output by the loss degree prediction model.

[0038] An embodiment of the present invention also provides a neural network-based automobile cable loss prediction device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements any of the above-mentioned neural network-based automobile cable loss prediction methods.

[0039] An embodiment of the present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any of the above-mentioned neural network-based automobile cable loss prediction methods.

[0040] Compared to the prior art, the neural network-based automotive cable loss prediction method, device, and storage medium provided by the embodiments of the present invention have the following advantages: by acquiring environmental data, first temperature data, and second temperature data of the automotive cable; wherein the first temperature data is temperature variation data over time, and the second temperature data is temperature variation data over space; calculating the current stability coefficient of the automotive cable based on the first temperature data, and calculating the temperature distribution coefficient of the automotive cable based on the second temperature data; multiplying the current stability coefficient and the temperature distribution coefficient to obtain a risk factor for the automotive cable; calculating a loss degree sequence for the automotive cable based on the environmental data and the risk factor; inputting the loss degree sequence into a preset loss degree prediction model to obtain the loss degree of the automotive cable at the next moment output by the loss degree prediction model. The embodiments of the present invention determine the loss of the cable during use based on the temporal current stability coefficient and spatial temperature distribution characteristics of the automotive cable, combined with the influence of automotive environmental factors, and predict the loss degree of the automotive cable in real time using a neural network model, thereby enabling timely detection of cable problems and risk warnings. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flow chart of a preferred embodiment of a method for predicting automobile cable loss based on a neural network provided by the present invention;

[0042] Figure 2 This is a schematic structural diagram of a preferred embodiment of a neural network-based automobile cable loss prediction device provided by the present invention;

[0043] Figure 3 It is a structural schematic diagram of another preferred embodiment of a neural network-based automobile cable loss prediction device provided by the present invention. DETAILED DESCRIPTION

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0045] See also Figure 1 , Figure 1 The figure is a flow chart of a preferred embodiment of a method for predicting automobile cable loss based on a neural network provided by the present invention. The method for predicting automobile cable loss based on a neural network includes:

[0046] S1, obtaining environmental data, first temperature data, and second temperature data of the vehicle cable; wherein the first temperature data is temperature variation data over time, and the second temperature data is temperature variation data over space;

[0047] S2, calculating a current stability coefficient of the automotive cable according to the first temperature data, and calculating a temperature distribution coefficient of the automotive cable according to the second temperature data;

[0048] S3, multiplying the current stability coefficient and the temperature distribution coefficient to obtain a risk coefficient of the automotive cable;

[0049] S4, calculating a loss degree sequence of the automotive cable according to the environmental data and the risk coefficient;

[0050] S5, inputting the loss degree sequence into a preset loss degree prediction model to obtain the loss degree of the automobile cable at the next moment output by the loss degree prediction model.

[0051] Specifically, due to the power consumption requirements of new energy vehicles for endurance driving, several cables are laid in the vehicle frame. When the current in the battery is output, a large amount of current is continuously output, which will cause the temperature of the cable to rise. Since the cable itself will generate heat when it is powered on, when it is under long-term load operation, the excessively high temperature will accelerate the aging of the insulation layer of the cable itself and affect the service life of the cable. Therefore, it is necessary to collect the temperature data of the new energy vehicle cable over time. The embodiment of the present invention uses a temperature probe sensor to collect the temperature data of the cables in the new energy vehicle. For example, the embodiment of the present invention uses one hour as the time length unit, and the data refresh frequency of the probe temperature sensor is set to 1 minute to obtain the first temperature data of the vehicle cable, that is, the temperature change sequence T = {T1, T2,....., T i}. Based on the fact that the length of the automobile cable also has a certain influence on the current, by installing temperature probes at different positions of each cable, the temperature data of each cable in the new energy vehicle is collected according to the probe temperature sensor to determine the temperature difference between different cables. The greater the difference in the temperature of each cable in space, it means that the insulation layer of the cable has obvious damage, so it is necessary to collect the temperature data of the new energy vehicle cable space. Exemplarily, the embodiment of the present invention uses a probe temperature sensor to read the temperature data of the cable in the new energy vehicle, and sets the temperature probe at a distance of 50 cm (which can be set according to actual conditions), and uses 10 distances as the length unit to obtain the second temperature data of the automobile cable, that is, the temperature distribution sequence H = {H1, H2,....., H i}. Due to the fact that long-term high temperature or humid climate will accelerate the loss of cable insulation, it is also necessary to collect environmental data of new energy vehicles. According to the current output, when the current output is unstable for a period of time, the temperature of the cable is the highest, and the damage to the service life of the cable is also the greatest. Therefore, the current stability coefficient of the automotive cable is calculated based on the first temperature data. The greater the spatial difference in the temperature of each cable, the more obvious the damage to the insulation layer of the cable. Therefore, the temperature distribution coefficient of the automotive cable is calculated based on the second temperature data. The current stability coefficient and the temperature distribution coefficient are multiplied to obtain the risk coefficient of the automotive cable during power-on use. Due to the fact that long-term high temperature or humid climate will accelerate the loss of cable insulation, the loss degree sequence of the automotive cable is calculated in combination with the environmental data and the risk coefficient. The loss degree sequence is input into the preset loss degree prediction model to obtain the loss degree of the automotive cable at the next moment output by the loss degree prediction model.

[0052] This embodiment determines the loss of the cable during use based on the temporal current stability coefficient and spatial temperature distribution characteristics of the automotive cable, combined with the influence of automotive environmental factors, and uses a neural network model to predict the degree of loss of the automotive cable in real time, thereby enabling timely detection of cable problems and risk warnings.

[0053] In another preferred embodiment, the current stability coefficient is calculated as follows:

[0054]

[0055] Where Q represents the current stability coefficient of the automotive cable, T i represents the first temperature, represents the average value of the first temperature data, n represents the number of the first temperature data, and e represents an exponent.

[0056] Specifically, in the embodiment of the present invention, the first temperature data of the automobile cable is obtained as T={T1, T2, ..., T i According to the current output, when the current output is unstable for a period of time, the temperature of the cable is the highest and the damage to the service life of the cable is also the greatest. Therefore, the current stability coefficient of the automotive cable is calculated based on the first temperature data. Where Q represents the current stability coefficient of the automotive cable, T i represents the first temperature, represents the average value of the first temperature data, n represents the number of the first temperature data, and e represents the exponent. To represent the temperature change caused by current fluctuations, the greater the cable temperature change per unit time, the greater the damage caused by frequent current fluctuations to the cable. Combined with the average value of the temperature change per unit time, the current stability coefficient Q is obtained to represent the stability change of the current per unit time.

[0057] In another preferred embodiment, the calculation formula of the temperature distribution coefficient is:

[0058]

[0059] Where W represents the temperature distribution coefficient of the automotive cable, H i represents the second temperature, represents the average value of the second temperature data, H Max Indicates the maximum value of the second temperature data, H Min represents the minimum value of the second temperature data, m represents the number of the second temperature data, and e represents an exponent.

[0060] Specifically, in the embodiment of the present invention, the second temperature data of the automobile cable is obtained as H={H1, H2, ....., H i}, when the temperature difference of each cable in space is greater, it means that the insulation layer of the cable has obvious damage. Therefore, the temperature distribution coefficient of the automotive cable is calculated based on the second temperature data: Where W represents the temperature distribution coefficient of the automotive cable, H i represents the second temperature, represents the average value of the second temperature data, H Max Indicates the maximum value of the second temperature data, H Minrepresents the minimum value of the second temperature data, m represents the number of second temperature data points, and e represents the exponent. Based on the temperature variance collected for each cable segment, the degree of dispersion between segments within that length unit is determined. A greater degree of dispersion indicates a greater temperature difference within that length unit, and therefore a greater risk to the cable's electrical performance. The maximum temperature difference is determined by combining the difference between the maximum and minimum temperature variations within that length unit. This yields W, the temperature distribution coefficient within the unit length.

[0061] After calculating the current stability coefficient Q and temperature distribution coefficient W of the automotive cable, multiply them together to obtain the risk coefficient U = Q * W during power-on use. The greater the difference in current stability per unit time, the greater the risk coefficient of the cable. The greater the temperature distribution coefficient of the cable, the greater the risk coefficient of the cable. U changes with changes in the current stability coefficient and the temperature distribution coefficient.

[0062] In another preferred embodiment, the environmental data includes environmental temperature data and environmental humidity data, and the calculation of the loss degree sequence of the automotive cable based on the environmental data and the risk coefficient specifically includes:

[0063] According to the ambient temperature data, the ambient humidity data and the formula Calculate the environmental impact factors;

[0064] In the formula, S represents the environmental impact factor, P i Indicates the ambient temperature, P Max Indicates the maximum value of the ambient temperature data, P Min Indicates the minimum value of the ambient temperature data, L i Indicates ambient humidity, L Max Indicates the maximum value of the ambient humidity data, L Min Indicates the minimum value of the ambient humidity data;

[0065] Multiplying the environmental impact factor and the risk coefficient to obtain the loss degree of each of the automotive cables;

[0066] The loss levels of all automobile cables in the automobile are summarized to obtain a loss level sequence of the automobile cables.

[0067] Specifically, long-term high temperature or humid climate will accelerate the loss of cable insulation, so it is also necessary to collect environmental data of new energy vehicles. First, collect the ambient temperature data of new energy vehicles and set a starting temperature that will affect the cable. The higher the temperature, the greater the impact on the cable. Collect the ambient temperature data according to the time unit of collecting cable temperature data, and obtain an ambient temperature change sequence P = {P1, P2, ..., P i Secondly, the humidity data of new energy vehicles is collected, and an initial humidity is set to obtain a humidity change sequence L = {L1, L2, ....., L i The higher the ambient temperature and humidity in a new energy vehicle, the greater the damage to the cable life. Therefore, the degree of impact of the new energy vehicle environment on the cable life is calculated based on the ambient temperature data and ambient humidity data to obtain the environmental impact factor. In the formula, S represents the environmental impact factor, P i Indicates the ambient temperature, P Max Indicates the maximum value of the ambient temperature data, P Min Indicates the minimum value of the ambient temperature data, L i Indicates ambient humidity, L Max Indicates the maximum value of the ambient humidity data, L Min Indicates the minimum value of the environmental humidity data. According to the influence of different humidity and temperature, the higher the ambient temperature and humidity in the new energy vehicle, the greater the damage to the cable life. Multiply the environmental impact factor S and the risk factor U to obtain the loss degree of each car cable Y = U * S. The loss degree of the cable changes with the change of the risk factor and environmental factors of the line, and the relationship is positively correlated. Repeat the above operation to collect and summarize the loss degree of all cables in the car, and obtain a cable loss degree change sequence Y = {Y1, Y2,....., Y i}.

[0068] In another preferred embodiment, the training method of the loss degree prediction model specifically includes:

[0069] Inputting the loss degree sequence into a temporal convolutional network to obtain an output value of the temporal convolutional network;

[0070] The output value is used as a label, and the temporal convolutional network is optimized according to a preset confidence level and loss function to obtain a loss degree prediction model.

[0071] Specifically, the embodiment of the present invention uses a TCN time series convolutional network to predict the degree of cable loss of new energy vehicles and predict the degree of cable loss of new energy vehicles at the next moment. The obtained new energy vehicle cable loss degree sequence is used as the first part of the feature sequence and input into the TCN neural network for training. The obtained output value is used as a label so that the TCN can learn the next predicted value of the current sequence. The remaining new energy vehicle cable loss degree sequence is obtained, and the time series convolutional network is optimized according to the preset confidence and loss function. The loss function of TCN is the mean square error loss.

[0072] For this sample sequence, use the confidence level C i As the quality score, and normalized to the sample weight C = {C1,..,C i}.

[0073] Among them, C is the normalized quality score, which serves as the loss weight. is the prediction sample, y i The goal is to ensure convergence of the loss function. Through continuous training, the loss is reduced, and the predicted trend is accurate. Similarly, different temperature variations correspond to different prediction results. The significance of training the TCN is that it can be used to predict the subsequent loss of new energy vehicle cables using data from small to large temperature differences. Based on the predicted results, users can use the TCN to predict the loss level of new energy vehicle cables when the loss level reaches the TCN output standard, indicating that the loss level is about to reach the threshold, and a risk warning should be issued.

[0074] As a preferred solution, the method further comprises:

[0075] Calculating the similarity between different cables according to the current stability coefficient and the risk coefficient;

[0076] Calculate the sample distance between the cables according to the similarity;

[0077] The automotive cables are grouped according to the sample distance based on a K-Means clustering method, and a corresponding threshold is set for each group.

[0078] Specifically, based on the fact that the cables in new energy vehicles themselves will generate heat when energized, the magnitude of the current is proportional to the heat generated by the wires. When laying the cables, according to real-time conditions, it may be necessary to bundle some cables together, and set the length and quantity of the bundles according to the actual situation of the implementer. When the cables are bundled together, the heat dissipation becomes poor, so the temperature will be much higher than that of a single cable, and the damage to the service life of the cables will naturally be greater. The longer the bundle length and the greater the number of cables, the greater the risk of the cables. Therefore, this embodiment groups the density of the cables. When the current stability fluctuations are close, it means that the distance between the cables is closer, and at the same time, the risk coefficient of the cables is closer. Therefore, the cables with close current stability and cable risk coefficient are grouped together, and the similarity between different cables is calculated. The calculation formula for the similarity is:

[0079]

[0080] Among them, A and B are two different cables, δ A δ B is the Pearson similarity coefficient, which is used to indicate the similarity between different data sequences. COV(T A ,T B ) is the overall difference of the temperature change series in the two cables. The greater the difference, the lower the Pearson correlation coefficient, and vice versa. abs(U A -U B ) is the absolute difference between the risk coefficients of cables A and B. The larger the difference, the lower the similarity.

[0081] Based on the obtained similarity between different cables, the sample distance between each cable is calculated, and the cable density is grouped according to the sample distance. The calculation formula of the sample distance is:

[0082]

[0083] Where D is the sample distance of different cable risk similarities. When the similarity of different cables is greater, the sample distance is smaller, and vice versa.

[0084] Based on the sample distance, the K-Means clustering method is used to group the cables of each density. The K value is defined as 3, and all the cables of different density are grouped into different density groups.

[0085] High-density group: The cables in this group are more concentrated and dense, and the risk factor is the highest.

[0086] Medium density group: The cable density in this group is average and the risk factor is medium.

[0087] Low-density group: The cables in this group are sparsely clustered, with low density and low risk factor.

[0088] Due to temperature fluctuations and cable density, different thresholds need to be set for cables of different densities. If a single threshold is set, the damage to the cables will vary from group to group, resulting in errors and causing cable failures. Therefore, three different thresholds are set for the three groups based on actual conditions.

[0089] As a preferred solution, the method further comprises:

[0090] comparing the loss level of the automotive cable with a threshold value of each group;

[0091] When the loss level of the automotive cable reaches a threshold value of one group, a risk warning process is performed.

[0092] Specifically, the present invention uses three threshold values ​​to predict the loss of automotive cables, as predicted by the loss prediction model, and conducts real-time cable loss detection for three types of new energy vehicles. When the loss of each type of new energy vehicle cable reaches the threshold, a risk warning is issued. It should be noted that the thresholds can be customized based on actual circumstances.

[0093] Correspondingly, the present invention also provides a vehicle cable loss prediction device based on a neural network, which can implement all the processes of the vehicle cable loss prediction method based on a neural network in the above embodiment.

[0094] See also Figure 2 , Figure 2 The figure is a schematic diagram of a preferred embodiment of a neural network-based automobile cable loss prediction device provided by the present invention. The neural network-based automobile cable loss prediction device comprises:

[0095] The data acquisition module 201 is used to acquire environmental data, first temperature data, and second temperature data of the vehicle cable; wherein the first temperature data is the temperature variation data over time, and the second temperature data is the temperature variation data over space;

[0096] A first calculation module 202 is configured to calculate a current stability coefficient of the automotive cable based on the first temperature data, and to calculate a temperature distribution coefficient of the automotive cable based on the second temperature data;

[0097] A second calculation module 203 is configured to multiply the current stability coefficient and the temperature distribution coefficient to obtain a risk coefficient of the automotive cable;

[0098] A third calculation module 204 is configured to calculate a loss degree sequence of the automotive cable according to the environmental data and the risk coefficient;

[0099] The prediction module 205 is configured to input the loss degree sequence into a preset loss degree prediction model to obtain the loss degree of the automotive cable at the next moment output by the loss degree prediction model.

[0100] Preferably, the calculation formula of the current stability coefficient is:

[0101]

[0102] Where Q represents the current stability coefficient of the automotive cable, T i represents the first temperature, represents the average value of the first temperature data, n represents the number of the first temperature data, and e represents an exponent.

[0103] Preferably, the calculation formula of the temperature distribution coefficient is:

[0104]

[0105] Where W represents the temperature distribution coefficient of the automotive cable, H i represents the second temperature, represents the average value of the second temperature data, H Max Indicates the maximum value of the second temperature data, H Min represents the minimum value of the second temperature data, m represents the number of the second temperature data, and e represents an exponent.

[0106] Preferably, the environmental data includes environmental temperature data and environmental humidity data, and the step of calculating the loss degree sequence of the automotive cable based on the environmental data and the risk coefficient specifically includes:

[0107] According to the ambient temperature data, the ambient humidity data and the formula Calculate the environmental impact factors;

[0108] In the formula, S represents the environmental impact factor, P i Indicates the ambient temperature, P Max Indicates the maximum value of the ambient temperature data, P Min Indicates the minimum value of the ambient temperature data, L i Indicates ambient humidity, L Max Indicates the maximum value of the ambient humidity data, L Min Indicates the minimum value of the ambient humidity data;

[0109] Multiplying the environmental impact factor and the risk coefficient to obtain the loss degree of each of the automotive cables;

[0110] The loss levels of all automobile cables in the automobile are summarized to obtain a loss level sequence of the automobile cables.

[0111] Preferably, the training method of the loss degree prediction model specifically includes:

[0112] Inputting the loss degree sequence into a temporal convolutional network to obtain an output value of the temporal convolutional network;

[0113] The output value is used as a label, and the temporal convolutional network is optimized according to a preset confidence level and loss function to obtain a loss degree prediction model.

[0114] Preferably, the device is further used for:

[0115] Calculating the similarity between different cables according to the current stability coefficient and the risk coefficient;

[0116] Calculate the sample distance between the cables according to the similarity;

[0117] The automotive cables are grouped according to the sample distance based on a K-Means clustering method, and a corresponding threshold is set for each group.

[0118] Preferably, the device is further used for:

[0119] comparing the loss level of the automotive cable with a threshold value of each group;

[0120] When the loss level of the automotive cable reaches a threshold value of one group, a risk warning process is performed.

[0121] In the specific implementation, the working principle, control process and technical effect of the neural network-based automobile cable loss prediction device provided in the embodiment of the present invention are the same as those of the neural network-based automobile cable loss prediction method in the above embodiment, and will not be repeated here.

[0122] See also Figure 3 , Figure 3 This is a schematic diagram of another preferred embodiment of a neural network-based vehicle cable loss prediction device provided by the present invention. The neural network-based vehicle cable loss prediction device includes a processor 301, a memory 302, and a computer program stored in the memory 302 and configured to be executed by the processor 301. When the processor 301 executes the computer program, it implements the neural network-based vehicle cable loss prediction method described in any of the aforementioned embodiments.

[0123] Preferably, the computer program can be divided into one or more modules / units (e.g., computer program 1, computer program 2, ...). These modules / units are stored in the memory 302 and executed by the processor 301 to implement the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, and these instruction segments are used to describe the execution process of the computer program in the neural network-based automotive cable loss prediction device.

[0124] The processor 301 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor, or the processor 301 can also be any conventional processor. The processor 301 is the control center of the neural network-based automobile cable loss prediction device, and uses various interfaces and lines to connect the various parts of the neural network-based automobile cable loss prediction device.

[0125] The memory 302 mainly includes a program storage area and a data storage area. The program storage area can store an operating system, at least one application required for a function, and the data storage area can store related data. In addition, the memory 302 can be a high-speed random access memory or a non-volatile memory, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, and a Flash Card. Alternatively, the memory 302 can be other volatile solid-state storage devices.

[0126] It should be noted that the above-mentioned vehicle cable loss prediction device based on neural network may include, but is not limited to, a processor and a memory. Those skilled in the art will understand that Figure 3 The structural diagram is only an example of the above-mentioned neural network-based automobile cable loss prediction device and does not constitute a limitation of the above-mentioned neural network-based automobile cable loss prediction device. It may include more or fewer components than shown in the figure, or combine certain components, or different components.

[0127] An embodiment of the present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the neural network-based automobile cable loss prediction method described in any of the above embodiments.

[0128] Embodiments of the present invention provide a neural network-based automotive cable loss prediction method, device, and storage medium. The method involves obtaining environmental data, first temperature data, and second temperature data of an automotive cable; wherein the first temperature data represents temporal temperature variation data, and the second temperature data represents spatial temperature variation data; calculating the current stability coefficient of the automotive cable based on the first temperature data, and calculating the temperature distribution coefficient of the automotive cable based on the second temperature data; multiplying the current stability coefficient and the temperature distribution coefficient to obtain a risk factor for the automotive cable; calculating a loss degree sequence for the automotive cable based on the environmental data and the risk factor; and inputting the loss degree sequence into a preset loss degree prediction model to obtain the loss degree of the automotive cable at the next moment output by the loss degree prediction model. The embodiment of the present invention determines the loss of the automotive cable during use based on the temporal current stability coefficient and spatial temperature distribution characteristics of the automotive cable, combined with the influence of automotive environmental factors. The method then uses a neural network model to predict the loss degree of the automotive cable in real time, thereby enabling timely detection of cable problems and risk warnings.

[0129] It should be noted that the system embodiment described above is merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the drawings of the system embodiment provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which can be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement the present invention without inventive work.

[0130] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for predicting automobile cable loss based on neural network, characterized in that: include: Acquire environmental data, first temperature data, and second temperature data of the vehicle cable; wherein the first temperature data is temperature change data over time, and the second temperature data is temperature change data over space; Calculating a current stability coefficient of the automotive cable according to the first temperature data, and calculating a temperature distribution coefficient of the automotive cable according to the second temperature data; Multiplying the current stability coefficient and the temperature distribution coefficient to obtain a risk coefficient of the automotive cable; Calculate the loss degree sequence of the automotive cable according to the environmental data and the risk coefficient; The loss degree sequence is input into a preset loss degree prediction model to obtain the loss degree of the automobile cable at the next moment output by the loss degree prediction model.

2. The method for predicting automobile cable loss based on a neural network according to claim 1, wherein: The calculation formula of the current stability coefficient is: Where Q represents the current stability coefficient of the automotive cable, T i represents the first temperature, represents the average value of the first temperature data, n represents the number of the first temperature data, and e represents an exponent.

3. The method for predicting automobile cable loss based on neural network according to claim 2, characterized in that: The calculation formula of the temperature distribution coefficient is: Where W represents the temperature distribution coefficient of the automotive cable, H i represents the second temperature, represents the average value of the second temperature data, H Max Indicates the maximum value of the second temperature data, H Min represents the minimum value of the second temperature data, m represents the number of the second temperature data, and e represents an exponent.

4. The method for predicting automobile cable loss based on neural network according to claim 3, characterized in that: The environmental data includes environmental temperature data and environmental humidity data, and the loss degree sequence of the automotive cable is calculated based on the environmental data and the risk coefficient, specifically including: According to the ambient temperature data, the ambient humidity data and the formula Calculate the environmental impact factors; In the formula, S represents the environmental impact factor, P i Indicates the ambient temperature, P Max Indicates the maximum value of the ambient temperature data, P Min Indicates the minimum value of the ambient temperature data, L i Indicates ambient humidity, L Max Indicates the maximum value of the ambient humidity data, L Min Indicates the minimum value of the ambient humidity data; Multiplying the environmental impact factor and the risk coefficient to obtain the loss degree of each of the automotive cables; The loss levels of all automobile cables in the automobile are summarized to obtain a loss level sequence of the automobile cables.

5. The method for predicting automobile cable loss based on neural network according to claim 4, characterized in that: The training method of the loss degree prediction model specifically includes: Inputting the loss degree sequence into a temporal convolutional network to obtain an output value of the temporal convolutional network; The output value is used as a label, and the temporal convolutional network is optimized according to a preset confidence level and loss function to obtain a loss degree prediction model.

6. The method for predicting automobile cable loss based on neural network according to claim 1, characterized in that: The method further comprises: Calculating the similarity between different cables according to the current stability coefficient and the risk coefficient; Calculate the sample distance between the cables according to the similarity; The automotive cables are grouped according to the sample distance based on a K-Means clustering method, and a corresponding threshold is set for each group.

7. The method for predicting automobile cable loss based on neural network according to claim 6, characterized in that: The method further comprises: comparing the loss level of the automotive cable with a threshold value of each group; When the loss level of the automotive cable reaches a threshold value of one group, a risk warning process is performed.

8. A vehicle cable loss prediction device based on neural network, characterized in that: include: A data acquisition module, configured to acquire environmental data, first temperature data, and second temperature data of the vehicle cable; wherein the first temperature data is temperature variation data over time, and the second temperature data is temperature variation data over space; a first calculation module, configured to calculate a current stability coefficient of the automotive cable according to the first temperature data, and to calculate a temperature distribution coefficient of the automotive cable according to the second temperature data; a second calculation module, configured to multiply the current stability coefficient and the temperature distribution coefficient to obtain a risk coefficient of the automotive cable; A third calculation module is configured to calculate a loss degree sequence of the automotive cable according to the environmental data and the risk coefficient; The prediction module is used to input the loss degree sequence into a preset loss degree prediction model to obtain the loss degree of the automobile cable at the next moment output by the loss degree prediction model.

9. A vehicle cable loss prediction device based on neural network, characterized in that: The method comprises a processor and a memory, wherein a computer program is stored in the memory and configured to be executed by the processor, and wherein the neural network-based automobile cable loss prediction method according to any one of claims 1 to 7 is implemented when the processor executes the computer program.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, wherein when the device where the computer-readable storage medium is located executes the computer program, the neural network-based automobile cable loss prediction method according to any one of claims 1 to 7 is implemented.

Images