Endurance warning method and system based on BMS data

Abstract

The present invention discloses a battery life warning method and system based on BMS data, and relates to the field of battery management technology. The battery life warning method based on BMS data includes the following steps: establishing an equivalent circuit model of the battery; generating a corresponding charge and discharge curve; evaluating the charge state and health state of each battery; generating balanced charging information of each battery in the BMS based on the equivalent circuit model; constructing a charging fluctuation change chart, analyzing the results to determine whether the battery is abnormal; and performing a battery life warning based on the judgment result. The present invention helps to more accurately understand the charge state and health state of the battery by establishing an equivalent circuit model and simulating the charging and discharging process, and performs a timing analysis based on the charging fluctuation change trend, which can assist in discovering battery abnormalities or potential problems, thereby improving the performance, safety and user experience of electric vehicles, while reducing energy consumption and maintenance costs.

Description

Technical Field

[0001] The present invention relates to the field of battery management technology, and in particular to a battery life warning method and system based on BMS data. Background Art

[0002] With the development of lithium-ion battery technology, the application scope of lithium-ion batteries is becoming wider and wider. In addition to portable electronic products, they are also widely used in electric vehicles, large power sources and energy storage fields. Except for small portable electronic products, other applications require the battery voltage to be higher than the existing single cell voltage. For example, pure electric vehicles require the battery voltage to reach more than 100V, and electric bicycles require the battery voltage to be 36 or 48V. High power output and high capacity requirements require the prepared single cells to be used in combination. However, due to differences in raw materials, production processes, production batches and manufacturing technologies, single cells of the same model and specification will have differences in performance parameters such as voltage, capacity and its attenuation rate, internal resistance and its rate of change over time during charging and discharging, life, self-discharge rate, etc.

[0003] In recent years, new energy vehicles have developed rapidly. Because new energy vehicles meet people's daily travel needs, are pollution-free, and are beneficial to environmental protection, they are considered to be one of the effective ways to achieve energy conservation and emission reduction. Among all the components of new energy vehicles, the battery system is the most core component, which directly determines the service life and safety of new energy vehicles. The service life of the battery system depends to a large extent on the service life of the single cell, and is also affected by the use conditions such as the vehicle's operating temperature and SOC state. In the prior art, traditional BMS mainly relies on empirical models and fixed parameters to manage the charging and discharging behavior of the battery. However, this method is not accurate enough in managing battery efficiency when facing complex use environments and multiple battery types, and it is not easy to evaluate the battery's charge state and health state, and it is not easy to find battery abnormalities or potential problems, thereby increasing the risk of battery failure.

[0004] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention proposes a battery life warning method and system based on BMS data, which solves the problem that in the prior art proposed in the above background technology, the traditional BMS mainly relies on empirical models and fixed parameters to manage the charging and discharging behavior of the battery. However, this method is not accurate enough in managing battery efficiency when facing complex usage environments and multiple battery types, and is not convenient for evaluating the charge state and health state of the battery, and is also not convenient for discovering battery abnormalities or potential problems, thereby increasing the risk of battery failure.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0007] According to one aspect of the present invention, a battery life warning method based on BMS data is provided, and the battery life warning method based on BMS data comprises the following steps:

[0008] S1. Obtain the parameter data of the battery in the BMS, perform preprocessing, and establish an equivalent circuit model of the battery based on the processing results;

[0009] S2. Simulate the charge and discharge process of each battery through an equivalent circuit model to generate a corresponding charge and discharge curve;

[0010] S3, using an optimization algorithm to compare the generated charge and discharge curves with the actual charge and discharge curves to evaluate the charge state and health state of each battery;

[0011] S4. Generate balanced charging information of each battery in the BMS based on the evaluation result and the equivalent circuit model;

[0012] S5. Obtain the charging fluctuation change trend according to the balanced charging information, construct a charging fluctuation change chart, perform time series analysis, and determine whether the battery is abnormal according to the analysis results;

[0013] S6. Based on the judgment results, formulate a preset adjustment plan and issue a life warning.

[0014] Furthermore, obtaining the parameter data of the battery in the BMS and preprocessing it, and establishing the equivalent circuit model of the battery according to the processing results includes the following steps:

[0015] S11, deploying data acquisition nodes in the network environment, and obtaining the original parameter data of the battery in the BMS through the communication protocol;

[0016] S12, extracting original parameter data and performing denoising processing;

[0017] S13, performing data verification on the original parameter data after denoising, extracting valid data, and performing normalization processing to generate accurate data;

[0018] S14. Extracting parameter data features of the battery from accurate data using Hilbert transform;

[0019] S15, determining the components of the battery according to the number, connection mode and model information of the battery, and reading the parameter data characteristics of the battery;

[0020] S16. According to the actual situation of the battery, an equivalent circuit model is established based on the PNGV model;

[0021] S17, importing the acquired parameter data of each battery into the circuit model to construct an equivalent circuit model of the battery.

[0022] Furthermore, extracting the original parameter data and performing denoising processing includes the following steps:

[0023] S121, performing wavelet transform on the original parameter data, converting the original parameter data from the time domain to the frequency domain, and screening out wave field records of each frequency band containing interference waves;

[0024] S122, tracking interference waves one by one in the wave field records of each frequency band, and obtaining the interference wave direction at the calculation sample point;

[0025] S123, selecting a number of windows centered on the calculation sample point to perform median filtering, and restoring the interference signal at the calculation sample point;

[0026] S124, recovering the interference signals in the wave field records of each frequency band one by one, and performing inverse wavelet transform to obtain the entire wave field of the interference wave;

[0027] S125 . Subtract the interference wave field from the original parameter data to obtain the effective signal wave field, and generate the denoised original parameter data based on the effective signal wave field.

[0028] Furthermore, simulating the charge and discharge process of each battery by an equivalent circuit model and generating a corresponding charge and discharge curve includes the following steps:

[0029] S21, determining parameter data of an equivalent circuit model, and importing the parameter data of the battery into the equivalent circuit model;

[0030] S22, using an equivalent circuit model to simulate the charging and discharging process of each battery;

[0031] S23. During the simulation process, the parameter data of voltage, current and charge state in the charge and discharge curve of each battery are recorded;

[0032] S24, generating a corresponding charge and discharge curve according to the parameter data of voltage, current and charge state in the charge and discharge curve of each battery.

[0033] Furthermore, the generated charge and discharge curves are compared with the actual charge and discharge curves using an optimization algorithm, and the charge state and health state of each battery are evaluated, including the following steps:

[0034] S31, initializing the optimization algorithm, setting the initial temperature and the initial charge and discharge curve of the battery;

[0035] S32, repeating steps S33 to S36 according to the set initial temperature;

[0036] S33, randomly perturbing the initial charge and discharge curve to generate a new charge and discharge curve;

[0037] S34, calculating the difference between the fitness value of the new charge and discharge curve and the fitness value of the initial charge and discharge curve;

[0038] S35. If the fitness value of the new charge-discharge curve is higher than the fitness value of the initial charge-discharge curve, the new charge-discharge curve is accepted as the current charge-discharge curve. Otherwise, according to the Metropolis rule, the acceptance probability of the new charge-discharge curve is calculated. If the acceptance probability is greater than a random number, the new charge-discharge curve is accepted as the current charge-discharge curve. Otherwise, the current charge-discharge curve is retained.

[0039] S36, if the set termination condition is met, the current charge-discharge curve is output as the optimal charge-discharge curve, and the termination condition is that the new charge-discharge curve is not accepted in several consecutive Metropolis chains or reaches the end temperature, otherwise, the temperature is attenuated by the attenuation function and then the process returns to step S32;

[0040] S37, repeating steps S33 to S36 until the optimal charge and discharge curve is found and compared with the actual charge and discharge curve, and the charge state and health state of each battery are evaluated.

[0041] Furthermore, the calculation formula of the attenuation function is:

[0042]

[0043] Among them, t s It is expressed as the temperature value at the θth step;

[0044] t0 represents the initial temperature.

[0045] Further, generating the balanced charging information of each battery in the BMS based on the evaluation result and the equivalent circuit model includes the following steps:

[0046] S41, generating battery historical charging process data based on the equivalent circuit model and the evaluation result;

[0047] S42, analyzing the charging status of each battery based on historical charging process data, and generating a balanced charging parameter according to the battery charging status;

[0048] S43: verify the balanced charging parameters, and generate balanced charging information according to the verification result.

[0049] Further, obtaining the charging fluctuation change trend according to the balanced charging information, constructing a charging fluctuation change chart, performing a time series analysis, and judging whether the battery is abnormal according to the analysis result includes the following steps:

[0050] S51, performing charging fluctuation statistical analysis according to the balanced charging information to obtain a charging fluctuation change trend;

[0051] S52, extracting change features based on the charging fluctuation change trend, and constructing a charging fluctuation change chart according to the extracted feature results;

[0052] S53, marking the charging fluctuation change chart with information, and presetting the charging fluctuation safety threshold;

[0053] S54, extracting battery fluctuation parameters from the charging fluctuation change chart, and performing spectrum analysis on the battery fluctuation parameters to obtain a change pattern and trend of the battery fluctuation parameters;

[0054] S55, extracting change characteristics according to the change mode and change trend of the battery fluctuation parameter to obtain change characteristic data;

[0055] S56. Analyze the change characteristic data based on the abnormality analysis algorithm, and determine whether the battery has an abnormality based on the analysis result.

[0056] Further, analyzing the change characteristic data based on the abnormality analysis algorithm and judging whether the battery is abnormal according to the analysis result includes the following steps:

[0057] S561, setting preset parameters, and generating battery change characteristic data and a corresponding trend matrix according to the preset parameters;

[0058] S562, calculating the fitness function value of the change feature data, where the fitness function is used to measure the quality of the change feature data;

[0059] S563, calculate the population fitness variance of all the change feature data, and when the preset conditions are met, go to step S565 if a mutation occurs, otherwise go to step S564;

[0060] S564, updating the inertia weight point according to the formula for calculating the inertia weight, and updating the speed and position of the change characteristic data according to the speed update formula and the position update formula of the change characteristic data respectively;

[0061] S565, check whether the end condition is met, if so, end the optimization search, otherwise go to S562 to continue the optimization search;

[0062] S566, assigning the found optimal parameter vector to the SVM model;

[0063] S567, inputting the processed sample data into the SVM model, and training the SVM to obtain a trained SVM classification model;

[0064] S568, using the trained SVM classification model to predict the prediction sample, obtain the final analysis result, and stop the calculation;

[0065] S569. Determine whether there is any abnormality in the battery based on the final analysis result.

[0066] According to another aspect of the present invention, a battery life warning system based on BMS data is also provided, and the battery life warning system based on BMS data includes: a data processing and model building module, a battery charge and discharge simulation and curve generation module, a model evaluation and battery status analysis module, a balance charging information generation module, a timing analysis and abnormality judgment module, and a decision support and warning module;

[0067] The data processing and model building module is used to obtain the parameter data of the battery in the BMS, perform preprocessing, and build an equivalent circuit model of the battery based on the processing results;

[0068] The battery charge and discharge simulation and curve generation module is used to simulate the charge and discharge process of each battery through an equivalent circuit model and generate the corresponding charge and discharge curve;

[0069] Model evaluation and battery status analysis module, which is used to compare the generated charge and discharge curves with the actual charge and discharge curves using an optimization algorithm to evaluate the charge state and health state of each battery;

[0070] A balanced charging information generation module is used to generate balanced charging information of each battery in the BMS according to the evaluation results and based on the equivalent circuit model;

[0071] The timing analysis and abnormality judgment module is used to obtain the charging fluctuation change trend based on the balanced charging information, build a charging fluctuation change chart, perform timing analysis, and judge whether the battery is abnormal based on the analysis results;

[0072] The decision support and early warning module is used to formulate preset adjustment plans and issue endurance warnings based on the judgment results.

[0073] The beneficial effects of the present invention are:

[0074] 1. The present invention helps to more accurately understand the charge state and health state of the battery by establishing an equivalent circuit model and simulating the charging and discharging process. It also performs timing analysis based on the charging fluctuation trend, which can assist in discovering battery anomalies or potential problems, thereby improving the performance, safety and user experience of electric vehicles, while reducing energy consumption and maintenance costs.

[0075] 2. The present invention obtains battery parameter data in the BMS, establishes an equivalent circuit model and simulates the charging and discharging process, so as to more accurately understand the charge state and health state of the battery, and optimizes the charging strategy based on the balanced charging information generated according to the evaluation results, thereby improving the performance and service life of the battery.

[0076] 3. The present invention can detect possible abnormalities in the battery through time-series analysis of the charging fluctuation trend, and timely detection of abnormalities helps to take measures in advance to prevent battery failure. At the same time, by analyzing the charging fluctuation change chart, the problems with the battery can be located more accurately. And through real-time monitoring and analysis of the battery charging fluctuation, the remaining battery life can be predicted more accurately, which helps users make relevant preparations in advance and reasonably plan driving routes or arrange charging.

[0077] 4. The present invention can more accurately evaluate the SOC and SOH of the battery by comparing the charge and discharge curves through the optimization algorithm, thereby providing a reliable basis for further optimization of the charging strategy and abnormality detection. In addition, evaluating the battery health status helps to discover battery performance degradation problems, thereby adjusting the charging strategy or taking maintenance measures to extend the battery life and reduce the cost of use. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0079] Figure 1 4 is a flow chart of a range warning method based on BMS data according to an embodiment of the present invention. DETAILED DESCRIPTION

[0080] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0081] In the description of the present invention, unless otherwise specified, the meaning of "plurality" is two or more. In addition, the terms "first", "second", "third", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0082] According to an embodiment of the present invention, a range warning method and system based on BMS data are provided.

[0083] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to the endurance warning method based on BMS data of an embodiment of the present invention, the endurance warning method based on BMS data includes the following steps:

[0084] S1. Obtain the parameter data of the battery in the BMS, perform preprocessing, and establish an equivalent circuit model of the battery based on the processing results;

[0085] Specifically, the parameter data includes the total voltage, current, state of charge, health status of the battery, and real-time voltage data and real-time temperature data of the single cell.

[0086] S2. Simulate the charge and discharge process of each battery through an equivalent circuit model to generate a corresponding charge and discharge curve;

[0087] S3, using an optimization algorithm to compare the generated charge and discharge curves with the actual charge and discharge curves to evaluate the charge state and health state of each battery;

[0088] S4. Generate balanced charging information of each battery in the BMS based on the evaluation result and the equivalent circuit model;

[0089] S5. Obtain the charging fluctuation change trend according to the balanced charging information, construct a charging fluctuation change chart, perform time series analysis, and determine whether the battery is abnormal according to the analysis results;

[0090] S6. Based on the judgment results, formulate a preset adjustment plan and issue a life warning.

[0091] Preferably, obtaining the parameter data of the battery in the BMS and preprocessing it, and establishing the equivalent circuit model of the battery according to the processing results includes the following steps:

[0092] S11, deploying data acquisition nodes in the network environment, and obtaining the original parameter data of the battery in the BMS through the communication protocol;

[0093] S12, extracting original parameter data and performing denoising processing;

[0094] S13, performing data verification on the original parameter data after denoising, extracting valid data, and performing normalization processing to generate accurate data;

[0095] S14, extracting parameter data features of the battery from accurate data using Hilbert transform;

[0096] S15, determining the components of the battery according to the number, connection mode and model information of the battery, and reading the parameter data characteristics of the battery;

[0097] S16. According to the actual situation of the battery, an equivalent circuit model is established based on the PNGV model;

[0098] Specifically, the PNGV model (Partnership for a New Generation of Vehicles model) is a battery model that is mainly used for battery performance analysis and prediction of electric vehicles. The model can help study and understand the behavior of batteries, including the electrical characteristics, electrochemical reactions, thermal behavior, aging and life of the battery. When establishing a battery circuit model, the PNGV model can provide a theoretical framework and method. The battery circuit model usually includes a power supply (representing the open circuit voltage of the battery), a resistor (representing the internal resistance of the battery), and a capacitor (representing the charge storage capacity of the battery). By appropriately setting the parameters of these components, the voltage and current characteristics of the battery under different charging and discharging conditions can be simulated.

[0099] S17, importing the acquired parameter data of each battery into the circuit model to construct an equivalent circuit model of the battery.

[0100] Specifically, establishing an equivalent circuit model includes the following steps: using professional testing equipment to collect data such as voltage, current, temperature, etc. of the battery under different operating conditions to obtain a large amount of battery test data; preprocessing the collected battery test data, including noise removal, interpolation, filtering, etc., for subsequent modeling; determining the parameters in the equivalent circuit model, including resistance, capacitance, inductance, etc., based on experimental data and the selected equivalent circuit model type; using regression analysis or machine learning methods to fit the battery test data with the equivalent circuit model parameters to obtain the final equivalent circuit model; using new test data to verify and adjust the model to ensure the accuracy and reliability of the model; applying the constructed equivalent circuit model to the battery management system for battery state estimation and prediction, charge and discharge control, etc.

[0101] Preferably, extracting the original parameter data and performing denoising processing comprises the following steps:

[0102] S121, performing wavelet transform on the original parameter data, converting the original parameter data from the time domain to the frequency domain, and screening out wave field records of each frequency band containing interference waves;

[0103] S122, tracking interference waves one by one in the wave field records of each frequency band, and obtaining the interference wave direction at the calculation sample point;

[0104] S123, selecting a number of windows centered on the calculation sample point to perform median filtering, and restoring the interference signal at the calculation sample point;

[0105] S124, recovering the interference signals in the wave field records of each frequency band one by one, and performing inverse wavelet transform to obtain the entire wave field of the interference wave;

[0106] S125 . Subtract the interference wave field from the original parameter data to obtain the effective signal wave field, and generate the denoised original parameter data based on the effective signal wave field.

[0107] Preferably, simulating the charge and discharge process of each battery by an equivalent circuit model to generate a corresponding charge and discharge curve comprises the following steps:

[0108] S21, determining parameter data of an equivalent circuit model, and importing the parameter data of the battery into the equivalent circuit model;

[0109] S22, using an equivalent circuit model to simulate the charging and discharging process of each battery;

[0110] S23. During the simulation process, the parameter data of voltage, current and charge state in the charge and discharge curve of each battery are recorded;

[0111] S24, generating a corresponding charge and discharge curve according to the parameter data of voltage, current and charge state in the charge and discharge curve of each battery.

[0112] Preferably, using an optimization algorithm to compare the generated charge and discharge curves with the actual charge and discharge curves, and evaluating the charge state and health state of each battery includes the following steps:

[0113] S31, initializing the optimization algorithm, setting the initial temperature and the initial charge and discharge curve of the battery;

[0114] S32, repeating steps S33 to S36 according to the set initial temperature;

[0115] S33, randomly perturbing the initial charge and discharge curve to generate a new charge and discharge curve;

[0116] S34, calculating the difference between the fitness value of the new charge and discharge curve and the fitness value of the initial charge and discharge curve;

[0117] S35. If the fitness value of the new charge-discharge curve is higher than the fitness value of the initial charge-discharge curve (i.e., the difference is less than 0), the new charge-discharge curve is accepted as the current charge-discharge curve; otherwise, according to the Metropolis rule, the acceptance probability of the new charge-discharge curve is calculated; if the acceptance probability is greater than a random number, the new charge-discharge curve is accepted as the current charge-discharge curve; otherwise, the current charge-discharge curve is retained;

[0118] S36, if the set termination condition is met, the current charge-discharge curve is output as the optimal charge-discharge curve, and the termination condition is that the new charge-discharge curve is not accepted in several consecutive Metropolis chains or reaches the end temperature, otherwise, the temperature is attenuated by the attenuation function and then the process returns to step S32;

[0119] S37, repeating steps S33 to S36 until the optimal charge and discharge curve is found and compared with the actual charge and discharge curve, and the charge state and health state of each battery are evaluated.

[0120] Specifically, the optimization algorithm is a simulated annealing algorithm (SA). The simulated annealing algorithm is a heuristic optimization algorithm that gradually optimizes the objective function value by simulating the cooling process during solid annealing. In the present invention, the simulated annealing algorithm is used to find the optimal charge and discharge curve for comparison with the actual charge and discharge curve. By setting the initial temperature, decay function and termination condition, the algorithm searches for the best solution in the global search space. The Metropolis rule is used to determine the probability of accepting a new charge and discharge curve, so as to perform a more refined search near the local optimal solution. Finally, the simulated annealing algorithm finds the optimal charge and discharge curve for evaluating the charge state and health state of the battery.

[0121] Specifically, the Metropolis rule is an acceptance-rejection criterion used in Monte Carlo simulation. The basic idea of ​​this rule is: if the new state has lower energy (or higher probability) than the current state, then accept the new state; otherwise, accept the new state with a probability that is the negative exponent of the energy difference (or probability ratio) between the new state and the current state.

[0122] Preferably, the calculation formula of the attenuation function is:

[0123]

[0124] Among them, t s It is represented as the temperature value at the θth step, i.e. the above-mentioned steps S31 to S37;

[0125] t0 represents the initial temperature.

[0126] Preferably, generating the balanced charging information of each battery in the BMS based on the evaluation result and the equivalent circuit model includes the following steps:

[0127] S41, generating battery historical charging process data based on the equivalent circuit model and the evaluation result;

[0128] S42, analyzing the charging status of each battery based on historical charging process data, and generating a balanced charging parameter according to the battery charging status;

[0129] Specifically, the battery charging state is determined by analyzing the battery charging current, voltage, temperature and other data. For example, if the battery charging current gradually decreases and the voltage gradually increases, the battery may be undergoing constant current charging; if the battery charging current remains constant and the voltage gradually increases, the battery may be undergoing constant voltage charging. Then, according to the battery charging state, the balanced charging parameters are generated. For example, if the battery is undergoing constant current charging, a larger charging current can be set; if the battery is undergoing constant voltage charging, a smaller charging current can be set. At the same time, the battery temperature needs to be considered to prevent the battery from overheating.

[0130] S43: verify the balanced charging parameters, and generate balanced charging information according to the verification result.

[0131] Specifically, the generated balanced charging parameters are verified through the actual charging process, that is, the generated parameters are applied to the actual charging process to observe whether the battery charging current, voltage, temperature and other data are in line with expectations. The verification results are analyzed. If the battery charging current, voltage, temperature and other data are in line with expectations, then the generated balanced charging parameters are valid. If they are not in line with expectations, then the balanced charging parameters may need to be adjusted. The balanced charging information is generated according to the verification results. For example, if the verification results show that the generated balanced charging parameters are valid, then the balance is generated.

[0132] Preferably, obtaining the charging fluctuation change trend according to the balanced charging information, constructing a charging fluctuation change chart, performing a time series analysis, and judging whether the battery is abnormal according to the analysis result includes the following steps:

[0133] S51, performing charging fluctuation statistical analysis according to the balanced charging information to obtain a charging fluctuation change trend;

[0134] Specifically, balance charging information is collected, which may include charging current, voltage, temperature and other parameters. Then, the collected data is processed to calculate the fluctuations of current, voltage, temperature and other parameters for each charge. The processed data is statistically analyzed to calculate the mean value, standard deviation and other statistics of current, voltage, temperature and other parameters. According to the results of the statistical analysis, the changing trend of charging fluctuations is analyzed by regression analysis method to predict future charging fluctuations.

[0135] S52, extracting change features based on the charging fluctuation change trend, and constructing a charging fluctuation change chart according to the extracted feature results;

[0136] Specifically, the current, voltage, time and other information of the charging fluctuations are collected, and then the sliding mean and sliding standard deviation are calculated to extract the frequency characteristics of the charging fluctuations. The extracted characteristics are then used to construct a charging fluctuation change chart, draw a time series graph of the charging fluctuations, or draw a box plot or violin plot of the charging fluctuations.

[0137] S53, marking the charging fluctuation change chart with information, and presetting the charging fluctuation safety threshold;

[0138] S54, extracting battery fluctuation parameters from the charging fluctuation change chart, and performing spectrum analysis on the battery fluctuation parameters to obtain a change pattern and trend of the battery fluctuation parameters;

[0139] Specifically, the time domain signal of the battery fluctuation parameter is converted into a frequency domain signal, and the battery fluctuation parameter is analyzed by a spectrum analysis method.

[0140] Specifically, the time domain signal of the battery fluctuation parameters is collected, and the time domain signal is converted into a frequency domain signal using Fourier transform, and the signal is decomposed into a series of sine waves and cosine waves through Fourier transform. After the frequency, amplitude and phase of each wave are determined to obtain the frequency domain signal, its amplitude spectrum and phase spectrum are calculated, and then the amplitude spectrum and phase spectrum are analyzed to understand the frequency characteristics of the battery fluctuation parameters.

[0141] Specifically, the fluctuation parameters of the battery are collected, the time domain signal is converted into a frequency domain signal through Fourier transform, and the frequency component of the signal is analyzed, the amplitude spectrum and the phase spectrum are calculated, and then the amplitude spectrum and the phase spectrum are analyzed. The frequency corresponding to the peak value in the amplitude spectrum is the main frequency component, and the height of the peak value is the amplitude of the corresponding frequency component.

[0142] Specifically, the frequency components and frequency amplitudes are brought into the time series analysis model for calculation to generate the change pattern and trend of the battery fluctuation parameters.

[0143] S55, extracting change characteristics according to the change mode and change trend of the battery fluctuation parameter to obtain change characteristic data;

[0144] Specifically, understand the changing patterns and trends of battery fluctuation parameters, including viewing statistical summaries of the data, such as mean, median, standard deviation, etc., and drawing graphs, such as time series graphs, autocorrelation graphs, and partial autocorrelation graphs. When the data has obvious periodic changes, the frequency and amplitude can be extracted; if the data has an obvious trend, the slope and intercept of the trend can be extracted; apply the selected machine learning algorithm to the data of battery fluctuation parameters. After the feature extraction is completed, the result is the number of changing features.

[0145] S56. Analyze the change characteristic data based on the abnormality analysis algorithm, and determine whether the battery has an abnormality based on the analysis result.

[0146] Preferably, analyzing the change characteristic data based on the abnormality analysis algorithm and judging whether the battery has an abnormality according to the analysis result includes the following steps:

[0147] S561, setting preset parameters, and generating battery change characteristic data and a corresponding trend matrix according to the preset parameters;

[0148] S562, calculating the fitness function value of the change feature data, where the fitness function is used to measure the quality of the change feature data;

[0149] S563, calculate the population fitness variance of all the change feature data, and when the preset conditions are met, go to step S565 if a mutation occurs, otherwise go to step S564;

[0150] S564, updating the inertia weight point according to the formula for calculating the inertia weight, and updating the speed and position of the change characteristic data according to the speed update formula and the position update formula of the change characteristic data respectively;

[0151] S565, check whether the end condition is met, if so, end the optimization search, otherwise go to S562 to continue the optimization search;

[0152] S566, assigning the found optimal parameter vector to the SVM model;

[0153] S567, inputting the processed sample data into the SVM model, and training the SVM to obtain a trained SVM classification model;

[0154] S568, using the trained SVM classification model to predict the prediction sample, obtain the final analysis result, and stop the calculation;

[0155] S569. Determine whether there is any abnormality in the battery based on the final analysis result.

[0156] Specifically, the anomaly analysis algorithm is an adaptive particle swarm optimization support vector machine algorithm, which is a machine learning algorithm that integrates particle swarm optimization (PSO) and support vector machine (SVM). Its main ideas are:

[0157] The particle swarm optimization algorithm optimizes the parameters of SVM to find the optimal value;

[0158] Apply the optimized SVM parameters to the SVM model;

[0159] Use the optimized SVM model for training and prediction.

[0160] According to another embodiment of the present invention, a battery life warning system based on BMS data is also provided, and the battery life warning system based on BMS data includes: a data processing and model building module, a battery charge and discharge simulation and curve generation module, a model evaluation and battery status analysis module, a balance charging information generation module, a timing analysis and abnormality judgment module, and a decision support and warning module;

[0161] The data processing and model building module is used to obtain the parameter data of the battery in the BMS, perform preprocessing, and build an equivalent circuit model of the battery based on the processing results;

[0162] The battery charge and discharge simulation and curve generation module is used to simulate the charge and discharge process of each battery through an equivalent circuit model and generate the corresponding charge and discharge curve;

[0163] Model evaluation and battery status analysis module, which is used to compare the generated charge and discharge curves with the actual charge and discharge curves using an optimization algorithm to evaluate the charge state and health state of each battery;

[0164] A balanced charging information generation module is used to generate balanced charging information of each battery in the BMS according to the evaluation results and based on the equivalent circuit model;

[0165] The timing analysis and abnormality judgment module is used to obtain the charging fluctuation change trend based on the balanced charging information, build a charging fluctuation change chart, perform timing analysis, and judge whether the battery is abnormal based on the analysis results;

[0166] The decision support and early warning module is used to formulate preset adjustment plans and issue endurance warnings based on the judgment results.

[0167] In summary, with the help of the above technical solutions of the present invention, the present invention obtains the battery parameter data in the BMS, establishes an equivalent circuit model and simulates the charging and discharging process, so as to more accurately understand the charge state and health state of the battery, and according to the balanced charging information generated by the evaluation results, the charging strategy can be optimized, thereby improving the performance and service life of the battery. The present invention can find possible abnormalities in the battery through the time series analysis of the charging fluctuation change trend, and timely discovering the abnormality helps to take measures in advance to prevent battery failure. At the same time, by analyzing the charging fluctuation change chart, the problems of the battery can be more accurately located, and by real-time monitoring and analysis of the battery charging fluctuation, the remaining battery life can be more accurately predicted, thereby helping users to make relevant preparations in advance, reasonably plan driving routes or arrange charging. The present invention can more accurately evaluate the SOC and SOH of the battery by comparing the charging and discharging curves through the optimization algorithm, thereby providing a reliable basis for further charging strategy optimization and abnormality detection, and evaluating the battery health status helps to discover the problem of battery performance degradation, thereby taking adjustment of the charging strategy or maintenance measures to extend the battery life and reduce the cost of use.

[0168] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. The endurance warning method based on BMS data is characterized by: The endurance warning method based on BMS data includes the following steps: S1. Obtain the parameter data of the battery in the BMS, perform preprocessing, and establish an equivalent circuit model of the battery based on the processing results; S2. Simulate the charge and discharge process of each battery through an equivalent circuit model to generate a corresponding charge and discharge curve; S3, using an optimization algorithm to compare the generated charge and discharge curves with the actual charge and discharge curves to evaluate the charge state and health state of each battery; S4. Generate balanced charging information of each battery in the BMS based on the evaluation result and the equivalent circuit model; S5. Obtain the charging fluctuation change trend according to the balanced charging information, construct a charging fluctuation change chart, perform time series analysis, and determine whether the battery is abnormal according to the analysis results; S6. According to the judgment results, formulate a preset adjustment plan and issue a life warning; The method of simulating the charge and discharge process of each battery by using an equivalent circuit model to generate a corresponding charge and discharge curve comprises the following steps: S21, determining parameter data of an equivalent circuit model, and importing the parameter data of the battery into the equivalent circuit model; S22, using an equivalent circuit model to simulate the charging and discharging process of each battery; S23. During the simulation process, the parameter data of voltage, current and charge state in the charge and discharge curve of each battery are recorded; S24, generating a corresponding charge and discharge curve according to the parameter data of voltage, current, and charge state in the charge and discharge curve of each battery; comparing the generated charge and discharge curve with the actual charge and discharge curve by using the optimization algorithm to evaluate the charge state and health state of each battery includes the following steps: S31, initializing the optimization algorithm, setting the initial temperature and the initial charge and discharge curve of the battery; S32, repeating steps S33 to S36 according to the set initial temperature; S33, randomly perturbing the initial charge and discharge curve to generate a new charge and discharge curve; S34, calculating the difference between the fitness value of the new charge and discharge curve and the fitness value of the initial charge and discharge curve; S35. If the fitness value of the new charge-discharge curve is higher than the fitness value of the initial charge-discharge curve, the new charge-discharge curve is accepted as the current charge-discharge curve. Otherwise, according to the Metropolis rule, the acceptance probability of the new charge-discharge curve is calculated. If the acceptance probability is greater than a random number, the new charge-discharge curve is accepted as the current charge-discharge curve. Otherwise, the current charge-discharge curve is retained. S36, if the set termination condition is met, the current charge-discharge curve is output as the optimal charge-discharge curve, and the termination condition is that the new charge-discharge curve is not accepted in several consecutive Metropolis chains or reaches the end temperature, otherwise, the temperature is attenuated by the attenuation function and then the process returns to step S32; S37, repeating steps S33 to S36 until an optimal charge-discharge curve is found and compared with the actual charge-discharge curve, and the charge state and health state of each battery are evaluated; The calculation formula of the attenuation function is: Among them, t s It is expressed as the temperature value at the θth step; t0 represents the initial temperature; The step of generating the balanced charging information of each battery in the BMS according to the evaluation result and based on the equivalent circuit model includes the following steps: S41, generating battery historical charging process data based on the equivalent circuit model and the evaluation result; S42, analyzing the charging status of each battery based on historical charging process data, and generating a balanced charging parameter according to the battery charging status; S43, verifying the balanced charging parameters, and generating balanced charging information according to the verification results; The method of obtaining the charging fluctuation change trend according to the balanced charging information, constructing a charging fluctuation change chart, performing a time series analysis, and judging whether the battery is abnormal according to the analysis result includes the following steps: S51, performing charging fluctuation statistical analysis according to the balanced charging information to obtain a charging fluctuation change trend; S52, extracting change features based on the charging fluctuation change trend, and constructing a charging fluctuation change chart according to the extracted feature results; S53, marking the charging fluctuation change chart with information, and presetting the charging fluctuation safety threshold; S54, extracting battery fluctuation parameters from the charging fluctuation change chart, and performing spectrum analysis on the battery fluctuation parameters to obtain a change pattern and trend of the battery fluctuation parameters; S55, extracting change characteristics according to the change mode and change trend of the battery fluctuation parameter to obtain change characteristic data; S56, analyzing the change characteristic data based on an abnormality analysis algorithm, and determining whether the battery has an abnormality according to the analysis result; The method of analyzing the change characteristic data based on the abnormality analysis algorithm and judging whether the battery has an abnormality according to the analysis result includes the following steps: S561, setting preset parameters, and generating battery change characteristic data and a corresponding trend matrix according to the preset parameters; S562, calculating the fitness function value of the change feature data, where the fitness function is used to measure the quality of the change feature data; S563, calculate the population fitness variance of all the change feature data, and when the preset conditions are met, go to step S565 if a mutation occurs, otherwise go to step S564; S564, updating the inertia weight point according to the formula for calculating the inertia weight, and updating the speed and position of the change characteristic data according to the speed update formula and the position update formula of the change characteristic data respectively; S565, check whether the end condition is met, if so, end the optimization search, otherwise go to S562 to continue the optimization search; S566, assigning the found optimal parameter vector to the SVM model; S567, inputting the processed sample data into the SVM model, and training the SVM to obtain a trained SVM classification model; S568, using the trained SVM classification model to predict the prediction sample, obtain the final analysis result, and stop the calculation; S569. Determine whether there is any abnormality in the battery based on the final analysis result.

2. The endurance warning method based on BMS data according to claim 1 is characterized in that: The method of obtaining the parameter data of the battery in the BMS, performing preprocessing, and establishing an equivalent circuit model of the battery according to the processing results includes the following steps: S11, deploying data acquisition nodes in the network environment, and obtaining the original parameter data of the battery in the BMS through the communication protocol; S12, extracting original parameter data and performing denoising processing; S13, performing data verification on the original parameter data after denoising, extracting valid data, and performing normalization processing to generate accurate data; S14, extracting parameter data features of the battery from accurate data using Hilbert transform; S15, determining the components of the battery according to the number, connection mode and model information of the battery, and reading the parameter data characteristics of the battery; S16. According to the actual situation of the battery, an equivalent circuit model is established based on the PNGV model; S17, importing the acquired parameter data of each battery into the circuit model to construct an equivalent circuit model of the battery.

3. The endurance warning method based on BMS data according to claim 2 is characterized in that: The extraction of original parameter data and the denoising process include the following steps: S121, performing wavelet transform on the original parameter data, converting the original parameter data from the time domain to the frequency domain, and screening out the wave field records of each frequency band containing the interference wave; S122, tracking interference waves one by one in the wave field records of each frequency band, and obtaining the interference wave direction at the calculation sample point; S123, selecting a number of windows centered on the calculation sample point to perform median filtering, and restoring the interference signal at the calculation sample point; S124, recovering the interference signals in the wave field records of each frequency band one by one, and performing inverse wavelet transform to obtain the entire wave field of the interference wave; S125 . Subtract the interference wave field from the original parameter data to obtain the effective signal wave field, and generate the denoised original parameter data based on the effective signal wave field.

4. A battery life warning system based on BMS data, used to implement the battery life warning method based on BMS data according to any one of claims 1 to 3, characterized in that: The battery life warning system based on BMS data includes: data processing and model building module, battery charge and discharge simulation and curve generation module, model evaluation and battery status analysis module, balance charging information generation module, timing analysis and abnormality judgment module and decision support and warning module; The data processing and model building module is used to obtain the parameter data of the battery in the BMS, perform preprocessing, and establish an equivalent circuit model of the battery according to the processing results; The battery charge and discharge simulation and curve generation module is used to simulate the charge and discharge process of each battery through an equivalent circuit model and generate a corresponding charge and discharge curve; The model evaluation and battery status analysis module is used to compare the generated charge and discharge curves with the actual charge and discharge curves using an optimization algorithm to evaluate the charge state and health state of each battery; The balanced charging information generating module is used to generate balanced charging information of each battery in the BMS according to the evaluation result and based on the equivalent circuit model; The timing analysis and abnormality judgment module is used to obtain the charging fluctuation change trend according to the balanced charging information, construct a charging fluctuation change chart, perform timing analysis, and judge whether the battery has abnormality according to the analysis results; The decision support and warning module is used to formulate a preset adjustment plan based on the judgment result and to issue a life warning.

Images