Abnormality detection method for power battery based on important sample screening under big data framework

Abstract

The application discloses a power battery abnormality detection method based on important sample screening under a big data framework, which comprises the following steps: 1, preprocessing vehicle operation history data; 2, obtaining a training set and a test set according to the preprocessed vehicle operation data; determining a neural network model structure and inputting the training set into the neural network model for pretraining; 3, calculating the importance score of each sample in the training set; 4, determining important samples according to the importance score and the sample screening proportion; 5, inputting the important samples into the neural network model for training; 6, inputting real-time vehicle data into the prediction model; calculating the residual error of the test data and the model output, if the residual error is greater than a set abnormality threshold, it is determined that the vehicle is abnormal at the current time, otherwise, it is normal. The application reduces the influence of redundant samples on the model efficiency, and provides support for the deployment of deep learning algorithms in large-scale vehicle safety monitoring.

Description

Technical Field

[0001] The present invention belongs to the field of power battery safety technology and relates to a battery anomaly detection method, specifically a power battery anomaly detection method based on important sample screening under a big data framework. Background Art

[0002] With increasing public interest in green energy, new energy vehicles are becoming a mainstream trend in the automotive industry. In the application of onboard power battery systems, abnormal battery parameter performance often indicates a decrease in battery safety. Therefore, early and accurate identification of battery anomalies is crucial for ensuring safe vehicle operation and providing early warning of battery failures. With the rapid growth of electric vehicles in recent years, achieving large-scale vehicle safety monitoring has become a pressing issue. With the continuous advancement of AI technology, deep learning algorithms have been widely applied in academia for electric vehicle safety testing. However, in the real-world vehicle big data environment, the deployment of deep learning algorithms faces practical bottlenecks due to factors such as training time and computational resources resulting from the massive amount of training samples. For large-scale real-time vehicle monitoring, the less computational resources a monitoring method consumes, the more vehicles can be monitored. Therefore, accurate monitoring and efficient computation should be given equal importance. Summary of the Invention

[0003] This invention aims to provide a power battery anomaly detection method based on important sample screening to address the challenges of deploying deep learning algorithms in large-scale vehicle monitoring due to resource and time constraints. This method evaluates sample importance and selects important samples for model training, replacing all samples. This reduces the impact of redundant samples on model performance and provides support for the deployment of deep learning algorithms in large-scale vehicle safety monitoring.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] A power battery anomaly detection method based on important sample screening in a big data framework specifically includes the following steps:

[0006] Step 1: preprocess the vehicle operation history data to obtain preprocessed vehicle operation data, and select model input item data from the preprocessed data;

[0007] Step 2: Divide the model input data obtained in step 1 into a training set and a test set; determine the neural network model structure and input the training set into the neural network model for pre-training, determine the number of pre-training iterations and extract the network parameter w at the current number of iterations;

[0008] Step 3: Input the training set obtained in step 2 into the neural network model with network parameter w in sequence, and calculate the importance score of each sample, including the following sub-steps;

[0009] In step 31, the importance score is composed of the gradient change score, the global stability error score, and the local sensitivity error score of each input sample;

[0010] Step 32: Use the Z-score method to standardize GCS, GSS, and LSS respectively, and the sum of the standardized GCS, GSS, and LSS is used as the sample importance score;

[0011] Step 4: Sort the training set samples according to the importance score of each sample obtained in step 3, and determine the appropriate sample screening ratio to form important samples, including the following sub-steps;

[0012] Step 41, determining the sample screening ratio, which is the proportion of important samples to the training set samples;

[0013] Step 42: sort the training set samples from largest to smallest according to the sample importance scores obtained in step 3. Based on the sample screening ratio k determined in step 41, select the top k% samples with the largest importance scores from the sorting as important samples.

[0014] Step 5: Input the important samples obtained in step 4 into the neural network model obtained in step 2 for training to obtain a trained neural network model, i.e., a prediction model;

[0015] Step 6: Input the real-time vehicle data as test data into the prediction model to obtain the model output; calculate the residual between the test data and the model output. If it is greater than the set abnormality threshold, it is determined that the vehicle is abnormal at the current moment, otherwise it is normal.

[0016] Furthermore, in step 1, the vehicle operation history data includes mileage, vehicle speed, SOC, battery probe temperature and battery cell voltage; and the preprocessing includes data cleaning and feature importance analysis.

[0017] Furthermore, the feature importance analysis specifically uses the Pearson correlation coefficient to evaluate the correlation between data.

[0018] Furthermore, step 2 includes the following sub-steps:

[0019] Step 21: divide the model input data obtained in step 1 into a training set and a test set; randomly set initial hyperparameters and define a model loss function; the model loss function Loss is defined as:

[0020]

[0021] Where n is the number of samples tested, and U i are the predicted value and the true value respectively;

[0022] Step 22: Select multiple possible hyperparameter combinations from the initial hyperparameters, use a grid search method to test the model performance of each hyperparameter combination one by one, select the hyperparameter combination with the best performance for use in the neural network model, and determine the model structure;

[0023] In step 23, the training set is input into the neural network model obtained in step 22 for pre-training to extract the network parameters required for sample importance evaluation; the following principles are used for selection:

[0024] 1) As the number of pre-training iterations increases, when the loss function Loss is less than 2.5*10 -4 Stop pre-training when , and save the network parameters w at the current number of iterations;

[0025] 2) If the conditions in 1) are not met in the first 10 iterations of pre-training, stop pre-training and save the network parameters w corresponding to the iteration number with the minimum training set loss in the first 10 iterations.

[0026] Furthermore, in step 31, the gradient change fraction is as follows:

[0027]

[0028] Among them, f w (X) is the output predicted by the neural network model with the network parameter w at the current iteration of the input sample X. is the gradient of the loss function with respect to the network parameter w;

[0029] The global stability error fraction and the local sensitivity error fraction are as follows:

[0030] GSS=||Yf w (X)||2

[0031] LSS=||Yf w (X)||1

[0032] Furthermore, in step 41 , if the average standard deviation of the voltage of each single cell in the training set data is greater than 0.17, the optimal sample screening ratio is determined to be 40%; if it is less than 0.17, the optimal sample screening ratio is determined to be 30%.

[0033] Furthermore, step 6 includes the following operations: inputting the test set data obtained in step 21 into the prediction model obtained in step 5, calculating the residual between the model prediction value and the true value; using the box plot method to set the abnormal threshold, the calculation formula for setting the abnormal threshold of the box plot is as follows:

[0034] Thr = Q3 + 1.5 * IQR

[0035] Where Q3 is the third quartile of the residual of the test set data, IQR is the interquartile range, and Thre is the abnormal threshold.

[0036] Compared with the existing technology, the present invention focuses on the method for identifying abnormalities in electric vehicle power batteries. Based on the big data of electric vehicle cloud monitoring, it proposes a power battery abnormality detection method based on important sample screening to address the problem that deep learning algorithms cannot be deployed on a large scale due to resource and time constraints. This method achieves abnormality detection by predicting normal battery parameters. It is not limited to the type of battery parameters and can be applied to parameters such as battery voltage and temperature. It uses important samples instead of full samples for model training, which can achieve similar model performance while greatly improving the timeliness of the model. At the same time, this method realizes sample importance evaluation at the sample level and does not involve specific network models. Therefore, it is also applicable to other neural network models optimized by gradient descent method, and has extremely strong generalization ability. It provides support for large-scale vehicle safety monitoring and fault prediction based on cloud data. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is a flow chart of the method of the present invention;

[0038] Figure 2 is the voltage prediction effect of the full sample training model;

[0039] Figure 3 It is the voltage prediction effect of the model trained on the top 30% of samples with the highest sample importance;

[0040] Figure 4 It is the voltage prediction effect of the model trained on the first 30% of samples with the lowest sample importance. DETAILED DESCRIPTION

[0041] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0042] like Figure 1 As shown, the present invention provides a power battery anomaly detection method based on important sample screening under a big data framework, which specifically includes the following steps:

[0043] Step 1: Preprocess the historical vehicle operation data to obtain preprocessed vehicle operation data, and select model input data from the preprocessed data. This step is used to ensure that the data format can meet the needs of the model.

[0044] In this embodiment, the historical data uploaded by electric vehicles, in accordance with the national standard GB / T 32960 "Technical Specification for Electric Vehicle Remote Service and Management Systems," includes multiple key data items, including mileage, speed, SOC, battery probe temperature, and battery cell voltage. These data items will serve as candidate model input items and need to undergo preprocessing such as data cleaning and feature importance analysis before they can be used as model input. The specific operations are as follows:

[0045] 1) For vehicle operation history data, if a data frame contains a null value, linear interpolation is performed on the two adjacent frames of data to fill the data frame;

[0046] 2) Considering that different data items have different effective ranges and distribution characteristics, this embodiment uses the Z-score method to normalize the data to ensure that each feature has a similar scale and range. The formula of the Z-score method is as follows:

[0047]

[0048] In the formula, x is the data to be standardized, μ is the mean value of all data under the data item corresponding to x, σ is the standard deviation of all data under the data item corresponding to x, and Z is the standardized data.

[0049] 3) In order to obtain better model performance, it is necessary to screen important features through feature importance analysis. In this embodiment, the Pearson correlation coefficient is used to evaluate the correlation between data. A stronger correlation represents a greater contribution to model performance. Therefore, it is necessary to screen data with strong correlation as model input. In the present invention, the total battery voltage, SOC, and battery cell voltage data items after Z-score standardization are determined as model input items. In addition, considering that the present invention performs battery anomaly detection by predicting battery cell voltage, the battery cell voltage is used as the target feature item.

[0050] The calculation formula of Pearson correlation coefficient is:

[0051]

[0052] Where n is the number of samples, x is i and y i are the i-th value of the model candidate input item and target feature item data, respectively. and are the average values ​​of the model candidate input items and target feature item data, respectively, and coef is the Pearson correlation coefficient.

[0053] Step 2: Use the model input data obtained in step 1 to obtain a training set and a test set; determine the neural network model structure and input the training set into the neural network model for pre-training, determine the number of pre-training iterations and extract the network parameters w under the current number of iterations.

[0054] Specifically, step 2 includes the following sub-steps:

[0055] Step 21: Divide the model input data obtained in step 1 into a training set and a test set. Randomly set the initial hyperparameters and define the model loss function. Specifically, the model loss function Loss is defined as:

[0056]

[0057] Where n is the number of samples tested, and U i are the predicted value and the true value respectively;

[0058] Step 22: To ensure that the model with the best performance is obtained, the model's hyperparameters need to be optimized. In the present invention, multiple possible hyperparameter combinations are selected from the initial hyperparameters. The model performance of each hyperparameter combination is tested one by one using a grid search method. The best performing set of hyperparameters is selected for use in a neural network model (such as LSTM) to determine the model structure.

[0059] Step 23: input the training set into the neural network model obtained in step 22 for pre-training to extract the network parameters required for sample importance evaluation;

[0060] The network parameters corresponding to a certain number of pre-training iterations affect the effect of importance evaluation. Therefore, how to select an appropriate number of pre-training iterations is particularly important. In order to select the relatively optimal number of pre-training iterations in the shortest possible time, this embodiment adopts the following principles:

[0061] 1) As the number of pre-training iterations increases, when the loss function Loss is less than 2.5*10 -4 Stop pre-training when , and save the network parameters w at the current number of iterations.

[0062] 2) If the conditions in 1) are not met in the first 10 iterations of pre-training, stop pre-training and save the network parameters w corresponding to the iteration number with the minimum training set loss in the first 10 iterations.

[0063] Step 3: Input the training set obtained in step 2 into the neural network model with network parameter w in sequence, and calculate the importance score of each sample, including the following sub-steps;

[0064] Step 31, the sample importance is evaluated from two angles of network parameters and model output. The importance score is composed of gradient change score (GCS), global stability error score (GSS) and local sensitivity error score (LSS) of each input sample.

[0065] Specifically, the gradient descent method is a commonly used neural network model optimization method at present, and the gradient update formula thereof is as follows:

[0066]

[0067] In the formula, w t and w t-1 are network parameters in the current iteration and the last iteration respectively, η is a learning rate, l(·) is a loss function Loss, X and Y are input samples and model output, is the output of the neural network prediction of the input sample X under the network parameters w t-1 in the last iteration, is the gradient of the loss function to the network parameters w t-1 ;

[0068] From the angle of network parameters, the sample that has greater influence on model performance will cause greater gradient change after entering the neural network, therefore, the gradient change score (GCS) is used as an index for evaluating sample importance, as follows:

[0069]

[0070] In the formula, f w (X) is the output of the neural network model prediction of the input sample X under the network parameters w in the current iteration,

[0071] From the angle of model output, the greater the residual error between the model prediction value and the true value, the more information that is not considered by the model, therefore, such sample has greater influence on model performance. In order to represent the residual error between the model prediction value and the true value from multiple angles, the global stability error score (GSS) and the local sensitivity error score (LSS) are used as indexes for evaluating sample importance, as follows:

[0072] GSS=||Y-f w (X)||2

[0073] LSS=||Y-f w (X)||1

[0074] Step 32: Since the fluctuation range and distribution characteristics of the importance scores of different samples are different, the present invention adopts the Z-score method to standardize GCS, GSS and LSS respectively, and the sum of the standardized GCS, GSS and LSS is used as the sample importance score.

[0075] Step 4: Sort the training set samples according to the importance score of each sample obtained in step 3, and determine the appropriate sample screening ratio to form important samples, including the following sub-steps;

[0076] Step 41, determine the sample screening ratio, which is the proportion of important samples to the training set samples. For different training set samples, an increase in the amount of data does not mean an increase in the amount of information contained in the data. Generally, the discreteness of a data set can be used to characterize the variability and volatility between data points. For data sets with higher discreteness, the model usually requires more data to accurately represent the data characteristics. Therefore, the present invention uses the standard deviation of the training set data to quantify the amount of information contained in the data set to determine the optimal sample screening ratio. Accordingly, if the average standard deviation of the voltage of each single cell in the training set data is greater than 0.17, the optimal sample screening ratio is determined to be 40%; if it is less than 0.17, the optimal sample screening ratio is determined to be 30%.

[0077] In step 42, the training set samples are sorted from large to small according to the sample importance scores obtained in step 3, and according to the sample screening ratio k determined in step 41, the top k% samples with the largest importance scores are selected from the sorting as important samples.

[0078] In step 5, the important samples obtained in step 4 are input into the neural network model obtained in step 22 for training to obtain a trained neural network model, i.e., a prediction model.

[0079] Step 6: Input the real-time vehicle data as test data into the prediction model to obtain the model output; calculate the residual between the test data and the model output. If it is greater than the set abnormality threshold, it is determined that the vehicle is abnormal at the current moment, otherwise it is normal.

[0080] Step 6 includes the following operations: Since the proportion of vehicle operation data in normal conditions is much higher than the proportion in abnormal conditions, the acquired vehicle data is generally biased towards normal conditions. Therefore, the prediction results of the model trained on the vehicle data can be considered normal values. When the prediction result differs significantly from the actual value (i.e., greater than the set threshold), this can be considered an anomaly. The real-time vehicle data is input into the prediction model obtained in step 5, and the residual between the model prediction value and the actual value is calculated; the anomaly threshold is set using the box plot method.

[0081] Specifically, the calculation formula for setting the abnormal threshold of the box plot is as follows:

[0082] Thr = Q3 + 1.5 * IQR

[0083] Where Q3 is the third quartile of the residual of the test set data, IQR is the interquartile range, and Thre is the abnormal threshold.

[0084] In summary, the present invention proposes a power battery anomaly detection method based on important sample screening. It evaluates the importance of training samples from two perspectives: the network parameters of the neural network model and the model output. It uses important samples instead of all samples for training, thereby shortening the model training time while ensuring accuracy, and providing support for the deployment of neural network models in large-scale vehicle safety monitoring.

[0085] The anomaly detection effect of the method of the present invention depends on the performance of the prediction model. Therefore, taking voltage prediction as an example, the actual operation data of a real vehicle in 2020 is verified using a model trained with all samples, a model trained with the top 30% of samples with the highest sample importance, and a model trained with the top 30% of samples with the lowest sample importance. The verification effect is as follows: Figures 2-4 As shown. Figures 2-4 It can be seen that the effect obtained by training the model with the top 30% of samples with the highest sample importance is almost the same as that of the full sample training model, which indicates that the important samples extracted by the method of the present invention can achieve the ability to cover the information volume of the entire sample with only 30% of the data volume; by comparing the training model with the top 30% of samples with the highest sample importance and the training model with the top 30% of samples with the lowest sample importance, it can be found that the performance gap between the two models is large, which also verifies the accuracy of the extraction of important samples in the method of the present invention, thereby proving the feasibility and effectiveness of the method of the present invention.

Claims

1. A power battery anomaly detection method based on important sample screening under a big data framework, characterized in that: The specific steps include: Step 1: preprocess the vehicle operation history data to obtain preprocessed vehicle operation data, and select model input item data from the preprocessed data; The vehicle operation history data includes mileage, vehicle speed, SOC, battery probe temperature and battery cell voltage; the preprocessing includes data cleaning and feature importance analysis; Step 2: Divide the model input data obtained in step 1 into a training set and a test set; Determine the neural network model structure and input the training set into the neural network model for pre-training, determine the number of pre-training iterations and extract the network parameters under the current number of iterations ; Step 3: Input the training set obtained in step 2 into the network parameters In the neural network model, calculating the importance score of each sample includes the following sub-steps: Step 31, the importance score is composed of the gradient change score, global stability error score and local sensitivity error score of each input sample; The gradient change fraction is as follows: in, For input samples The network parameters at the current iteration The output predicted by the neural network model, The loss function is the network parameters gradient; The global stability error fraction and the local sensitivity error fraction are as follows: Step 32: Use the Z-score method to standardize GCS, GSS, and LSS respectively, and the sum of the standardized GCS, GSS, and LSS is used as the sample importance score; Step 4: Sort the training set samples according to the importance score of each sample obtained in step 3, and determine the appropriate sample screening ratio to form important samples, including the following sub-steps; Step 41, determining the sample screening ratio, which is the proportion of important samples to the training set samples; Step 42: sort the training set samples from largest to smallest according to the sample importance scores obtained in step 3. Based on the sample screening ratio k determined in step 41, select the top k% samples with the largest importance scores from the sorting as important samples. Step 5: Input the important samples obtained in step 4 into the neural network model obtained in step 2 for training to obtain a trained neural network model, i.e., a prediction model; Step 6: Input the real-time vehicle data as test data into the prediction model to obtain the model output; calculate the residual between the test data and the model output. If it is greater than the set abnormality threshold, it is determined that the vehicle is abnormal at the current moment, otherwise it is normal.

2. The power battery anomaly detection method based on important sample screening under the big data framework of claim 1 is characterized in that: The feature importance analysis specifically uses the Pearson correlation coefficient to evaluate the correlation between data.

3. The power battery anomaly detection method based on important sample screening under the big data framework of claim 1 is characterized in that: Step 2 includes the following sub-steps: Step 21: divide the model input data obtained in step 1 into a training set and a test set; randomly set initial hyperparameters and define a model loss function; the model loss function Loss is defined as: Where, is the number of samples tested, and are the predicted value and the true value respectively; Step 22: Select multiple possible hyperparameter combinations from the initial hyperparameters, use a grid search method to test the model performance of each hyperparameter combination one by one, select the hyperparameter combination with the best performance for use in the neural network model, and determine the model structure; Step 23: input the training set into the neural network model obtained in step 22 for pre-training to extract the network parameters required for sample importance evaluation; The following principles are used for selection: 1) As the number of pre-training iterations increases, when the loss function Loss is less than Stop pre-training and save the network parameters at the current number of iterations ; 2) If the conditions corresponding to 1) are not met in the first 10 iterations of pre-training, stop pre-training and save the network parameters corresponding to the iteration number with the minimum training set loss in the first 10 iterations. .

4. The power battery anomaly detection method based on important sample screening under the big data framework of claim 1 is characterized in that: In step 41 , if the average standard deviation of the voltage of each single cell in the training set data is greater than 0.17, the optimal sample screening ratio is determined to be 40%; if it is less than 0.17, the optimal sample screening ratio is determined to be 30%.

5. The power battery anomaly detection method based on important sample screening under the big data framework according to claim 1 is characterized in that: Step 6 includes the following operations: input the test set data obtained in step 21 into the prediction model obtained in step 5, and calculate the residual between the model prediction value and the true value; use the box plot method to set the abnormal threshold. The calculation formula for setting the abnormal threshold of the box plot is as follows: Where, is the third quartile of the residual of the test set data, is the interquartile range, is the abnormal threshold.

Images