A new energy vehicle charging safety monitoring method, system, device and medium

Abstract

The application relates to a new energy vehicle charging safety monitoring method, system, device and medium. The method comprises the following steps: acquiring battery data and charging pile data of a vehicle; performing standardization processing on the battery data and the charging pile data of the vehicle to obtain standardized vehicle battery data and standardized charging pile data; fusing the standardized vehicle battery data and the standardized charging pile data to obtain vehicle-pile fusion data; performing feature extraction on the vehicle-pile fusion data to obtain vehicle battery feature parameters and charging pile feature parameters; inputting the vehicle battery feature parameters and the charging pile feature parameters into a power battery voltage prediction model to obtain a power battery voltage prediction value; obtaining a fault diagnosis result according to the power battery voltage prediction value and a power battery voltage; and comprehensively evaluating the charging safety risk of the vehicle according to the fault diagnosis result to obtain a safety risk evaluation result. The method can improve the charging safety monitoring accuracy and improve the charging safety.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of new energy vehicles, and particularly relates to a new energy vehicle charging safety monitoring method, system, device and medium. BACKGROUND

[0002] With the rapid development of new energy vehicle technology, the number of new energy vehicles is increasing, and the charging safety problem is increasingly concerned. At present, new energy vehicle charging safety monitoring mainly relies on experimental data or simulation data, and the research objects are mostly limited number of power battery monomers or power battery groups. In the traditional technology, various tests are generally carried out on the power battery under laboratory conditions to obtain its characteristic parameters under different working conditions, and then a charging safety monitoring model is constructed.

[0003] However, there is a big difference between laboratory conditions and actual charging environment. There are problems such as noise interference, data quality and variable working condition environment in real vehicle data. It is difficult to completely express the safety characteristics in the actual charging process by directly using the characteristic parameters under laboratory conditions, which leads to the limitation of the existing monitoring method in actual application. In addition, in the aspect of charging fault diagnosis, there is currently a lack of charging voltage fault diagnosis model research oriented to vehicle and pile collaborative operation data. The charging safety main body includes new energy vehicles and charging piles, and the state of the power battery of the new energy vehicle, the working state of the charging pile and the coupling effect between them need to be considered comprehensively to make a reasonable evaluation of the charging safety risk. However, the research on the integration of data resources of new energy vehicles and charging piles and the fusion of safety indexes in the prior art is not deep enough, and it cannot meet the needs of actual charging safety monitoring. SUMMARY

[0004] Therefore, it is necessary to provide a new energy vehicle charging safety monitoring method, system, device and medium capable of improving charging safety in view of the above technical problems.

[0005] In a first aspect, the application provides a new energy vehicle charging safety monitoring method, comprising:

[0006] obtaining battery data and charging pile data of the vehicle; the battery data of the vehicle includes power battery voltage, current and temperature parameters; the charging pile data includes charging pile output voltage, charging pile current, charging pile temperature and insulation resistance;

[0007] standardizing the battery data and charging pile data of the vehicle to obtain standardized vehicle battery data and standardized charging pile data;

[0008] fusing the standardized vehicle battery data and the standardized charging pile data to obtain vehicle-pile fusion data;

[0009] Feature extraction is performed on the vehicle-pile fusion data to obtain automobile battery feature parameters and charging pile feature parameters;

[0010] The automobile battery feature parameters and the charging pile feature parameters are input into a preset power battery voltage prediction model to obtain a power battery voltage prediction value; the power battery voltage prediction model comprises a convolution layer, a bidirectional gate control recurrent unit, an attention mechanism layer, and a full connection layer;

[0011] According to the power battery voltage prediction value and the power battery voltage, fault diagnosis is performed on the power battery of the automobile to obtain a fault diagnosis result;

[0012] According to the fault diagnosis result, a comprehensive evaluation is performed on the charging safety risk of the automobile to obtain a safety risk evaluation result; the safety risk evaluation result comprises a safety risk evaluation score.

[0013] In one of the embodiments, the automobile battery feature parameters and the charging pile feature parameters are input into a preset power battery voltage prediction model to obtain a power battery voltage prediction value, comprising:

[0014] The automobile battery feature parameters and the charging pile feature parameters are spliced to obtain a joint feature vector; the joint feature vector is input into the power battery voltage prediction model;

[0015] Local features are extracted from the joint feature vector by the convolution layer to output a convolution feature map;

[0016] Time sequence features are extracted from the convolution feature map by the bidirectional gate control recurrent unit to obtain a time sequence feature vector; the time sequence feature vector comprises a forward hidden state and a backward hidden state;

[0017] The time sequence feature vector is weighted by the attention mechanism layer to obtain a weighted feature vector;

[0018] The weighted feature vector is input into the full connection layer to obtain the power battery voltage prediction value.

[0019] In one of the embodiments, the standardized automobile battery data and the standardized charging pile data are fused to obtain vehicle-pile fusion data, comprising:

[0020] Charging order data of the automobile is obtained; the charging order data comprises a vehicle identification code, an order creation time, a charging start time, and a charging end time;

[0021] The standardized automobile battery data is matched with the charging order data according to the vehicle identification code and the order creation time to obtain preliminary matching data;

[0022] The preliminary matching data is matched in a time window according to the charging start time and the charging end time to obtain time window matching data;

[0023] The time window matching data is data-aligned and spliced to obtain preliminary vehicle-pile fusion data;

[0024] The preliminary vehicle-pile fusion data is denoised to obtain vehicle-pile fusion data.

[0025] In one of the embodiments, feature extraction is performed on the vehicle-pile fusion data to obtain automobile battery feature parameters and charging pile feature parameters, including:

[0026] The vehicle-pile fusion data is divided by a sliding window to obtain a plurality of charging segment data;

[0027] Feature parameters are extracted from each charging segment data to obtain automobile battery feature parameters and charging pile feature parameters; the automobile battery feature parameters include voltage mean, current variance and temperature slope; the charging pile feature parameters include output voltage fluctuation, output current stability and insulation resistance change rate.

[0028] In one of the embodiments, according to the power battery voltage prediction value and the power battery voltage, the power battery of the automobile is diagnosed for fault to obtain a fault diagnosis result, including:

[0029] The residual value of the power battery voltage prediction value and the power battery voltage is calculated using the following formula to obtain a residual sequence:

[0030] ;

[0031] Wherein, is the residual value of the i th time step, is the power battery voltage of the i th time step, is the power battery voltage prediction value of the i th time step; The residual mean and the residual standard deviation of the residual sequence are obtained using the following formula:

[0032] ;

[0033] ;

[0034] Wherein, is the residual mean, is the residual standard deviation, is the window size, is the i th residual value in the window; According to the residual mean and the residual standard deviation, the upper limit of the dynamic threshold and the lower limit of the dynamic threshold are obtained using the following formula:

[0035]

[0036] ; ​​​

[0037] wherein, is an upper limit of the dynamic threshold, is a lower limit of the dynamic threshold, is a residual mean, is a residual standard deviation;

[0038] a dynamic threshold range is obtained according to the upper limit of the dynamic threshold and the lower limit of the dynamic threshold;

[0039] a fault anomaly result is obtained when the residual value exceeds the dynamic threshold range; the fault anomaly result includes an anomaly time point and an anomaly type;

[0040] a fault diagnosis result is obtained based on the fault anomaly result; the fault diagnosis result includes a fault type, a fault occurrence frequency and a fault duration time.

[0041] In one of the embodiments, the fault diagnosis result is obtained based on the fault anomaly result, including:

[0042] a cluster analysis is performed on the anomaly time point in the fault anomaly result to obtain an anomaly event cluster;

[0043] a feature extraction is performed on the anomaly event cluster to obtain a fault frequency feature and a fault duration time; the fault frequency feature is used to represent the number of times of the fault occurrence in a preset time period; the fault duration time is used to represent a time period of the anomaly event occurrence;

[0044] a fault type classification result is obtained according to the fault frequency feature and the fault duration time; the fault type classification result is used to indicate the type of the fault;

[0045] the fault diagnosis result is obtained according to the fault type classification result.

[0046] In one of the embodiments, according to the fault diagnosis result, a comprehensive evaluation is performed on the charging safety risk of the automobile to obtain a safety risk evaluation result, including:

[0047] a risk evaluation index is obtained according to the fault diagnosis result; the risk evaluation index includes an automobile battery characteristic parameter and a charging pile characteristic parameter;

[0048] a comprehensive weight of each risk evaluation index is calculated using the following formula:

[0049] ;

[0050] wherein, is a comprehensive weight of the i-th risk evaluation index, is a judgment matrix element, indicating that the i-th risk evaluation index is relative to the j-th risk evaluation index, is a comprehensive weight of the i-th risk evaluation index, is a judgment matrix element, indicating that the i-th risk evaluation index is relative to the j-th risk evaluation index, importance of the risk evaluation index, total number of risk evaluation indexes, order of the judgment matrix;

[0051] correcting the comprehensive weight of each risk evaluation index to obtain a corrected comprehensive weight of each risk evaluation index;

[0052] According to the corrected comprehensive weight of each risk evaluation index and each risk evaluation index, using the following formula, the safety risk evaluation score is obtained:

[0053] ;

[0054] wherein, safety risk evaluation score, corrected comprehensive weight of the i-th risk evaluation index, quantitative value of the i-th risk evaluation index, total number of risk evaluation indexes; According to the safety risk evaluation score, the safety risk level is obtained; the safety risk level includes a first safety level, a second safety level and a third safety level.

[0055] According to the safety risk level, the safety risk evaluation result is obtained; the safety risk evaluation result is used to indicate the safety risk of the power battery of the automobile during charging.

[0056]

[0057] In a second aspect, the present application also provides a new energy automobile charging safety monitoring system, comprising:

[0058] a data acquisition module, configured to acquire battery data and charging pile data of the automobile; the battery data of the automobile includes power battery voltage, current and temperature parameters; the charging pile data includes charging pile output voltage, charging pile current, charging pile temperature and insulation resistance;

[0059] a data standardization processing module, configured to perform standardization processing on the battery data and the charging pile data of the automobile to obtain standardized battery data and standardized charging pile data;

[0060] a vehicle-pile data fusion module, configured to fuse the standardized battery data and the standardized charging pile data to obtain vehicle-pile fusion data;

[0061] a feature parameter extraction module, configured to extract features from the vehicle-pile fusion data to obtain automobile battery feature parameters and charging pile feature parameters;

[0062] ​​The battery voltage prediction module is used to input the characteristic parameters of the vehicle battery and the characteristic parameters of the charging pile into the preset power battery voltage prediction model to obtain the predicted value of the power battery voltage. The power battery voltage prediction model includes a convolutional layer, a bidirectional gate control loop unit, an attention mechanism layer and a fully connected layer.

[0063] The fault diagnosis module is used to diagnose faults in the vehicle's power battery based on the predicted value of the power battery voltage and the power battery voltage, and obtain the fault diagnosis results.

[0064] The safety evaluation module is used to comprehensively evaluate the charging safety risks of automobiles based on the fault diagnosis results, and obtain the safety risk evaluation results, which include a safety risk evaluation score.

[0065] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method described in the first aspect.

[0066] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described in the first aspect.

[0067] The aforementioned method, system, equipment, and medium for monitoring the charging safety of new energy vehicles acquire multi-dimensional data from vehicle batteries and charging piles. This data is then standardized to eliminate the impact of data discrepancies, and finally fused to form vehicle-pile integrated data to break down data silos. Subsequently, feature parameters are extracted and input into a voltage prediction model containing structures such as convolutional layers and bidirectional gate control loop units. Combined with actual voltage data, fault diagnosis is achieved, and finally, a safety risk assessment is completed based on the diagnostic results. This method effectively solves problems such as significant differences between laboratory and actual charging environments and noise interference with real-vehicle data. It fully integrates vehicle-pile data resources, fuses safety indicators, considers the coupling effect between the two, avoids the limitations of traditional monitoring methods, improves the accuracy and applicability of charging safety monitoring, meets the safety monitoring needs in actual charging scenarios, and provides reliable protection for the charging safety of new energy vehicles. Attached Figure Description

[0068] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0069] Figure 1 This is a flowchart illustrating a method for monitoring the charging safety of new energy vehicles in one embodiment.

[0070] Figure 2 FIG. 1 is a structural schematic diagram of a new energy vehicle charging safety monitoring system according to an embodiment;

[0071] Figure 3 FIG. 2 is a schematic diagram of a computer device according to an embodiment. DETAILED DESCRIPTION

[0072] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not to limit the present application.

[0073] In one embodiment, as shown in FIG. 1, a new energy vehicle charging safety monitoring method is provided, and the present embodiment takes the method applied to a safety monitoring terminal (referred to as a terminal) as an example for illustration. It should be understood 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 realized through the interaction of the terminal and the server. In the present embodiment, the method includes the following steps: Figure 1 S1, obtaining battery data and charging pile data of the vehicle.

[0074] Exemplarily, the safety monitoring terminal establishes a data interaction link with the battery management system of the new energy vehicle and the communication module of the charging pile, and obtains the battery data and the charging pile data of the vehicle. The battery data of the vehicle includes power battery voltage, current and temperature parameters. The power battery voltage is used to reflect the potential difference between the positive and negative electrodes of the battery, and is a basic parameter for judging whether the battery is in a normal charging and discharging interval. The current represents the internal charge flow rate of the battery, and is directly related to the charging power and the battery heating condition. The temperature parameters include the battery monomer temperature and the battery pack overall temperature, and both excessively high and excessively low temperatures will affect the battery performance and safety. The charging pile data includes charging pile output voltage, charging pile current, charging pile temperature and insulation resistance. The charging pile output voltage needs to be matched with the required voltage of the vehicle battery to ensure the stability of the charging process. The charging pile current needs to be controlled within the tolerance range of the battery to determine the charging speed. The charging pile temperature monitoring can timely find the overheating problem of the internal circuit of the charging pile. The insulation resistance is used to evaluate the electrical insulation performance between the charging pile and the vehicle, and a too low resistance value may cause a risk of electric leakage.

[0075] S2, standardizing the battery data and the charging pile data of the vehicle to obtain standardized battery data and standardized charging pile data.

[0076]

[0077] ​Exemplarily, the safety monitoring terminal standardizes the automobile battery data and the charging pile data. Through the standardization processing, the influence of the differences in dimensions and numerical ranges between different data types is eliminated, thereby laying a unified foundation for subsequent data fusion and analysis. Specifically, since the units and numerical magnitudes of the voltage, current, and temperature in the automobile battery data and the output voltage, current, temperature, and insulation resistance in the charging pile data are different, for example, the unit of voltage can be volt (V), the unit of current can be ampere (A), and the unit of temperature can be Celsius degree (℃), and the unit of insulation resistance can be mega-ohm (MΩ), if the original data is directly used for analysis, the numerical differences can cause some parameters to be excessively amplified or weakened in the calculation, thereby affecting the accuracy of the analysis result. The safety monitoring terminal maps all kinds of data into the same data interval, calculates the mean and standard deviation of the original data of each parameter, and then converts the original data into dimensionless standardized data according to a preset standardization formula, thereby obtaining standardized automobile battery data and standardized charging pile data, so that different types of data are comparable and fusible.

[0078] S3, fusing the standardized automobile battery data and the standardized charging pile data to obtain vehicle-pile fusion data.

[0079] Exemplarily, the safety monitoring terminal fuses the standardized automobile battery data and the standardized charging pile data, so as to break the data islands of the vehicle and the pile and construct vehicle-pile fusion data that comprehensively reflects the state of the charging process. During the charging process, the automobile battery state and the charging pile operating state have strong correlation. Specifically, the change of the charging pile output current directly affects the change of the battery charging current and temperature, and it is difficult to comprehensively evaluate the charging safety by relying on single data. Through the data fusion technology, the safety monitoring terminal first analyzes the data correlation of the two types of standardized data, identifies the battery data and the charging pile data that match each other in the same time dimension, and then adopts data splicing and integration strategies to correspondingly associate the battery parameters and the charging pile parameters that reflect the same charging moment or the same charging stage, thereby forming vehicle-pile fusion data containing multi-dimensional state information of the vehicle and the pile. The vehicle-pile fusion data not only contains the core parameters such as the voltage, current, and temperature of the battery, but also integrates the output characteristics and insulation performance parameters of the charging pile, so as to more comprehensively and accurately reflect the overall operating state during the charging process.

[0080] S4, extracting features from the vehicle-pile fusion data to obtain automobile battery feature parameters and charging pile feature parameters.

[0081] Exemplarily, the security monitoring terminal extracts features from the vehicle-pile fusion data, that is, extracts key information capable of representing the operating states of the vehicle battery and the charging pile from the massive fusion data to obtain vehicle battery feature parameters and charging pile feature parameters. The vehicle-pile fusion data contains a large amount of time-series raw data, and direct use for analysis may have problems such as data redundancy and non-prominence of key information. Among them, the vehicle battery feature parameters include power battery voltage parameters, current feature parameters, temperature feature parameters, state of charge (SOC) feature parameters, capacity degradation rate feature parameters, internal resistance growth rate feature parameters, and voltage consistency feature parameters. The power battery voltage parameters reflect the battery health state by analyzing the voltage trend; the current feature parameters pay attention to the current fluctuation to determine whether the charging process is stable; the temperature feature parameters track the temperature change rate to warn of overheating risks; the SOC feature parameters are calculated by combining the charging time and current data to reflect the current power level of the battery; the capacity degradation rate feature parameters evaluate the battery aging degree by comparing the current capacity and the initial capacity of the battery; the internal resistance growth rate feature parameters are calculated based on the voltage and current data, and the internal resistance growth is an important indicator of battery performance decline; and the voltage consistency feature parameters analyze the voltage difference between battery monomers, and excessive difference will affect the overall performance and safety of the battery. Among them, the charging pile feature parameters include charging pile output voltage feature parameters, charging pile current feature parameters, charging pile temperature feature parameters, insulation resistance feature parameters, charging efficiency feature parameters, and fault frequency feature parameters. The output voltage feature parameters monitor the voltage stability to avoid voltage fluctuations damaging the battery; the output current feature parameters evaluate the smoothness of the current output; the temperature feature parameters monitor the self-heating of the charging pile; the insulation resistance feature parameters track the changes in insulation performance; the charging efficiency feature parameters are obtained by calculating the ratio of the actual charging capacity to the theoretical charging capacity; and the fault frequency feature parameters are used to represent the historical fault occurrence frequency of the charging pile.

[0082] S5, input the vehicle battery feature parameters and the charging pile feature parameters into a preset power battery voltage prediction model to obtain a power battery voltage prediction value.

[0083] Specifically, the safety monitoring terminal inputs the battery characteristic parameters and the charging pile characteristic parameters into a preset power battery voltage prediction model to realize accurate prediction of future voltage changes of the power battery and obtain a power battery voltage prediction value. The power battery voltage prediction model combines multiple deep learning technologies, and each layer structure bears different functions: the convolution layer extracts local features from the input characteristic parameters by setting multiple convolution kernels, can capture local correlation information between the characteristic parameters, such as the correlation between voltage and current changes in a short time, and the output convolution feature map integrates local key features; the bidirectional gated recurrent unit (Bi-GRU, a recurrent neural network unit that can utilize forward and backward time sequence information) is used to extract time sequence features from the convolution feature map. Since the battery voltage changes during charging have obvious time sequence, the forward hidden state can capture the time sequence information from the beginning of charging to the current time, and the backward hidden state can capture the time sequence information from the current time to the end of charging. The time sequence feature vector formed by the combination of the two can comprehensively reflect the time sequence law of voltage changes; the attention mechanism layer calculates the importance weight of each element in the time sequence feature vector, gives higher attention to key time sequence features, and gives greater weight to the time sequence features at the voltage mutation moment, realizes the strengthening of important information, and obtains a weighted feature vector; the fully connected layer is used to map the weighted feature vector to a preset output dimension, and through the calculation of multiple neurons, finally outputs the power battery voltage prediction value, which provides a basis for comparison for subsequent fault diagnosis.

[0084] S6, according to the power battery voltage prediction value and the power battery voltage, the power battery of the automobile is diagnosed for fault to obtain a fault diagnosis result;

[0085] Exemplarily, the safety monitoring terminal diagnoses the power battery of the automobile for fault according to the difference between the power battery voltage prediction value and the actual power battery voltage. First, the residual value between the two is calculated to obtain a residual sequence. The terminal determines a dynamic threshold range by calculating the mean and standard deviation of the residual sequence. When the residual value exceeds the dynamic threshold range, it is judged as a fault anomaly, and the abnormal time point and the abnormal type are recorded. Based on these fault anomaly results, further analysis is performed to obtain a fault diagnosis result, including fault type, fault occurrence frequency and fault duration. Through the fault diagnosis of the prediction model and the dynamic threshold, potential faults of the power battery during charging can be found in time and accurately, and the charging safety is improved.

[0086] S7, according to the fault diagnosis result, the charging safety risk of the automobile is comprehensively evaluated to obtain a safety risk evaluation result.

[0087] Specifically, the safety monitoring terminal performs comprehensive evaluation on the charging safety risk of the automobile according to the fault diagnosis result through multi-dimensional index quantification and weighted calculation, and obtains a safety risk evaluation result containing a safety risk evaluation score. Among them, the fault types, fault occurrence frequency, fault duration and other information in the fault diagnosis result have different degrees of influence on charging safety. For example, the risk of internal short circuit fault is much higher than that of voltage slight fluctuation. The higher the fault occurrence frequency and the longer the fault duration, the higher the safety risk. The safety monitoring terminal first converts each item of information in the fault diagnosis result into a quantitative index, and then sets a weight according to the degree of influence of each index on safety. The safety risk evaluation score is calculated by weighted summation. The higher the score, the higher the safety risk. Finally, the safety risk level is determined in combination with the interval in which the score is located, and a complete safety risk evaluation result is formed, providing a clear safety decision basis for the user or the charging management system.

[0088] In the new energy automobile charging safety monitoring method, the vehicle-pile dual-end data is acquired and standardized, effectively solving the problems of real vehicle data noise interference, uneven data quality, and monitoring limitations caused by differences between laboratory and actual environment; in the vehicle-pile data fusion and feature extraction link, the vehicle-pile data resources are deeply integrated, the safety indexes are fused, and the coupling effect of the two is fully considered, so as to make up for the defects of insufficient utilization of vehicle-pile collaborative data in the prior art; in combination with the voltage prediction model with structures such as convolution layer and bidirectional gate control recurrent unit and the fault diagnosis mechanism, the battery fault can be accurately identified, the subsequent safety risk evaluation provides quantitative basis for charging safety, and the accuracy and reliability of monitoring in actual charging scene are improved as a whole, meeting the actual charging safety monitoring demand.

[0089] In an optional embodiment, the automobile battery feature parameters and the charging pile feature parameters are input into a preset power battery voltage prediction model to obtain a power battery voltage prediction value, including the following steps:

[0090] S11, the automobile battery feature parameters and the charging pile feature parameters are spliced to obtain a joint feature vector.

[0091] Exemplarily, the safety monitoring terminal first performs data format verification on the automobile battery characteristic parameters and the charging pile characteristic parameters, ensures that the data dimensions and data types in the two types of parameters are consistent, and avoids splicing failure due to format differences. The automobile battery characteristic parameters include multiple dimensions of characteristic parameters such as voltage, current, and temperature, and the charging pile characteristic parameters include characteristic parameters such as output voltage, output current, and insulation resistance. The safety monitoring terminal corresponds the characteristic parameters in the two types of parameter sets one by one according to a preset parameter sorting rule. Specifically, the automobile battery voltage characteristic parameters and the charging pile output voltage characteristic parameters are arranged adjacently, and then the two types of parameters are spliced into a unified joint feature vector through a data splicing algorithm. The joint feature vector integrates the key feature information of the vehicle and the pile, can comprehensively reflect the cooperative operation state of the battery and the charging pile during the charging process, and provides complete input data for feature processing of each layer of the subsequent model.

[0092] S12, extracting local features from the joint feature vector through a convolution layer, and outputting a convolution feature map;

[0093] Exemplarily, when the safety monitoring terminal extracts local features from the joint feature vector through the convolution layer, a plurality of convolution kernels of different sizes are preset according to the data dimensions and feature types of the joint feature vector, for example, 3x1 and 5x1 convolution kernels. Each convolution kernel corresponds to extract a local feature mode, such as the change correlation feature of voltage and current in a short time, the cooperative change feature of temperature and insulation resistance, and the like. The convolution kernel performs sliding calculation on the joint feature vector, and maps the local features to feature values through matrix multiplication and an activation function. The feature values output by the plurality of convolution kernels are integrated to form a convolution feature map. The convolution feature map condenses the local key information in the joint feature vector and eliminates part of the redundant data.

[0094] S13, extracting time sequence features from the convolution feature map through a bidirectional gate control recurrent unit to obtain a time sequence feature vector.

[0095] Exemplarily, when the safety monitoring terminal extracts time sequence features from the convolution feature map through the bidirectional gate control recurrent unit, since the feature data in the convolution feature map is arranged in time sequence, the forward calculation process of the bidirectional gate control recurrent unit advances from the feature data at the charging start time, captures the forward change law in the time sequence, and the backward calculation process traces back from the feature data at the current charging time, captures the reverse correlation information in the time sequence. The forward hidden state and the backward hidden state respectively store the forward and reverse time sequence feature information, and the time sequence feature vector formed by splicing the two can completely reflect the change trend of the convolution feature map in the time dimension, and provide key features in the time dimension for processing of the subsequent attention mechanism layer.

[0096] S14, weighting the time sequence feature vector through an attention mechanism layer to obtain a weighted feature vector;

[0097] Exemplarily, the safety monitoring terminal weights the time sequence feature vector through the attention mechanism layer, identifies the feature elements in the time sequence feature vector that have an important influence on the prediction of the power battery voltage, and gives a higher weight to improve the prediction accuracy of the model. The safety monitoring terminal first calculates the attention score of each element in the time sequence feature vector. The attention score is obtained through a preset attention calculation function. The function comprehensively considers the feature value size of the element itself, the relevance with adjacent elements, and the position information in the entire time sequence. Specifically, the time sequence feature element corresponding to the voltage mutation moment will obtain a higher attention score. The terminal assigns weights to each element in the time sequence feature vector according to the attention score. The higher the score, the greater the weight of the element. Then each element is multiplied by its corresponding weight to obtain the weighted feature element. All weighted feature elements are integrated to form a weighted feature vector. Through the weighting process, the time sequence features that play a key role in voltage prediction can be strengthened, and the influence of irrelevant or secondary features can be weakened, so that the subsequent fully connected layer can focus on core features for calculation, improving the accuracy of the power battery voltage prediction value.

[0098] S15, input the weighted feature vector into the fully connected layer to obtain the power battery voltage prediction value.

[0099] Exemplarily, the safety monitoring terminal inputs the weighted feature vector into the fully connected layer. The fully connected layer is composed of multiple layers of neurons. Each layer of neurons is connected with all neurons of the previous layer to realize the deep feature conversion and integration of the weighted feature vector. Specifically, the first layer of neurons performs linear transformation and nonlinear activation processing on the weighted feature vector, maps the feature vector to a new feature space, and extracts more abstract feature information. The intermediate layer neurons continue to process the features output by the previous layer, gradually optimize the feature expression, and enhance the ability to capture the voltage change rule. The last layer of neurons maps the features processed by multiple layers to a single output dimension, and the output value is the power battery voltage prediction value. The fully connected layer continuously optimizes the parameters through the preset weight parameters and bias terms combined with the back propagation algorithm, to ensure that the output prediction value can accurately fit the actual voltage change trend of the power battery.

[0100] In an optional embodiment, the standardized automobile battery data and the standardized charging pile data are fused to obtain vehicle-pile fusion data, including the following steps:

[0101] S21, obtaining charging order data of an automobile.

[0102] Specifically, the safety monitoring terminal obtains the charging order data of the automobile by establishing a data communication connection with the charging operation management platform. The charging order data is structured data recording key information of a single charging service, including a vehicle identification number (VIN, an encoding uniquely identifying each automobile), an order creation time, a charging start time, and a charging end time. The vehicle identification number is a core identifier linking the automobile and the charging order, and through the VIN, the specific new energy automobile can be accurately located. The order creation time records the time point at which the user initiates the charging request, and is used to reflect the starting application time of the charging service. The charging start time is the time at which the charging pile actually starts to supply power to the automobile, marking the formal start of the charging process. The charging end time is the time at which the charging pile stops supplying power, marking the termination of the charging process. By obtaining the charging order data of the automobile, not only can the time range of a single charging service be determined, but also the charging order can be associated with the specific vehicle through the VIN, providing a key basis for subsequent matching of the vehicle-pile data and the order data, and ensuring the accuracy and relevance of data fusion.

[0103] S22, matching the standardized automobile battery data with the charging order data according to the vehicle identification number and the order creation time to obtain preliminary matching data;

[0104] Exemplarily, the safety monitoring terminal matches the standardized automobile battery data with the charging order data according to the vehicle identification number and the order creation time, establishes the correspondence between the automobile battery data and the specific charging order, and obtains preliminary matching data. Specifically, since there may be multiple automobiles charging at the same time in the same time period, relying only on time dimension matching is easy to cause data confusion, and the vehicle identification number is unique and can be used as a key identifier to distinguish different automobiles. The safety monitoring terminal first takes the vehicle identification number as an index to filter out the standardized battery data of the automobile corresponding to each charging order, and then combines the order creation time to further filter out the standardized automobile battery data within a preset time range before and after the order creation time, excluding battery data unrelated to the charging order. Through the double matching of the vehicle identification number and the order creation time, it is ensured that each piece of standardized automobile battery data can be accurately associated with the corresponding charging order.

[0105] S23, time window matching of the preliminary matching data according to the charging start time and the charging end time to obtain time window matching data.

[0106] Exemplarily, the safety monitoring terminal matches the preliminary matching data in a time window according to the charging start time and the charging end time, so as to accurately locate the battery data corresponding to the single charging process, and obtain time window matching data. The preliminary matching data may include preparation data before charging starts and redundant data after charging ends, which are not effective data in the charging process and need to be screened through the time window. Specifically, the safety monitoring terminal takes the charging start time in the charging order as the starting point of the time window and takes the charging end time as the ending point of the time window, so as to construct a time window completely consistent with the duration of the single charging process. Then, the safety monitoring terminal extracts the standardized automobile battery data with the time stamp in the time window from the preliminary matching data, which corresponds to the battery state change of the automobile in the current charging process, and eliminates the invalid data before and after charging, so as to ensure that the subsequent data fusion is based on only the effective data in the charging process and improve the data fusion quality.

[0107] S24, data alignment and splicing are performed on the time window matching data to obtain preliminary vehicle and pile fusion data;

[0108] Exemplarily, the safety monitoring terminal performs data alignment and splicing on the time window matching data, integrates the standardized automobile battery data in the time window and the corresponding standardized charging pile data according to the one-to-one correspondence of the time stamp, and obtains preliminary vehicle and pile fusion data. Specifically, since the collection frequencies of the automobile battery data and the charging pile data may be different, the battery data is collected every 1 second and the charging pile data is collected every 2 seconds, and direct splicing may cause data time misalignment. The safety monitoring terminal first uniformly processes the time stamps of the two types of data, uses an interpolation algorithm to complete the data with a lower collection frequency, so that the time stamps of the two types of data are completely synchronized, and then splices the standardized automobile battery parameters and the standardized charging pile parameters at the same time point according to the time stamp sequence, so as to form a structured data entry containing the time stamp, the battery parameter and the charging pile parameter. All data entries are arranged in time sequence to form the preliminary vehicle and pile fusion data, so as to accurately align the vehicle and pile data in the time dimension.

[0109] S25, denoising is performed on the preliminary vehicle and pile fusion data to obtain vehicle and pile fusion data.

[0110] Specifically, the safety monitoring terminal performs denoising processing on the preliminary vehicle-pile fusion data to eliminate noise data introduced in the data collection process due to sensor errors, communication interference and other factors, and obtains pure vehicle-pile fusion data. There may be abnormal data points in the preliminary vehicle-pile fusion data, such as voltage mutation caused by temporary sensor failure, zero current and other data that do not conform to the actual charging rules. These noise data will affect the accuracy of subsequent feature extraction and analysis. The safety monitoring terminal uses a preset denoising algorithm, such as a sliding average algorithm or an outlier detection algorithm, to analyze each point of the preliminary vehicle-pile fusion data: the sliding average algorithm calculates the average value of adjacent data points to smooth the data fluctuations and eliminate small noise; the outlier detection algorithm identifies and eliminates abnormal data points that are outside the normal range. Through denoising processing, it is ensured that the final vehicle-pile fusion data can truly and accurately reflect the actual operating status of the vehicle and the pile during charging.

[0111] In an optional embodiment, feature extraction is performed on the vehicle-pile fusion data to obtain vehicle battery feature parameters and charging pile feature parameters, including the following steps:

[0112] S31, the vehicle-pile fusion data is divided into a plurality of charging segment data by a sliding window;

[0113] Exemplarily, the safety monitoring terminal divides the vehicle-pile fusion data by a sliding window, the purpose of which is to divide the continuous time-series vehicle-pile fusion data into a plurality of charging segment data with time correlation. The vehicle-pile fusion data is a continuous data sequence arranged in chronological order, and it is difficult to capture the feature differences of different charging stages by directly extracting features from the whole data, for example, the battery and charging pile state change rules are different in the initial, middle and final stages of charging. The safety monitoring terminal presets a fixed length of sliding window and sliding step, the length of the sliding window is set according to the typical stage time length of the charging process, to ensure that the data in each window can reflect a complete local charging stage; the sliding step controls the interval of window movement to avoid excessive overlap or omission of data. By moving the sliding window on the vehicle-pile fusion data sequence, the continuous data is divided into a plurality of non-overlapping or slightly overlapping charging segment data, each segment data corresponds to a specific charging time period, which facilitates subsequent feature extraction for different charging stages.

[0114] S32, feature parameters are extracted from each charging segment data to obtain vehicle battery feature parameters and charging pile feature parameters.

[0115] Specifically, the safety monitoring terminal extracts feature parameters from each charging segment data, and mines key indicators from each local charging segment that can represent the operating state of the battery and the charging pile in this stage, to obtain the automobile battery feature parameters and the charging pile feature parameters. For the automobile battery feature parameters, the voltage mean is obtained by calculating the average value of all voltage data in the charging segment, which is used to reflect the overall level of the battery voltage in this segment; the current variance is obtained by calculating the average value of the square sum of the deviation of the current data from the mean, which is used to reflect the degree of current fluctuation, the larger the variance, the more unstable the current; the temperature slope is obtained by linear fitting the temperature change trend with time in this segment, which is used to reflect the temperature change rate, the positive slope indicates temperature rise, and the larger the absolute value of the slope, the faster the temperature change. For the charging pile feature parameters, the output voltage fluctuation is obtained by calculating the difference between the maximum and minimum values of the voltage, the larger the difference, the worse the voltage stability; the output current stability is obtained by calculating the coefficient of variation (ratio of standard deviation to mean) of the current data, the smaller the coefficient of variation, the more stable the current; the insulation resistance change rate is obtained by calculating the difference between the insulation resistance at the start time and the end time of the segment, and then dividing by the insulation resistance at the start time, which reflects the change trend of the insulation performance, when the change rate is negative and the absolute value is large, it indicates that the insulation performance decreases significantly. These feature parameters can accurately depict the operating state of the car and the pile in each charging segment.

[0116] In an optional embodiment, according to the power battery voltage prediction value and the power battery voltage, the power battery of the automobile is diagnosed for fault to obtain a fault diagnosis result, including the following steps:

[0117] S41, the residual value of the power battery voltage prediction value and the power battery voltage is calculated using the following formula to obtain a residual sequence:

[0118] ;

[0119] wherein, is the residual value of the th time step, is the power battery voltage of the th time step, is the power battery voltage prediction value of the th time step.

[0120] Specifically, the safety monitoring terminal calculates the residual value of the power battery voltage prediction value and the power battery voltage according to the given formula to form a residual sequence. The residual value is a core indicator for measuring the deviation degree of the prediction value from the actual value. In the calculation formula of the above residual value, is the residual value of the th time step, is the power battery voltage of the th time step, is the power battery voltage prediction value of the i-th time step. By calculating the difference between the two by time step, the continuous deviation information is converted into an ordered residual sequence, which can intuitively reflect the error distribution law of the prediction model. If the residual sequence presents random fluctuations and stable range, it means that the model prediction effect is good; if there is systematic deviation or mutation, it may be that the battery is abnormal or the model needs to be optimized.

[0121] S42, using the following formula, the residual mean and residual standard deviation of the residual sequence are obtained:

[0122] ;

[0123] wherein, is the residual mean, is the residual standard deviation, is the window size, is the i-th residual value in the window.

[0124] Specifically, the safety monitoring terminal calculates the residual mean and residual standard deviation of the residual sequence according to the formula. In the calculation formula of the residual mean and the residual standard deviation, is the residual mean, which is used to reflect the overall trend of the deviation. If the mean deviates from the zero value by a large margin, there may be systematic errors in the prediction model. is the residual standard deviation, which measures the dispersion degree of the residual value deviating from the mean value, reflects the fluctuation amplitude of the error, and the larger the standard deviation, the more dispersed the error distribution. is the window size, is the i-th residual value in the window. The window size needs to consider the continuity of the time sequence and the local feature capturing ability. By segmenting the residual sequence with a sliding window, the mean and standard deviation corresponding to each window are obtained, which dynamically reflects the error characteristics in different time periods and provides a quantitative basis for the determination of the subsequent dynamic threshold. S43, according to the residual mean and the residual standard deviation, using the following formula, the upper limit of the dynamic threshold and the lower limit of the dynamic threshold are obtained:

[0125]

[0126] ;

[0127] wherein, is the upper limit of the dynamic threshold, is the lower limit of the dynamic threshold, is the residual mean, is the residual standard deviation.

[0128] ​​​Specifically, the safety monitoring terminal calculates the upper and lower limits of the dynamic threshold using a formula based on the residual mean and residual standard deviation. The setting of the dynamic threshold follows statistical principles. The principle is that normal data is likely distributed within the range of the mean plus or minus... Data outside the range of one-times the standard deviation is considered outlier. In the above formulas for calculating the upper and lower limits of the dynamic threshold, The mean of the residuals, The standard deviation of the residuals. The upper limit of the dynamic threshold is Add 3 times , The lower limit of the dynamic threshold is Reduce by 3 times This setting method allows the threshold to be dynamically adjusted according to the mean and fluctuation characteristics of the residual sequence. Compared with a fixed threshold, it is more adaptable to changes in the battery state during charging. The residual distribution characteristics may be different in the early and late stages of charging. The dynamic threshold can accurately match the error characteristics of different stages and improve the accuracy of anomaly detection.

[0129] S44. Based on the upper limit and lower limit of the dynamic threshold, the dynamic threshold range is obtained.

[0130] For example, the security monitoring terminal defines the range between the upper and lower limits of the dynamic threshold as the dynamic threshold range. The dynamic threshold range serves as the benchmark for determining whether residual values ​​are normal and is updated in real time as the time window moves. The width of the dynamic threshold range is determined by the standard deviation of the residuals; the greater the error fluctuation, the wider the range; the smaller the error fluctuation, the narrower the range. By setting this range, the distribution interval of normal residuals is clearly defined, providing a clear criterion for subsequent anomaly detection. This ensures that only residual values ​​that significantly deviate from the normal pattern are identified as abnormal, reducing the probability of false positives.

[0131] S45, when the residual value exceeds the dynamic threshold range, an abnormal fault result is obtained.

[0132] For example, the safety monitoring terminal compares the residual value with the dynamic threshold range. When the residual value exceeds the upper limit of the dynamic threshold or falls below the lower limit, an anomaly is identified, resulting in a fault anomaly result. The anomaly time point is the specific time step in which the residual value exceeds the threshold range; the time of occurrence can be precisely located using a timestamp. The anomaly type is initially determined based on the direction and magnitude of the residual exceeding the threshold. For example, a residual value significantly higher than the upper limit of the dynamic threshold may correspond to a voltage surge anomaly, while a residual value significantly lower than the lower limit may correspond to a voltage drop anomaly. By monitoring the relationship between the residual and the threshold in real time, rapid identification of abnormal battery states is achieved, providing initial anomaly information for subsequent fault diagnosis.

[0133] S46. Based on the fault anomaly results, the fault diagnosis results are obtained.

[0134] The safety monitoring terminal integrates and analyzes the fault abnormality results to obtain fault diagnosis results. The fault diagnosis results include fault type, fault occurrence frequency, and fault duration. The terminal aggregates abnormal time points and abnormal types, and combines other characteristic parameters such as battery current and temperature to determine whether the abnormality is caused by battery failure and to exclude false abnormalities caused by data collection errors. Then, the fault type is determined, such as voltage out of control and single battery abnormality. The fault occurrence frequency, i.e., the number of abnormal events per unit time, is counted. The fault duration, i.e., the time interval from the occurrence of the abnormality to the recovery to normal, is recorded. Through comprehensive analysis of these information, complete fault diagnosis results including fault type, occurrence frequency, and duration are formed.

[0135] In an optional embodiment, based on the fault abnormality results, the fault diagnosis results are obtained, including the following steps:

[0136] S51, clustering analysis is performed on the abnormal time points in the fault abnormality results to obtain abnormal event clusters.

[0137] Specifically, the safety monitoring terminal performs clustering analysis on the abnormal time points in the fault abnormality results, groups abnormal points with time correlation into the same abnormal event cluster, and identifies complete fault events. A single abnormal time point may only be a transient disturbance, while continuous or intensive abnormal points often constitute a complete fault event. The safety monitoring terminal uses a clustering algorithm, such as Density-Based Spatial Clustering of Applications with Noise (DBSCAN, a clustering algorithm based on data point density), to group abnormal time points according to their time intervals: when the interval between two abnormal time points is less than a preset threshold, it is determined to be continuous abnormality of the same event; when the interval is greater than the threshold, it is divided into different events. The abnormal event cluster obtained by clustering can accurately reflect the complete time distribution of the fault event, avoiding misjudgment of multiple abnormal points of the same fault as multiple independent faults.

[0138] S52, feature extraction is performed on the abnormal event cluster to obtain fault frequency features and fault duration.

[0139] The fault frequency feature is used to represent the number of times of the fault occurring in a preset time period, and the fault duration is used to represent the time period of the abnormal event occurring. Specifically, the safety monitoring terminal extracts features of the abnormal event cluster to obtain the fault frequency feature and the fault duration. The fault frequency feature is obtained by counting the number of abnormal event clusters in a preset time period, and is used to reflect the intensive degree of the fault occurrence. The higher the frequency is, the worse the battery stability is. The fault duration is obtained by calculating the time difference between the first abnormal time point and the last abnormal time point in each abnormal event cluster, and reflects the continuous time length of the fault event. The longer the duration is, the more serious the fault impact is.

[0140] S53, obtaining a fault type classification result according to the fault frequency feature and the fault duration.

[0141] Exemplarily, the safety monitoring terminal obtains the fault type classification result according to the fault frequency feature and the fault duration, and in combination with a preset fault type classification rule. Different types of battery faults have different time characteristics. Specifically, the cell aging fault can be manifested as low frequency and short duration voltage abnormality, and the internal short circuit fault can be manifested as high frequency and long duration voltage drop and temperature rise coexisting abnormality. The safety monitoring terminal compares the extracted fault frequency and duration features with preset fault feature templates of various types, determines the most possible fault type through similarity matching, and the classification result clearly indicates the specific attributes of the fault, thereby providing a basis for accurately diagnosing the battery problem.

[0142] S54, obtaining a fault diagnosis result according to the fault type classification result.

[0143] Specifically, the safety monitoring terminal integrates the fault frequency feature, the fault duration and other information according to the fault type classification result to form a final fault diagnosis result. The fault diagnosis result not only contains the specific type of the fault, but also covers the time law (frequency) and the impact range (duration) of the fault occurrence, and is a comprehensive characterization of the battery fault state. Exemplarily, the diagnosis result can be a cell aging fault, which occurs twice in the past 1 hour, and each time lasts about 30 seconds. The structured result not only clearly indicates the nature of the fault, but also provides quantitative fault features, thereby providing detailed and accurate input data for subsequent safety risk evaluation, and ensuring the scientificity and pertinence of the risk evaluation.

[0144] In an optional embodiment, according to the fault diagnosis result, a comprehensive evaluation is performed on the charging safety risk of the automobile to obtain a safety risk evaluation result, including the following steps:

[0145] S61, obtaining a risk evaluation index according to the fault diagnosis result.

[0146] For example, the safety monitoring terminal extracts key information from fault diagnosis results and combines it with the characteristic parameters of the vehicle battery and the charging pile to determine risk assessment indicators. These indicators cover multi-dimensional state information of the battery and the charging pile. Vehicle battery characteristic parameters include voltage stability, current fluctuation, temperature change rate, state of charge, and capacity degradation rate, directly reflecting the battery's health status and safety boundaries. Charging pile characteristic parameters include output voltage stability, output current stability, temperature, insulation resistance, charging efficiency, and fault frequency, reflecting the impact of the charging pile's operating status on charging safety. Specifically, the terminal integrates these indicators to construct a charging safety risk assessment system containing target, project, and indicator layers. It then determines the weights of the safety risk assessment indicators, uses the analytic hierarchy process (AHP) and entropy weighting method to determine the combined weights of each indicator, and calculates the charging safety risk indicators for each vehicle based on the constructed risk assessment system. The indicators are then normalized to obtain a safety risk assessment score for each vehicle. The superiority of this method is verified by comparing the results with the AHP analysis, thus achieving charging safety risk assessment.

[0147] S62, use the following formula to calculate the overall weight of each risk assessment indicator:

[0148] ;

[0149] in, For the first The comprehensive weight of each risk assessment indicator, To determine the matrix element, the first element is represented as the first element. The risk assessment indicator is relative to the first The importance of each risk assessment indicator This represents the total number of risk assessment indicators. To determine the order of a matrix.

[0150] Specifically, the safety monitoring terminal uses a given formula to calculate the comprehensive weight of each risk assessment indicator. The weight reflects the importance of different indicators in the risk assessment. The formula for calculating the comprehensive weight is as follows: For the first The comprehensive weight of each risk assessment indicator, To determine the first element in the matrix The first indicator is relative to the first The importance of each indicator is scored using a judgment matrix constructed through the Analytic Hierarchy Process (AHP, a method for decomposing complex decision problems into hierarchical structures for analysis), quantifying the relative importance among the indicators. For example, the importance of fault type may be higher than that of fault frequency, reflected in... The values in the matrix are normalized, i.e., the numerator is the sum of the importance scores of the certain index and all other indexes, and the denominator is the sum of the importance scores of all indexes with respect to each other, to obtain initial comprehensive weights of the indexes, which provide a weight basis for subsequent risk calculation.

[0151] S63, the comprehensive weights of the risk evaluation indexes are corrected to obtain the corrected comprehensive weights of the risk evaluation indexes.

[0152] Specifically, the safety monitoring terminal corrects the comprehensive weights of the risk evaluation indexes by using the entropy weight method to obtain more objective and reasonable corrected weights. The entropy weight method is based on the information entropy theory and determines the weights by analyzing the dispersion degree of the index data. The greater the dispersion degree of the data (i.e., the more significant the change in the index value), the greater the amount of information it carries, and the higher the weight. Conversely, the weight is lower. By correcting through the entropy weight method, the subjective bias that may exist in the analytic hierarchy process can be effectively made up. Illustratively, when the actual data of the battery insulation resistance value index fluctuates greatly (indicating that it has a significant impact on safety), the entropy weight method will increase its weight. In the correction process, the terminal will first calculate the information entropy of each index, then calculate the entropy weight according to the information entropy, and finally combine the initial weight obtained by the analytic hierarchy process with the entropy weight to obtain the corrected comprehensive weight, so that the weight distribution reflects both expert experience and actual data characteristics.

[0153] S64, according to the corrected comprehensive weights of the risk evaluation indexes and the risk evaluation indexes, the following formula is used to obtain the safety risk evaluation score:

[0154] ;

[0155] wherein, is the safety risk evaluation score, is the corrected comprehensive weight of the i-th risk evaluation index, is the quantitative value of the i-th risk evaluation index, and is the total number of risk evaluation indexes. Specifically, the safety monitoring terminal calculates the safety risk evaluation score according to the corrected comprehensive weight and the quantitative value of each risk evaluation index

[0156] through the above formula of the safety risk evaluation score. ​​​is a numerical processing of the risk indicators, and the failure type is assigned according to the risk level, the failure frequency is assigned according to the occurrence interval, and the failure duration is assigned according to the time interval. The formula integrates the quantitative values of each indicator into a single score by weighted summation, and the greater the weight of the indicator, the greater the impact on the total score. The quantitative value of the high-risk failure type is multiplied by a higher weight, which significantly increases the risk score. The safety risk evaluation score is used to comprehensively reflect the overall level of charging safety risk and provide a quantitative basis for risk level classification.

[0157] S65, according to the safety risk evaluation score, the safety risk level is obtained; the safety risk level includes a first safety level, a second safety level and a third safety level;

[0158] Specifically, the safety monitoring terminal determines the corresponding safety risk level according to the interval in which the safety risk evaluation score is located, including a first safety level (low risk), a second safety level (medium risk), and a third safety level (high risk). The interval threshold of the level classification is set according to historical failure data and safety standards. For example, a score below a preset lower limit is classified as the first level, indicating that the charging process is safe and controllable; a score between the lower limit and the upper limit is classified as the second level, indicating that there is potential risk that needs to be closely monitored; and a score above the upper limit is classified as the third level, indicating that the risk is high and needs to be intervened immediately. Different levels correspond to different safety states, and the level classification converts the abstract risk score into an intuitive risk state identifier, making it easy to quickly understand the risk level.

[0159] S66, according to the safety risk level, the safety risk evaluation result is obtained; the safety risk evaluation result is used to indicate the safety risk of the power battery of the automobile during charging.

[0160] Specifically, the safety monitoring terminal forms a complete safety risk evaluation result according to the safety risk level and the specific circumstances of the risk evaluation indicators. The safety risk evaluation result not only includes the risk level, but also includes the key factors that lead to the level, and gives specific safety suggestions. The key factor can be the third safety level, mainly due to the high frequency of internal short circuit failure of the battery and the decrease of the insulation resistance of the charging pile. The safety suggestion can be to suggest stopping charging immediately, checking the internal state of the battery and the insulation performance of the charging pile. The safety risk evaluation result provides a clear decision basis for users, charging operators or vehicle control systems, effectively controls the charging safety risk, and ensures the safety and reliability of the charging process of new energy vehicles.

[0161] In the new energy vehicle charging safety monitoring method, the multi-dimensional data of the vehicle battery and the charging pile are acquired, the data difference influence is eliminated through standardization processing, and the vehicle-pile fusion data is fused to break the data island. Subsequently, the feature parameters are extracted and input into the voltage prediction model with structures such as convolution layer and bidirectional gate control recurrent unit, fault diagnosis is realized in combination with the actual voltage, and finally, the safety risk evaluation is completed according to the diagnosis result. The method effectively solves the problems of large difference between laboratory and actual charging environment, vehicle data interference by noise and the like, fully integrates vehicle-pile data resources and fuses safety indexes, considers the coupling effect of the two, avoids the limitations of traditional monitoring methods, improves the accuracy and applicability of charging safety monitoring, meets the safety monitoring needs in the actual charging scene, and provides reliable protection for new energy vehicle charging safety.

[0162] It should be understood that, although each step in the flowchart involved in each embodiment as described above is displayed in sequence according to the arrow, these steps are not necessarily executed in sequence according to the arrow. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other sequences. Moreover, at least part of the steps in the flowchart involved in each embodiment as described above can include multiple steps or stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of these steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least part of other steps or steps or stages in other steps.

[0163] Based on the same inventive concept, the embodiments of the present application also provide a new energy vehicle charging safety monitoring system for implementing the new energy vehicle charging safety monitoring method described above. The implementation scheme for solving the problem provided by the system is similar to the implementation scheme described in the above method, so the specific limitations in one or more new energy vehicle charging safety monitoring system embodiments provided below can refer to the limitations of the new energy vehicle charging safety monitoring method described above, which will not be repeated here.

[0164] In one exemplary embodiment, as shown in Figure 2 A new energy vehicle charging safety monitoring system 200 is provided, comprising:

[0165] The data acquisition module 201 is configured to acquire battery data and charging pile data of a vehicle. The battery data of the vehicle includes power battery voltage, current and temperature parameters. The charging pile data includes charging pile output voltage, charging pile current, charging pile temperature and insulation resistance.

[0166] The data standardization processing module 202 is configured to perform standardization processing on the battery data and the charging pile data of the automobile to obtain standardized battery data and standardized charging pile data.

[0167] The vehicle-pile data fusion module 203 is configured to fuse the standardized battery data and the standardized charging pile data to obtain vehicle-pile fusion data.

[0168] The feature parameter extraction module 204 is configured to perform feature extraction on the vehicle-pile fusion data to obtain automobile battery feature parameters and charging pile feature parameters.

[0169] The battery voltage prediction module 205 is configured to input the automobile battery feature parameters and the charging pile feature parameters into a preset power battery voltage prediction model to obtain a power battery voltage prediction value. The power battery voltage prediction model includes a convolution layer, a bidirectional gate control recurrent unit, an attention mechanism layer, and a full connection layer.

[0170] The fault diagnosis module 206 is configured to perform fault diagnosis on the power battery of the automobile according to the power battery voltage prediction value and the power battery voltage to obtain a fault diagnosis result.

[0171] The safety evaluation module 207 is configured to comprehensively evaluate the charging safety risk of the automobile according to the fault diagnosis result to obtain a safety risk evaluation result. The safety risk evaluation result includes a safety risk evaluation score.

[0172] Further, the battery voltage prediction module 205 is further configured to:

[0173] splicing the automobile battery feature parameters and the charging pile feature parameters to obtain a joint feature vector; the joint feature vector is input into the power battery voltage prediction model;

[0174] extracting local features of the joint feature vector through the convolution layer to output a convolution feature map;

[0175] extracting time sequence features of the convolution feature map through the bidirectional gate control recurrent unit to obtain a time sequence feature vector; the time sequence feature vector includes a forward hidden state and a backward hidden state;

[0176] weighting the time sequence feature vector through the attention mechanism layer to obtain a weighted feature vector;

[0177] inputting the weighted feature vector into the full connection layer to obtain the power battery voltage prediction value.

[0178] Further, the vehicle-pile data fusion module 203 is further configured to:

[0179] obtain charging order data of the automobile; the charging order data includes a vehicle identification code, an order creation time, a charging start time, and a charging end time;

[0180] According to the vehicle identification code and the order creation time, the standardized automobile battery data is matched with the charging order data to obtain preliminary matching data;

[0181] According to the charging start time and the charging end time, the preliminary matching data is matched in a time window to obtain time window matching data;

[0182] The time window matching data is aligned and spliced to obtain preliminary vehicle-pile fusion data;

[0183] The preliminary vehicle-pile fusion data is denoised to obtain vehicle-pile fusion data.

[0184] Further, the feature parameter extraction module 204 is further configured to:

[0185] The vehicle-pile fusion data is segmented in a sliding window to obtain a plurality of charging segment data;

[0186] Feature parameters are extracted from each charging segment data to obtain automobile battery feature parameters and charging pile feature parameters; the automobile battery feature parameters include voltage mean, current variance and temperature slope; the charging pile feature parameters include output voltage fluctuation, output current stability and insulation resistance change rate.

[0187] Further, the fault diagnosis module 206 is further configured to:

[0188] The following formula is used to calculate the power battery voltage prediction value and the residual value of the power battery voltage to obtain a residual sequence:

[0189] ;

[0190] wherein, is the residual value of the i th time step, is the power battery voltage of the i th time step, is the power battery voltage prediction value of the i th time step; The following formula is used to obtain the residual mean and the residual standard deviation of the residual sequence:

[0191] ;

[0192] ;

[0193] wherein, is the residual mean, is the residual standard deviation, is the window size, is the i th residual value in the window;

[0194] ​​​According to the residual mean and the residual standard deviation, the upper limit of the dynamic threshold and the lower limit of the dynamic threshold are obtained using the following formula:

[0195] ;

[0196] wherein, the upper limit of the dynamic threshold, the lower limit of the dynamic threshold, the residual mean, the residual standard deviation;

[0197] According to the upper limit of the dynamic threshold and the lower limit of the dynamic threshold, the dynamic threshold range is obtained;

[0198] When the residual value exceeds the dynamic threshold range, a fault anomaly result is obtained; the fault anomaly result includes an abnormal time point and an abnormal type;

[0199] Based on the fault anomaly result, a fault diagnosis result is obtained; the fault diagnosis result includes a fault type, a fault occurrence frequency and a fault duration time.

[0200] Further, the fault diagnosis module 206 is further used for:

[0201] The abnormal time points in the fault anomaly result are subjected to cluster analysis to obtain an abnormal event cluster;

[0202] The abnormal event cluster is subjected to feature extraction to obtain a fault frequency feature and a fault duration time; the fault frequency feature is used to represent the number of times of occurrence of the fault within a preset time period; and the fault duration time is used to represent a time period of occurrence of the abnormal event;

[0203] According to the fault frequency feature and the fault duration time, a fault type classification result is obtained; the fault type classification result is used to indicate the type of the fault;

[0204] According to the fault type classification result, the fault diagnosis result is obtained.

[0205] Further, the safety evaluation module 207 is further used for:

[0206] According to the fault diagnosis result, a risk evaluation index is obtained; the risk evaluation index includes an automobile battery characteristic parameter and a charging pile characteristic parameter;

[0207] The comprehensive weight of each risk evaluation index is calculated using the following formula:

[0208] ;

[0209] wherein, the comprehensive weight of the i-th risk evaluation index, ​The judgment matrix element represents the importance of the jth risk evaluation index relative to the ith risk evaluation index. The judgment matrix element represents the importance of the jth risk evaluation index relative to the ith risk evaluation index. The judgment matrix element represents the importance of the jth risk evaluation index relative to the ith risk evaluation index. The judgment matrix element represents the importance of the jth risk evaluation index relative to the ith risk evaluation index. The judgment matrix element represents the importance of the jth risk evaluation index relative to the ith risk evaluation index. The judgment matrix element represents the importance of the jth risk evaluation index relative to the ith risk evaluation index.

[0210] The comprehensive weight of each risk evaluation index is modified to obtain the modified comprehensive weight of each risk evaluation index. The comprehensive weight of each risk evaluation index is modified to obtain the modified comprehensive weight of each risk evaluation index.

[0211] According to the modified comprehensive weight of each risk evaluation index and each risk evaluation index, the safety risk evaluation score is obtained by using the following formula:

[0212]

[0213] The safety risk evaluation score is obtained according to the safety risk evaluation score. The safety risk evaluation score is obtained according to the safety risk evaluation score. The safety risk evaluation score is obtained according to the safety risk evaluation score. The safety risk evaluation score is obtained according to the safety risk evaluation score. The safety risk evaluation score is obtained according to the safety risk evaluation score. The safety risk evaluation score is obtained according to the safety risk evaluation score. The safety risk evaluation score is obtained according to the safety risk evaluation score. The safety risk evaluation score is obtained according to the safety risk evaluation score.

[0214] The safety risk evaluation score is obtained according to the safety risk evaluation score. The safety risk evaluation score is obtained according to the safety risk evaluation score.

[0215] The safety risk evaluation score is obtained according to the safety risk evaluation score.

[0216] In one embodiment, a computer device 300 is provided, comprising: Figure 3 At least one processor 301 and at least one memory 302 connected in communication with the processor 301; the memory stores application program code executable by the at least one processor, and the application program code is executed by the at least one processor to enable the at least one processor to perform the steps of the new energy vehicle charging safety monitoring method as described above.

[0217] The computer device can further comprise a sensor 303.

[0218] The processor 301, the memory 302 and the sensor 303 can be connected through a bus 304 or other means, and in the figure, they are connected through the bus 304,

[0219] only one thick line is used to represent only one bus or one type of bus. Figure 3 only one thick line is used to represent only one bus or one type of bus. only one thick line is used to represent only one bus or one type of bus.

[0220] In one embodiment, a computer readable storage medium is provided, having stored thereon a computer program, which when executed by a processor implements the steps of any of the above method embodiments.

[0221] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts are described in the part of the method embodiments. The above described device embodiments are only illustrative, wherein the components described as separate components can or can not be physically separated, the components displayed as units can or can not be physical units, i.e. can be located in one place or can be distributed to multiple network units. Part or all of the modules can be selected to achieve the purpose of the present disclosure according to actual needs. Those skilled in the art can understand and implement it without creative labor.

[0222] The above described embodiments only express several implementation manners of the present application, the description is more specific and detailed, but it should not be understood as a limitation to the patent scope of the application. It should be pointed out that, for those skilled in the art, without departing from the concept of the present application, several modifications and improvements can be made, which are all within the protection scope of the present application.

Claims

1. A new energy vehicle charging safety monitoring method, characterized in that, The method comprises: acquiring battery data of the automobile and charging pile data; the battery data of the automobile comprises power battery voltage, current and temperature parameters; the charging pile data comprises charging pile output voltage, charging pile current, charging pile temperature and insulation resistance; standardizing the battery data of the automobile and the charging pile data to obtain standardized battery data of the automobile and standardized charging pile data; fusing the standardized battery data of the automobile and the standardized charging pile data to obtain vehicle-pile fusion data; extracting features from the vehicle-pile fusion data to obtain automobile battery feature parameters and charging pile feature parameters; inputting the automobile battery feature parameters and the charging pile feature parameters into a preset power battery voltage prediction model to obtain a power battery voltage prediction value; the power battery voltage prediction model comprises a convolution layer, a bidirectional gate control recurrent unit, an attention mechanism layer and a fully connected layer; performing fault diagnosis on the power battery of the automobile according to the power battery voltage prediction value and the power battery voltage to obtain a fault diagnosis result; comprehensively evaluating the charging safety risk of the automobile according to the fault diagnosis result to obtain a safety risk evaluation result; the safety risk evaluation result comprises a safety risk evaluation score; wherein, the fault diagnosis on the power battery of the automobile according to the power battery voltage prediction value and the power battery voltage to obtain a fault diagnosis result comprises: calculating the residual value of the power battery voltage prediction value and the power battery voltage using the following formula to obtain a residual sequence: ； wherein, is the residual value for the time step, is the power battery voltage for the time step, is the power battery voltage prediction value for the time step; using the following formula to obtain the residual mean and residual standard deviation of the residual sequence: ； ； in, The mean of the residuals, The standard deviation of the residual is... For window size, For the first in the window One residual value; using the following formula to obtain the upper limit of the dynamic threshold and the lower limit of the dynamic threshold according to the residual mean and the residual standard deviation: ； ； wherein, is an upper limit of the dynamic threshold, is a lower limit of the dynamic threshold, is the residual mean, is the residual standard deviation; obtaining a dynamic threshold range according to the upper limit of the dynamic threshold and the lower limit of the dynamic threshold; obtaining a fault anomaly result when the residual value exceeds the dynamic threshold range; the fault anomaly result comprises an abnormal time point and an abnormal type; obtaining a fault diagnosis result based on the fault anomaly result; the fault diagnosis result comprises a fault type, a fault occurrence frequency and a fault duration; wherein, the comprehensive evaluation of the charging safety risk of the automobile according to the fault diagnosis result to obtain a safety risk evaluation result comprises: obtaining risk evaluation indexes according to the fault diagnosis result; the risk evaluation indexes comprise the automobile battery feature parameters and the charging pile feature parameters; using the following formula to calculate the comprehensive weight of each risk evaluation index: ； in, For the first The comprehensive weight of each risk assessment indicator, To determine the matrix element, the first element is represented as the first element. The risk assessment indicator is relative to the first The importance of each risk assessment indicator This represents the total number of risk assessment indicators. To determine the order of a matrix; correcting the comprehensive weight of each risk evaluation index to obtain the corrected comprehensive weight of each risk evaluation index; using the following formula to obtain a safety risk evaluation score according to the corrected comprehensive weight of each risk evaluation index and each risk evaluation index: ； wherein, is the security risk assessment score, is the modified comprehensive weight of the risk assessment index, is the quantitative value of the risk assessment index, is the total number of risk assessment indexes. obtaining a safety risk level according to the safety risk evaluation score; the safety risk level comprises a first safety level, a second safety level and a third safety level; According to the safety risk level, a safety risk evaluation result is obtained; the safety risk evaluation result is used to indicate the safety risk of the power battery of the automobile during charging.

2. The method of claim 1, wherein, The method comprises the following steps: The automobile battery feature parameters and the charging pile feature parameters are spliced to obtain a joint feature vector; the joint feature vector is used as an input of the power battery voltage prediction model; Local features of the joint feature vector are extracted through the convolution layer, and a convolution feature map is output; Time sequence features of the convolution feature map are extracted through the bidirectional gate control cycle unit to obtain a time sequence feature vector; the time sequence feature vector comprises a forward hidden state and a backward hidden state; The time sequence feature vector is weighted through the attention mechanism layer to obtain a weighted feature vector; The weighted feature vector is input into the full connection layer to obtain a power battery voltage prediction value.

3. The method of claim 1, wherein, The method comprises the following steps: Charging order data of the automobile is obtained; the charging order data comprises a vehicle identification code, an order creation time, a charging start time, and a charging end time; The standardized automobile battery data and the charging order data are matched according to the vehicle identification code and the order creation time to obtain preliminary matching data; The preliminary matching data are matched in a time window according to the charging start time and the charging end time to obtain time window matching data; The time window matching data are aligned and spliced to obtain preliminary vehicle-pile fusion data; The preliminary vehicle-pile fusion data are denoised to obtain the vehicle-pile fusion data.

4. The method of claim 1, wherein, The method comprises the following steps: The vehicle-pile fusion data are divided into multiple charging segment data through a sliding window; Feature parameters of each charging segment data are extracted to obtain automobile battery feature parameters and charging pile feature parameters; the automobile battery feature parameters comprise a voltage mean value, a current variance, and a temperature slope; the charging pile feature parameters comprise an output voltage fluctuation, an output current stability, and an insulation resistance change rate.

5. The method of claim 1, wherein, The method comprises the following steps: The abnormal time points in the fault anomaly result are clustered to obtain an abnormal event cluster; Feature extraction is performed on the abnormal event cluster to obtain a fault frequency feature and a fault duration; the fault frequency feature is used to represent the number of times of fault occurrence in a preset time period; the fault duration is used to represent a time period of abnormal event occurrence; A fault type classification result is obtained according to the fault frequency feature and the fault duration; the fault type classification result is used to indicate the type of fault; The fault diagnosis result is obtained according to the fault type classification result.

6. A new energy vehicle charging safety monitoring system for implementing the method of any one of claims 1 to 5, characterized in that, The system comprises: The data acquisition module is configured to acquire battery data of the automobile and charging pile data; the battery data of the automobile includes power battery voltage, current and temperature parameters; the charging pile data includes charging pile output voltage, charging pile current, charging pile temperature and insulation resistance; The data standardization processing module is configured to perform standardization processing on the battery data of the automobile and the charging pile data to obtain standardized battery data of the automobile and standardized charging pile data; The vehicle-pile data fusion module is configured to fuse the standardized battery data of the automobile and the standardized charging pile data to obtain vehicle-pile fusion data; The feature parameter extraction module is configured to perform feature extraction on the vehicle-pile fusion data to obtain automobile battery feature parameters and charging pile feature parameters; The battery voltage prediction module is configured to input the automobile battery feature parameters and the charging pile feature parameters into a preset power battery voltage prediction model to obtain a power battery voltage prediction value; the power battery voltage prediction model includes a convolution layer, a bidirectional gate control recurrent unit, an attention mechanism layer and a full connection layer; The fault diagnosis module is configured to perform fault diagnosis on the power battery of the automobile according to the power battery voltage prediction value and the power battery voltage to obtain a fault diagnosis result; The safety evaluation module is configured to comprehensively evaluate the charging safety risk of the automobile according to the fault diagnosis result to obtain a safety risk evaluation result; the safety risk evaluation result includes a safety risk evaluation score. 7.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-6 when the computer program is executed by the processor. The processor executes the computer program to realize the method in any one of claims 1 to 5.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the method in any one of claims 1 to 5.

Images