Lithium battery discharge data analysis method and system based on data analysis

Abstract

The invention discloses a lithium battery discharge data analysis method and system based on data analysis, and belongs to the technical field of artificial intelligence. The method comprises the following steps: acquiring real-time data of a battery through a sensor, and performing multi-source data fusion by using Kalman filtering to obtain a battery fusion data set; based on the battery fusion data set, collecting state data of the battery under different discharge conditions to form a discharge curve data set; extracting time domain features, frequency domain features and statistical features from the discharge curve data set; taking the extracted statistical characteristics as input, predicting the remaining life and the health state of the battery as output, and training by using a linear regression model; defining an optimization target, acquiring statistical characteristics of a discharge curve and defining constraints of discharge conditions according to the predicted residual life and health state of the battery, and searching an optimal discharge condition under given constraint conditions; a PID controller is adopted to design a control strategy, and the discharging behavior of the battery is adjusted according to real-time data and the optimal discharging condition.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a lithium battery discharge data analysis method and system based on data analysis. Background Art

[0002] With the widespread application of portable electronic devices, electric vehicles and energy storage systems, lithium batteries have become one of the most important energy storage devices. Lithium batteries have the advantages of high energy density, long life and no memory effect, and are widely used in various fields. However, during the use of lithium batteries, their performance and life are affected by many factors, such as temperature, current, voltage, etc. In order to ensure the reliability and safety of lithium batteries, accurate status monitoring and life prediction are required.

[0003] Existing technologies usually rely on a single data source and ignore the advantages brought by multi-source data fusion. They lack comprehensiveness in feature extraction, usually only focusing on time domain features, ignoring the combination of frequency domain features and statistical features, and only using simple regression models or empirical formulas, which cannot fully capture the complex changes in battery status. Existing technologies lack dynamic adjustment capabilities in optimizing discharge conditions and cannot adapt to changes in batteries and the environment in real time. The optimization algorithms used have large limitations and cannot find the best discharge conditions under complex constraints. Existing control strategies are usually preset fixed strategies that lack flexibility and cannot be adjusted in real time. Summary of the invention

[0004] The purpose of the present invention is to provide a lithium battery discharge data analysis method and system based on data analysis to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: A lithium battery discharge data analysis method based on data analysis, the method steps are as follows: S100, acquiring real-time data of the battery through a sensor, the real-time data including battery status data and environmental data, performing data preprocessing on the real-time data, and using Kalman filtering to perform multi-source data fusion on the historical data of the battery and the real-time data to obtain a battery fusion data set; S200, based on the battery fusion data set, collect the battery status data under different discharge conditions to form a discharge curve data set; extract time domain features from the discharge curve data set, apply fast Fourier transform to extract frequency domain features, and calculate the statistical features of the discharge curve; use the extracted statistical features as input, output the predicted remaining life and health state of the battery, use the linear regression model for training, and use the mean square error and determination coefficient for evaluation; S300, defining an optimization goal according to the predicted remaining life and health state of the battery, the optimization goal being to maximize the remaining life of the battery and maintain the health state, obtaining statistical characteristics of the discharge curve, defining constraints on the discharge condition, and using gradient descent to find the optimal discharge condition under the given constraints; S400, obtaining the best discharge condition, wherein the best discharge condition indicates that the battery is at the best voltage, current and temperature, designing a control strategy based on the real-time data of the battery, adjusting the discharge behavior of the battery based on the real-time data and the best discharge condition, wherein the control strategy uses a PID controller to adjust the discharge current and voltage; during the implementation process, continuously monitoring the battery status and environmental conditions, dynamically adjusting the discharge conditions to adapt to the changes, recording the actual remaining life and health status of the battery, and comparing and evaluating them with the predicted values.

[0006] According to step S100, the state data includes voltage, current, temperature, charging state and health state, and the environmental data includes temperature and humidity of the environment; after collecting the real-time data, data preprocessing is required, and the method is as follows: If the sensor does not provide data at a certain point in time, interpolation is used to process missing values; z-Score is used to detect and process outliers; in order to reduce the impact of noise on the data, a moving average algorithm is used for data smoothing; the battery's historical data and real-time data are fused through multi-source data, and the Kalman filter is used for fusion processing.

[0007] When the Kalman filter is used for fusion processing, the Kalman filter fusion step is divided into a prediction phase and an update phase; The formula in the prediction stage is as follows: , , ; in, represents the voltage predicted at time k before time k-1, represents the current predicted at time k before time k-1, represents the temperature predicted at time k before time k-1, Describes the variation of voltage V over time. Describes the temporal variation of the current I. Describes the temporal variation of temperature T; The voltage during the update phase is given by: ; ; ; is the Kalman gain of voltage, which is used to weigh the weight of historical voltage data and real-time measured voltage data. is the voltage prediction covariance matrix, which represents the uncertainty of the voltage prediction value before time k, is the voltage measurement matrix, which is used to map the predicted voltage value to the measurement space. is the voltage measurement noise covariance matrix, describing the noise properties of the voltage measurement value, T represents the transpose of the matrix, is the updated voltage, which represents the optimal state estimate of the voltage at time k, is the updated voltage covariance matrix, describing the uncertainty of the updated voltage state estimate, is the voltage measured in real time, and IM is the identity matrix.

[0008] Step S200 includes: S201, based on the battery fusion data set, collecting battery state data under different discharge conditions to form a discharge curve data set, wherein the different discharge conditions are a combination of different battery state data and different environmental data; S202, analyzing the time series data of the discharge curve to extract key features, wherein the key features include discharge time, voltage change rate, and current change rate, wherein the discharge time represents the time length from the start to the end of the battery discharge, and the voltage change rate and current change rate represent the speed of change of the voltage and current over time, respectively; S203, applying fast Fourier transform to convert the discharge curve in the time domain into the frequency domain, extracting frequency domain features, and reflecting the periodicity and frequency characteristics in the discharge process; S204, calculating the statistical characteristics of the discharge curve, including mean, variance and peak value, wherein the mean value represents the average level of current, voltage and temperature during the discharge process, the variance represents the fluctuation range of current, voltage and temperature during the discharge process, and the peak value represents the maximum and minimum values ​​of current, voltage and temperature during the discharge process; S205, organizing the extracted time domain features, frequency domain features and statistical features into feature vectors as inputs of a linear regression model, each feature vector corresponding to a discharge curve data sample; S206, using a linear regression model, taking the feature vector as input, predicting the remaining life and health state of the battery, and determining the parameters of the model through a training process so that the prediction result is as close to the true value as possible; S207, using the mean square error to evaluate the prediction error of the model, and using the determination coefficient to evaluate the goodness of fit of the model; S208. Adjust model parameters according to evaluation indicators to optimize model performance.

[0009] According to step S201, the performance of the battery under different discharge conditions needs to be considered. The combination of these conditions may include but is not limited to: different voltage levels: low voltage discharge, medium voltage discharge, high voltage discharge; different current levels: small current discharge, medium current discharge, large current discharge; different external temperature conditions: low temperature environment, room temperature environment, high temperature environment; different humidity conditions: low humidity environment, medium humidity environment and high humidity environment. The combination of these conditions forms a multi-dimensional test matrix. For example, high voltage discharge with medium current in low temperature environment, or low voltage discharge with small current in high temperature environment.

[0010] According to step S202, the time points of the start and end of the discharge are determined from the discharge curve, and the time difference between the two time points is calculated as the discharge time; the voltage change at each time point is calculated by subtracting the voltage at the previous time point from the voltage at the current time point, and the time interval at each time point is calculated by subtracting the time at the previous time point from the time at the current time point, and the voltage change rate is the ratio of the voltage change to the time interval; the current change at each time point is calculated by subtracting the current at the previous time point from the current at the current time point, and the current change rate is the ratio of the current change to the time interval.

[0011] According to step S203, the time domain signal is converted into a frequency domain signal, and the calculation formula of the fast Fourier transform is as follows: ; in, is a time domain signal, is a frequency domain signal, N is the length of the signal, and f is the frequency. By fast Fourier transform, the main frequency components and corresponding energy are extracted to reflect the periodicity and frequency characteristics of the discharge process.

[0012] According to step S206, the linear regression model is used to predict the target variable, which represents the remaining life span and health status, and the formula is as follows: ; Where y is the predicted remaining battery life and health status, arrive is the extracted feature, arrive is the regression coefficient corresponding to different features, is the error term; According to step S207, the prediction error of the model is evaluated using the mean square error, and the formula is as follows: ; in, is the actual value, is the predicted value, is the sample size; The coefficient of determination evaluates the goodness of fit of the model and is as follows: in, is the mean of the actual values.

[0013] According to step S300, the optimization goal is to maximize the remaining life RUL of the battery and maintain the state of health SOH, and these two factors are comprehensively considered by setting an objective function J: ; in, and They are and The weight parameter is used to balance the importance of RUL and SOH; Obtain statistical features of the discharge curve, including voltage V, current I, and temperature T, and define constraints on discharge conditions to constrain the voltage range, current range, and temperature range of the battery: , , ; in, and They represent the agreed maximum and minimum values ​​of voltage respectively. and They represent the maximum and minimum values ​​of the current respectively. and They represent the agreed maximum and the agreed minimum values ​​of temperature respectively; Use gradient descent to find the best discharge condition under given constraints and initialize the discharge condition parameters , calculate the objective function The gradient is: Among them, the formula for calculating the partial derivative of the objective function J with respect to the voltage V is as follows: The formula for calculating the partial derivative of the objective function J with respect to the current I is as follows: The formula for calculating the partial derivative of the objective function J with respect to temperature T is as follows: Update the discharge condition parameters: ; in, is the learning rate.

[0014] According to step S400, a PID controller is used to adjust the discharge current and voltage to maintain the battery operating under optimal discharge conditions, wherein the proportion of the PID controller is used to control the immediate response of the current and voltage, the integral of the PID controller is used to correct the accumulated error and eliminate the deviation, and the differential of the PID controller is used to predict future errors and reduce overshoot; The sensor continuously collects real-time data from the battery, compares the real-time data with the optimal discharge conditions, and calculates the deviation; based on the deviation of the real-time data, the PID controller dynamically adjusts the discharge current and voltage to make them close to the optimal discharge conditions. By adjusting the current and voltage, the discharge behavior of the battery is optimized to ensure operation under optimal conditions; the actual remaining life and health status of the battery are recorded, and the discharge conditions and corresponding battery status data after each adjustment are saved. The actual remaining life and health status are compared with the model predicted values, the accuracy of the prediction model is evaluated, the source of deviation is analyzed, and the model and control strategy are optimized; based on the evaluation results, the discharge conditions and control strategy are continuously optimized.

[0015] A lithium battery discharge data analysis system based on data analysis, comprising: Data acquisition and fusion module: including: sensor data acquisition unit, data preprocessing unit and multi-source data fusion unit; wherein, the sensor data acquisition unit is responsible for real-time acquisition of battery status data and environmental data, the data preprocessing unit is responsible for cleaning, denoising and preprocessing real-time data, processing missing values ​​and abnormal values, and the multi-source data fusion unit is responsible for using Kalman filtering to fuse the battery's historical data with real-time data to generate a battery fusion data set; Feature extraction and prediction module: including: data collection unit, time domain feature extraction unit, frequency domain feature extraction unit, statistical feature calculation unit and prediction model training unit; wherein, the data collection unit is responsible for collecting the state data of the battery under different discharge conditions to form a discharge curve data set, the time domain feature extraction unit is responsible for extracting time domain features from the discharge curve data set, including discharge time, voltage change rate and current change rate, the frequency domain feature extraction unit is responsible for applying fast Fourier transform to convert the discharge curve in the time domain into the frequency domain and extract frequency domain features, the statistical feature calculation unit is responsible for calculating the statistical features of the discharge curve, including mean, variance and peak value, and the prediction model training unit is responsible for training using a linear regression model, taking the extracted statistical features as input, and predicting the remaining life and health status of the battery; Optimization condition module: including: optimization target definition unit, constraint condition definition unit and optimal discharge condition search unit; wherein, the optimization target definition unit defines the optimization target, maximizes the remaining life of the battery and maintains the healthy state, the constraint condition definition unit is responsible for defining the constraints of the discharge condition, including the range of voltage, current and temperature, and the optimal discharge condition search unit is responsible for using the gradient descent method to find the optimal discharge condition under the given constraint conditions; Control and adjustment module: including: real-time data monitoring unit, PID control unit, dynamic adjustment unit and data recording and evaluation unit; among them, the real-time data monitoring unit is responsible for monitoring the battery status and environmental conditions according to the real-time data of the battery, the PID control unit is responsible for designing the control strategy and using the PID controller to adjust the discharge current and voltage, the dynamic adjustment unit is responsible for adjusting the discharge behavior of the battery according to the real-time data and the optimal discharge conditions during the implementation process, and the data recording and evaluation unit is responsible for recording the actual remaining life and health status of the battery, and comparing and evaluating them with the predicted values.

[0016] Compared with the prior art, the beneficial effects achieved by the present invention are: By adopting an optimization strategy for dynamically adjusting discharge conditions and a flexible control strategy, the present invention can adapt to changes in the battery and the environment in real time, maintain an optimal discharge state, and extend the service life of the battery.

[0017] By dynamically optimizing discharge conditions, maximizing the remaining life of the battery and maintaining a healthy state, the present invention can optimize the performance of the battery, increase the number of cycles and energy density of the battery, and reduce the cost of use.

[0018] By adopting multi-source data fusion and dynamic adjustment strategies, the present invention can monitor the battery status and environmental conditions in real time, adjust the discharge conditions in time, maintain stable operation of the system, and reduce the risk of system failure and damage. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 It is a schematic diagram of the steps of a lithium battery discharge data analysis method based on data analysis of the present invention; Figure 2 It is a system structure diagram of a lithium battery discharge data analysis system based on data analysis of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] See also Figure 1 and Figure 2 , the present invention provides a technical solution: According to one embodiment of the present invention, Figure 1 A step diagram of a lithium battery discharge data analysis method based on data analysis is shown in the figure. A lithium battery discharge data analysis method based on data analysis, the method steps are as follows: S100, acquiring real-time data of the battery through a sensor, the real-time data including battery status data and environmental data, performing data preprocessing on the real-time data, and using Kalman filtering to perform multi-source data fusion on the historical data of the battery and the real-time data to obtain a battery fusion data set; S200, based on the battery fusion data set, collect the battery status data under different discharge conditions to form a discharge curve data set; extract time domain features from the discharge curve data set, apply fast Fourier transform to extract frequency domain features, and calculate the statistical features of the discharge curve; use the extracted statistical features as input, output the predicted remaining life and health state of the battery, use the linear regression model for training, and use the mean square error and determination coefficient for evaluation; S300, defining an optimization goal according to the predicted remaining life and health state of the battery, the optimization goal being to maximize the remaining life of the battery and maintain the health state, obtaining statistical characteristics of the discharge curve, defining constraints on the discharge condition, and using gradient descent to find the optimal discharge condition under the given constraints; S400, obtaining the best discharge condition, wherein the best discharge condition indicates that the battery is at the best voltage, current and temperature, designing a control strategy based on the real-time data of the battery, adjusting the discharge behavior of the battery based on the real-time data and the best discharge condition, wherein the control strategy uses a PID controller to adjust the discharge current and voltage; during the implementation process, continuously monitoring the battery status and environmental conditions, dynamically adjusting the discharge conditions to adapt to the changes, recording the actual remaining life and health status of the battery, and comparing and evaluating them with the predicted values.

[0022] According to step S100, the state data includes voltage, current, temperature, charging state and health state, and the environmental data includes temperature and humidity of the environment; after collecting the real-time data, data preprocessing is required, and the method is as follows: If the sensor does not provide data at a certain point in time, interpolation is used to process missing values; z-Score is used to detect and process outliers; in order to reduce the impact of noise on the data, a moving average algorithm is used for data smoothing; the battery's historical data and real-time data are fused through multi-source data, and the Kalman filter is used for fusion processing.

[0023] When the Kalman filter is used for fusion processing, the Kalman filter fusion step is divided into a prediction phase and an update phase; The formula in the prediction stage is as follows: , , ; in, represents the voltage predicted at time k before time k-1, represents the current predicted at time k before time k-1, represents the temperature predicted at time k before time k-1, Describes the variation of voltage V over time. Describes the temporal variation of the current I. Describes the temporal variation of temperature T; The voltage during the update phase is given by: ; ; ; is the Kalman gain of voltage, which is used to weigh the weight of historical voltage data and real-time measured voltage data. is the voltage prediction covariance matrix, which represents the uncertainty of the voltage prediction value before time k, is the voltage measurement matrix, which is used to map the predicted voltage value to the measurement space. is the voltage measurement noise covariance matrix, describing the noise properties of the voltage measurement value, T represents the transpose of the matrix, is the updated voltage, which represents the estimated value of the optimal voltage state at time k. is the updated voltage covariance matrix, describing the uncertainty of the updated voltage state estimate, is the voltage measured in real time, and IM is the identity matrix.

[0024] Step S200 includes: S201, based on the battery fusion data set, collecting battery state data under different discharge conditions to form a discharge curve data set, wherein the different discharge conditions are a combination of different battery state data and different environmental data; S202, analyzing the time series data of the discharge curve to extract key features, wherein the key features include discharge time, voltage change rate, and current change rate, wherein the discharge time represents the time length from the start to the end of the battery discharge, and the voltage change rate and current change rate represent the speed of change of the voltage and current over time, respectively; S203, applying fast Fourier transform to convert the discharge curve in the time domain into the frequency domain, extracting frequency domain features, and reflecting the periodicity and frequency characteristics in the discharge process; S204, calculating the statistical characteristics of the discharge curve, including mean, variance and peak value, wherein the mean value represents the average level of current, voltage and temperature during the discharge process, the variance represents the fluctuation range of current, voltage and temperature during the discharge process, and the peak value represents the maximum and minimum values ​​of current, voltage and temperature during the discharge process; S205, organizing the extracted time domain features, frequency domain features and statistical features into feature vectors as inputs of a linear regression model, each feature vector corresponding to a discharge curve data sample; S206, using a linear regression model, taking the feature vector as input, predicting the remaining life and health state of the battery, and determining the parameters of the model through a training process so that the prediction result is as close to the true value as possible; S207, using the mean square error to evaluate the prediction error of the model, and using the determination coefficient to evaluate the goodness of fit of the model; S208. Adjust model parameters according to evaluation indicators to optimize model performance.

[0025] According to step S202, the time points of the start and end of the discharge are determined from the discharge curve, and the time difference between the two time points is calculated as the discharge time; the voltage change at each time point is calculated by subtracting the voltage at the previous time point from the voltage at the current time point, and the time interval at each time point is calculated by subtracting the time at the previous time point from the time at the current time point, and the voltage change rate is the ratio of the voltage change to the time interval; the current change at each time point is calculated by subtracting the current at the previous time point from the current at the current time point, and the current change rate is the ratio of the current change to the time interval.

[0026] According to step S203, the time domain signal is converted into a frequency domain signal, and the calculation formula of the fast Fourier transform is as follows: ; in, is a time domain signal, is a frequency domain signal, N is the length of the signal, and f is the frequency. By fast Fourier transform, the main frequency components and corresponding energy are extracted to reflect the periodicity and frequency characteristics of the discharge process.

[0027] According to step S206, the linear regression model is used to predict the target variable, which represents the remaining life span and health status, and the formula is as follows: ; Where y is the predicted remaining battery life and health status, arrive is the extracted feature, arrive is the regression coefficient corresponding to different features, is the error term; According to step S207, the prediction error of the model is evaluated using the mean square error, and the formula is as follows: ; in, is the actual value, is the predicted value, is the sample size; The coefficient of determination evaluates the goodness of fit of the model and is as follows: in, is the mean of the actual values.

[0028] According to step S300, the optimization goal is to maximize the remaining life RUL of the battery and maintain the state of health SOH, and these two factors are comprehensively considered by setting an objective function J: ; in, and They are and The weight parameter is used to balance the importance of RUL and SOH; set , .

[0029] Obtain statistical features of the discharge curve, including voltage V, current I, and temperature T, and define constraints on discharge conditions to constrain the voltage range, current range, and temperature range of the battery: , , ; In this embodiment, it is set: , , , , , .

[0030] in, and They represent the agreed maximum and minimum values ​​of voltage respectively. and They represent the maximum and minimum values ​​of the current respectively. and They represent the agreed maximum and the agreed minimum values ​​of temperature respectively; Use gradient descent to find the best discharge condition under given constraints and initialize the discharge condition parameters ,in , , , calculate the objective function The gradient is: Among them, the formula for calculating the partial derivative of the objective function J with respect to the voltage V is as follows: The formula for calculating the partial derivative of the objective function J with respect to the current I is as follows: The formula for calculating the partial derivative of the objective function J with respect to temperature T is as follows: Update the discharge condition parameters: ; Update voltage: ; Update current: ; Update temperature: .

[0031] in, is the learning rate, set to 0.01.

[0032] According to step S400, a PID controller is used to adjust the discharge current and voltage to maintain the battery operating under optimal discharge conditions, wherein the proportion of the PID controller is used to control the immediate response of the current and voltage, the integral of the PID controller is used to correct the accumulated error and eliminate the deviation, and the differential of the PID controller is used to predict future errors and reduce overshoot; The sensor continuously collects real-time data from the battery, compares the real-time data with the optimal discharge conditions, and calculates the deviation; based on the deviation of the real-time data, the PID controller dynamically adjusts the discharge current and voltage to make them close to the optimal discharge conditions. By adjusting the current and voltage, the discharge behavior of the battery is optimized to ensure operation under optimal conditions; the actual remaining life and health status of the battery are recorded, and the discharge conditions and corresponding battery status data after each adjustment are saved. The actual remaining life and health status are compared with the model predicted values, the accuracy of the prediction model is evaluated, the source of deviation is analyzed, and the model and control strategy are optimized; based on the evaluation results, the discharge conditions and control strategy are continuously optimized.

[0033] According to another embodiment of the present invention, Figure 2 As shown in the system structure diagram of a lithium battery discharge data analysis system based on data analysis, a lithium battery discharge data analysis system based on data analysis includes: Data acquisition and fusion module: including: sensor data acquisition unit, data preprocessing unit and multi-source data fusion unit; wherein, the sensor data acquisition unit is responsible for real-time acquisition of battery status data and environmental data, the data preprocessing unit is responsible for cleaning, denoising and preprocessing real-time data, processing missing values ​​and abnormal values, and the multi-source data fusion unit is responsible for using Kalman filtering to fuse the battery's historical data with real-time data to generate a battery fusion data set; Feature extraction and prediction module: including: data collection unit, time domain feature extraction unit, frequency domain feature extraction unit, statistical feature calculation unit and prediction model training unit; wherein, the data collection unit is responsible for collecting the state data of the battery under different discharge conditions to form a discharge curve data set, the time domain feature extraction unit is responsible for extracting time domain features from the discharge curve data set, including discharge time, voltage change rate and current change rate, the frequency domain feature extraction unit is responsible for applying fast Fourier transform to convert the discharge curve in the time domain into the frequency domain and extract frequency domain features, the statistical feature calculation unit is responsible for calculating the statistical features of the discharge curve, including mean, variance and peak value, and the prediction model training unit is responsible for training using a linear regression model, taking the extracted statistical features as input, and predicting the remaining life and health status of the battery; In this embodiment, the data collection unit uses sensors to obtain the state data of the battery under different discharge conditions, including voltage, current, and temperature, etc., preprocesses and cleans the collected data, and organizes the preprocessed data into a discharge curve data set. The time domain feature extraction unit determines the time points of the start and end of the discharge according to the discharge curve data set, calculates the time difference between the two time points as the discharge time, calculates the voltage and current changes at each time point, and obtains the voltage change rate and current change rate by subtracting the voltage at the previous time point and the current change from the voltage at the current time point, and then dividing by the time interval. The frequency domain feature extraction unit performs a fast Fourier transform on the discharge curve data, converts the time domain signal into a frequency domain signal, and extracts the main frequency components and corresponding energy from the frequency domain signal to reflect the periodicity and frequency characteristics during the discharge process.

[0034] The statistical feature calculation unit calculates the mean of the current, voltage and temperature during the discharge process to reflect the average level during the discharge process; calculates the variance of the current, voltage and temperature during the discharge process to reflect the fluctuation amplitude during the discharge process; calculates the peak values ​​of the current, voltage and temperature during the discharge process to reflect the extreme value during the discharge process. The prediction model training unit organizes the extracted time domain features, frequency domain features and statistical features into feature vectors as the input of the linear regression model, uses the linear regression model training data, and adjusts the model parameters to make the prediction results as close to the true value as possible.

[0035] After implementing the above feature extraction and prediction modules, the following implementation results were obtained: time domain, frequency domain and statistical features were successfully extracted from the battery discharge curve data, and these features can effectively reflect the characteristics of the battery discharge process. The remaining life and health status of the battery were successfully predicted using the linear regression model. The prediction results of the model are consistent with the actual situation and have high accuracy and reliability. Through the evaluation of the prediction model, it was found that the mean square error was small and the determination coefficient was high, indicating that the prediction error of the model was small and the goodness of fit was high.

[0036] Optimization condition module: including: optimization target definition unit, constraint condition definition unit and optimal discharge condition search unit; wherein, the optimization target definition unit defines the optimization target, maximizes the remaining life of the battery and maintains the healthy state, the constraint condition definition unit is responsible for defining the constraints of the discharge condition, including the range of voltage, current and temperature, and the optimal discharge condition search unit is responsible for using the gradient descent method to find the optimal discharge condition under the given constraint conditions; Control and adjustment module: including: real-time data monitoring unit, PID control unit, dynamic adjustment unit and data recording and evaluation unit; among them, the real-time data monitoring unit is responsible for monitoring the battery status and environmental conditions according to the real-time data of the battery, the PID control unit is responsible for designing the control strategy and using the PID controller to adjust the discharge current and voltage, the dynamic adjustment unit is responsible for adjusting the discharge behavior of the battery according to the real-time data and the optimal discharge conditions during the implementation process, and the data recording and evaluation unit is responsible for recording the actual remaining life and health status of the battery, and comparing and evaluating them with the predicted values.

[0037] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0038] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A lithium battery discharge data analysis method based on data analysis, characterized in that: The method steps are as follows: S100, acquiring real-time data of the battery through a sensor, wherein the real-time data includes status data and environmental data of the battery, performing data preprocessing on the real-time data, and performing multi-source data fusion on the historical data of the battery and the real-time data using a Kalman filter to obtain a battery fusion data set; S200, based on the battery fusion data set, collecting battery state data under different discharge conditions to form a discharge curve data set; Extract time domain features from the discharge curve data set, apply fast Fourier transform to extract frequency domain features, and calculate the statistical characteristics of the discharge curve; The extracted statistical features are used as input, and the output is the predicted remaining battery life and health status. The linear regression model is used for training and the mean square error and determination coefficient are used for evaluation. S300, defining an optimization goal according to the predicted remaining life and health state of the battery, the optimization goal being to maximize the remaining life of the battery and maintain the health state, obtaining statistical characteristics of the discharge curve, defining constraints on the discharge condition, and using gradient descent to find the optimal discharge condition under the given constraints; S400, obtaining the best discharge condition, wherein the best discharge condition indicates that the battery is at the best voltage, current and temperature, designing a control strategy based on the real-time data of the battery, adjusting the discharge behavior of the battery based on the real-time data and the best discharge condition, wherein the control strategy uses a PID controller to adjust the discharge current and voltage; during the implementation process, continuously monitoring the battery status and environmental conditions, dynamically adjusting the discharge conditions to adapt to the changes, recording the actual remaining life and health status of the battery, and comparing and evaluating them with the predicted values.

2. The lithium battery discharge data analysis method based on data analysis according to claim 1, characterized in that: According to step S100, the state data includes voltage, current, temperature, charging state and health state, and the environmental data includes temperature and humidity of the environment; after collecting the real-time data, data preprocessing is required, and the method is as follows: If the sensor does not provide data at a certain point in time, interpolation is used to process missing values; z-Score is used to detect and process outliers; in order to reduce the impact of noise on the data, a moving average algorithm is used for data smoothing; the battery's historical data and real-time data are fused through multi-source data, and the Kalman filter is used for fusion processing.

3. A lithium battery discharge data analysis method based on data analysis according to claim 2, characterized in that: When the Kalman filter is used for fusion processing, the Kalman filter fusion step is divided into a prediction phase and an update phase; In the prediction stage, the voltage, current and temperature need to be predicted, and the voltage, current and temperature at the current moment are inferred by using the temporal variation law of the voltage, current and temperature at the previous moment; In the update phase, the real-time measured data is used to correct the predicted value to obtain a more accurate state estimate. After considering the real-time measured data, the predicted value is corrected. The Kalman gain is used to weigh the weights of the historical voltage data and the real-time measured voltage data, which affects the corrected estimated value. The measurement matrix is ​​used to map the predicted value of voltage to the measurement space, and the measurement noise covariance matrix describes the noise properties of the measured value; An updated voltage state estimation value and an updated covariance matrix are obtained through the Kalman gain, the measurement matrix, the measurement noise covariance matrix and the real-time measured voltage value.

4. The lithium battery discharge data analysis method based on data analysis according to claim 1, characterized in that: Step S200 includes: S201, based on the battery fusion data set, collecting battery state data under different discharge conditions to form a discharge curve data set, wherein the different discharge conditions are a combination of different battery state data and different environmental data; S202, analyzing the time series data of the discharge curve to extract key features, wherein the key features include discharge time, voltage change rate, and current change rate, wherein the discharge time represents the time length from the start to the end of the battery discharge, and the voltage change rate and current change rate represent the speed of change of the voltage and current over time, respectively; S203, applying fast Fourier transform to convert the discharge curve in the time domain into the frequency domain, extracting frequency domain features, and reflecting the periodicity and frequency characteristics in the discharge process; S204, calculating the statistical characteristics of the discharge curve, including mean, variance and peak value, wherein the mean value represents the average level of current, voltage and temperature during the discharge process, the variance represents the fluctuation range of current, voltage and temperature during the discharge process, and the peak value represents the maximum and minimum values ​​of current, voltage and temperature during the discharge process; S205, organizing the extracted time domain features, frequency domain features and statistical features into feature vectors as inputs of a linear regression model, each feature vector corresponding to a discharge curve data sample; S206, using a linear regression model, taking the feature vector as input, predicting the remaining life and health state of the battery, and determining the parameters of the model through a training process so that the prediction result is as close to the true value as possible; S207, using the mean square error to evaluate the prediction error of the model, and using the determination coefficient to evaluate the goodness of fit of the model; S208. Adjust model parameters according to evaluation indicators to optimize model performance.

5. The lithium battery discharge data analysis method based on data analysis according to claim 4, characterized in that: According to step S202, the time points of the start and end of the discharge are determined from the discharge curve, and the time difference between the two time points is calculated as the discharge time; the voltage change at each time point is calculated by subtracting the voltage at the previous time point from the voltage at the current time point, and the time interval at each time point is calculated by subtracting the time at the previous time point from the time at the current time point, and the voltage change rate is the ratio of the voltage change to the time interval; the current change at each time point is calculated by subtracting the current at the previous time point from the current at the current time point, and the current change rate is the ratio of the current change to the time interval.

6. A lithium battery discharge data analysis method based on data analysis according to claim 4, characterized in that: According to step S203, the time domain signal is converted into a frequency domain signal, and the fast Fourier transform represents the signal as a combination of a series of sine and cosine functions, decomposes the signal into components of different frequencies, and calculates the corresponding energy; in the discharge curve data analysis, the fast Fourier transform is used to extract the frequency characteristics of the discharge process to understand the periodicity and frequency characteristics of the battery discharge; in the calculation process, the fast Fourier transform is applied to the time domain signal to obtain the frequency domain signal; the main frequency components and the corresponding energy are extracted from the frequency domain signal, so as to obtain the oscillation conditions of different frequencies during the discharge process.

7. The lithium battery discharge data analysis method based on data analysis according to claim 4, characterized in that: According to step S206, the remaining life and health state of the battery are predicted using a linear regression model, where the prediction target variable of the model represents the remaining life and health state of the battery, and a series of extracted features are combined with corresponding regression coefficients through linear regression to obtain a predicted value of the target variable, wherein the series of extracted features include time domain features, frequency domain features, and statistical features; According to step S207, the prediction error of the model is evaluated using the mean square error, which represents the average of the squares of the differences between the actual values ​​and the predicted values. For a data set, the square of the prediction error of each sample is calculated, and then the square errors of all samples are averaged to obtain the mean square error. The determination coefficient is used to evaluate the goodness of fit of the model, which represents the degree to which the model explains the variability of the actual values. The value range is between 0 and 1. The closer it is to 1, the better the model fits, and the closer it is to 0, the worse the model fits.

8. The lithium battery discharge data analysis method based on data analysis according to claim 1, characterized in that: According to step S300, the optimization goal is to maximize the remaining life RUL of the battery and maintain the state of health SOH, and these two factors are comprehensively considered by setting an objective function; Obtain statistical features of the discharge curve, the statistical features including voltage, current and temperature, define constraints on discharge conditions, constrain the voltage range, current range and temperature range of the battery, and need to constrain the agreed maximum and agreed minimum values ​​of voltage, the agreed maximum and agreed minimum values ​​of current and the agreed maximum and agreed minimum values ​​of temperature; Use gradient descent to find the optimal discharge condition under given constraints, initialize the discharge condition parameters, calculate the partial derivatives of the objective function with respect to voltage, current and temperature, and obtain the rate of change of the objective function with respect to voltage, current and temperature; the partial derivatives represent the rate of change of the objective function in each direction, which can guide the adjustment of the discharge condition in the direction that increases the objective function; finally, through several iterative updates, the discharge condition that maximizes the objective function under given constraints can be found, thereby achieving the optimal discharge strategy for the battery.

9. The lithium battery discharge data analysis method based on data analysis according to claim 1, characterized in that: According to step S400, a PID controller is used to adjust the discharge current and voltage to maintain the battery operating under optimal discharge conditions, wherein the proportion of the PID controller is used to control the immediate response of the current and voltage, the integral of the PID controller is used to correct the accumulated error and eliminate the deviation, and the differential of the PID controller is used to predict future errors and reduce overshoot; The sensor continuously collects real-time data from the battery, compares the real-time data with the optimal discharge conditions, and calculates the deviation; based on the deviation of the real-time data, the PID controller dynamically adjusts the discharge current and voltage to make them close to the optimal discharge conditions. By adjusting the current and voltage, the discharge behavior of the battery is optimized to ensure operation under optimal conditions; the actual remaining life and health status of the battery are recorded, and the discharge conditions and corresponding battery status data after each adjustment are saved. The actual remaining life and health status are compared with the model predicted values, the accuracy of the prediction model is evaluated, the source of deviation is analyzed, and the model and control strategy are optimized; based on the evaluation results, the discharge conditions and control strategy are continuously optimized.

10. A lithium battery discharge data analysis system based on data analysis, using a lithium battery discharge data analysis method based on data analysis according to any one of claims 1 to 9, characterized in that: include: Data acquisition and fusion module: including: sensor data acquisition unit, data preprocessing unit and multi-source data fusion unit; wherein, the sensor data acquisition unit is responsible for real-time acquisition of battery status data and environmental data, the data preprocessing unit is responsible for cleaning, denoising and preprocessing real-time data, processing missing values ​​and abnormal values, and the multi-source data fusion unit is responsible for using Kalman filtering to fuse the battery's historical data with real-time data to generate a battery fusion data set; Feature extraction and prediction module: including: data collection unit, time domain feature extraction unit, frequency domain feature extraction unit, statistical feature calculation unit and prediction model training unit; wherein, the data collection unit is responsible for collecting the state data of the battery under different discharge conditions to form a discharge curve data set, the time domain feature extraction unit is responsible for extracting time domain features from the discharge curve data set, including discharge time, voltage change rate and current change rate, the frequency domain feature extraction unit is responsible for applying fast Fourier transform to convert the discharge curve in the time domain into the frequency domain and extract frequency domain features, the statistical feature calculation unit is responsible for calculating the statistical features of the discharge curve, including mean, variance and peak value, and the prediction model training unit is responsible for training using a linear regression model, taking the extracted statistical features as input, and predicting the remaining life and health status of the battery; Optimization condition module: including: optimization target definition unit, constraint condition definition unit and optimal discharge condition search unit; wherein, the optimization target definition unit defines the optimization target, maximizes the remaining life of the battery and maintains the healthy state, the constraint condition definition unit is responsible for defining the constraints of the discharge condition, including the range of voltage, current and temperature, and the optimal discharge condition search unit is responsible for using the gradient descent method to find the optimal discharge condition under the given constraint conditions; Control and adjustment module: including: real-time data monitoring unit, PID control unit, dynamic adjustment unit and data recording and evaluation unit; among them, the real-time data monitoring unit is responsible for monitoring the battery status and environmental conditions according to the real-time data of the battery, the PID control unit is responsible for designing the control strategy and using the PID controller to adjust the discharge current and voltage, the dynamic adjustment unit is responsible for adjusting the discharge behavior of the battery according to the real-time data and the optimal discharge conditions during the implementation process, and the data recording and evaluation unit is responsible for recording the actual remaining life and health status of the battery, and comparing and evaluating them with the predicted values.

Images