Method and system for extracting influence factors of service life of energy storage battery

Abstract

The invention discloses an energy storage battery life influence factor extraction method and system, and the method comprises the steps: obtaining first target data, and carrying out the first preprocessing of the first target data, and obtaining second target data; establishing a first feature parameter set, and performing first analysis in combination with the second target data to obtain a second feature parameter set; performing second analysis on the second feature parameter set to obtain a third feature data set; and establishing a first extraction model, and performing influence factor extraction according to the first extraction model. The service life of the lithium ion energy storage battery can be effectively prolonged, the maintenance cost is reduced, safe and stable operation of an energy storage system is guaranteed, and the method has important practical significance for promoting wide application of renewable energy sources and development of an energy storage technology.

Description

Technical Field

[0001] The present invention relates to the technical field of extracting factors influencing the life of an energy storage battery, and in particular to a method and system for extracting factors influencing the life of an energy storage battery. Background Art

[0002] With the widespread application of renewable energy and the growing demand for energy storage, lithium-ion energy storage batteries have been widely used due to their high energy density, long cycle life and other advantages. However, in actual use, the life of lithium-ion energy storage batteries will be affected by a variety of complex factors. Accurately extracting these influencing factors is extremely important for optimizing the design of lithium-ion energy storage batteries, improving their service life and ensuring the safe and stable operation of energy storage systems.

[0003] Currently, in the analysis of factors affecting the life of lithium-ion energy storage batteries, the life of lithium-ion energy storage batteries is affected by a combination of complex factors. These factors cover multiple dimensions of the battery, from static initial material properties and electrode structure to various real-time operating parameters during the dynamic charging and discharging process.

[0004] Existing analysis methods often only focus on some factors and fail to fully consider these diverse influencing factors, resulting in incomplete extraction of factors affecting battery life and difficulty in accurately grasping the actual changes in battery life. In addition, during the construction and analysis of characteristic parameters, either the constructed characteristic parameters cannot fully reflect the intrinsic connection between battery operating status and life from multiple angles, or the methods used to analyze the correlation between these characteristic parameters and battery life and further explore key influencing factors are not efficient and accurate enough, making it difficult to extract the factors that really play a key role in battery life from many complex parameters, which is not conducive to effective life prediction, performance optimization, and reasonable use and maintenance of lithium-ion energy storage batteries. Summary of the invention

[0005] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0006] In view of the above existing problems, the present invention is proposed.

[0007] Therefore, the present invention provides a method and system for extracting factors affecting the life of an energy storage battery, which can solve the problems mentioned in the background technology.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0009] In a first aspect, the present invention provides a method for extracting factors affecting the life of an energy storage battery, comprising:

[0010] Acquire first target data, and perform first preprocessing on the first target data to obtain second target data;

[0011] Establishing a first feature parameter set, and performing a first analysis in combination with the second target data to obtain a second feature parameter set;

[0012] Performing a second analysis on the second feature parameter set to obtain a third feature data set;

[0013] A first extraction model is established, and influencing factors are extracted according to the first extraction model.

[0014] As a preferred solution of the method for extracting factors affecting the life of energy storage batteries described in the present invention, the first extraction model includes:

[0015] The first extraction model is any model that takes the second target data as input and outputs several features in the third feature data set or can directly or indirectly obtain relevant parameters of several features in the third feature data set.

[0016] As a preferred solution of the method for extracting factors affecting the life of an energy storage battery described in the present invention, the first feature parameter set includes at least a plurality of static feature parameters and a plurality of dynamic feature parameters.

[0017] As a preferred solution of the method for extracting factors affecting the life of an energy storage battery according to the present invention, the step of establishing a first characteristic parameter set and performing a first analysis in combination with the second target data includes:

[0018] Presetting a first correlation algorithm and a first correlation threshold;

[0019] Solving the correlation of the first feature parameter set based on the second target data;

[0020] Comparing the solution result with the first correlation threshold;

[0021] The feature parameter set that meets the comparison result is recorded as the second feature parameter set.

[0022] As a preferred solution of the method for extracting factors affecting the life of an energy storage battery according to the present invention, the second analysis of the second characteristic parameter set includes:

[0023] Preset the first principal component analysis algorithm;

[0024] A second analysis is performed on the second feature parameter set according to the first principal component analysis algorithm.

[0025] As a preferred solution of the method for extracting factors affecting the life of an energy storage battery described in the present invention, wherein: the first feature parameter set includes the second feature parameter set, and the second feature parameter set includes the third feature parameter set.

[0026] As a preferred solution of the method for extracting factors affecting the life of an energy storage battery according to the present invention, the first preprocessing of the first target data includes:

[0027] Cleaning the first target data and identifying outliers;

[0028] performing outlier processing on the first target data after outlier identification;

[0029] The first target data after outlier processing is normalized to obtain second target data.

[0030] In a second aspect, the present invention provides a system for extracting factors affecting the life of an energy storage battery, comprising:

[0031] A data processing module, used for acquiring first target data, and performing first preprocessing on the first target data to obtain second target data;

[0032] A first analysis module, used for establishing a first characteristic parameter set, and performing a first analysis in combination with the second target data to obtain a second characteristic parameter set;

[0033] A second analysis module, used for performing a second analysis on the second feature parameter set to obtain a third feature data set;

[0034] The model building module is used to build a first extraction model and extract influencing factors according to the first extraction model.

[0035] In a third aspect, the present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned method when executing the computer program.

[0036] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the method described above when executed by a processor.

[0037] Compared with the prior art, the invention has the following beneficial effects: the invention proposes a method and system for extracting factors affecting the life of energy storage batteries, obtains first target data, performs first preprocessing on the first target data, obtains second target data; establishes a first feature parameter set, performs a first analysis in combination with the second target data, obtains a second feature parameter set; performs a second analysis on the second feature parameter set, obtains a third feature data set; establishes a first extraction model, and extracts influencing factors according to the first extraction model. The invention can effectively improve the service life of lithium-ion energy storage batteries, reduce maintenance costs, and ensure the safe and stable operation of energy storage systems, which has important practical significance for promoting the widespread application of renewable energy and the development of energy storage technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. 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 creative labor. Among them:

[0039] Figure 1 A method flow chart of a method and system for extracting factors affecting the life of an energy storage battery provided by an embodiment of the present invention;

[0040] Figure 2 An internal structural diagram of a computer device of a method and system for extracting factors affecting the life of an energy storage battery provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0041] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0042] Example 1

[0043] Reference Figure 1-Figure 2 , which is the first embodiment of the present invention, and provides a method and system for extracting factors affecting the life of an energy storage battery, including:

[0044] There are some problems in the existing related technologies, such as insufficient accuracy in battery life prediction and high complexity in the analysis of influencing factors. These problems limit the performance of battery management systems in practical applications and make it impossible to effectively guide battery maintenance and life extension.

[0045] The present application provides a method that can effectively solve the above-mentioned problems. Next, a plurality of embodiments will be combined to explain in detail how to implement the method for extracting factors affecting the life of the energy storage battery.

[0046] Figure 1 A method flow chart of a method and system for extracting factors affecting the life of an energy storage battery is shown, including:

[0047] S101, acquiring first target data, and performing first preprocessing on the first target data to obtain second target data;

[0048] In an optional embodiment, the first target data may be real-time operating data such as voltage, current, and temperature of the battery, as well as static data such as initial material composition and electrode structure parameters of the battery, which relevant technicians may obtain according to actual needs;

[0049] In an optional embodiment, the first target data may be acquired by a variety of sensors that can monitor the operating status of the battery in real time. For example, a voltage sensor is used to monitor the voltage change of the battery, a current sensor is used to monitor the charge and discharge current of the battery, and a temperature sensor is used to monitor the surface temperature of the battery.

[0050] In an optional embodiment, the first target data can be obtained by directly obtaining the static data from the battery manufacturing records through a data acquisition system, or by obtaining the initial material composition and electrode structure parameters of the battery through laboratory tests. After preprocessing, these data can be used for subsequent analysis and processing to extract the key factors affecting the battery life.

[0051] In the embodiment of the present application, the first preprocessing of the first target data includes:

[0052] Cleaning the first target data and identifying outliers;

[0053] performing outlier processing on the first target data after outlier identification;

[0054] The first target data after outlier processing is normalized to obtain second target data.

[0055] In an optional embodiment, outlier identification may be performed using a statistical method, such as a threshold judgment based on the standard deviation, to determine whether a data point is an outlier. Specifically, a threshold may be set, such as 2 times the standard deviation, and data points exceeding this range may be identified as outliers.

[0056] In an optional embodiment, a box plot analysis method may also be used to identify outliers by calculating quartiles.

[0057] In an optional embodiment, outlier processing may adopt a variety of strategies, such as replacing outliers with means, medians, or modes, or directly deleting these data points.

[0058] In an optional embodiment, the normalization process may use methods such as minimum-maximum normalization or Z-score normalization to ensure that the data is analyzed on a uniform scale, thereby improving the accuracy and efficiency of subsequent models.

[0059] In the embodiment of the present application, the specific steps of acquiring the first target data and performing the first preprocessing on the first target data to obtain the second target data may be as follows:

[0060] During the entire life cycle of lithium-ion energy storage batteries, high-precision sensors are used to collect various types of data, including real-time operating data such as battery voltage, current, and temperature, as well as static data such as the battery's initial material composition and electrode structure parameters. The collection frequency is set according to the battery's application scenario and actual needs to ensure that the battery's state changes under different operating conditions can be captured.

[0061] In an optional embodiment, a suitable high-precision sensor is selected. A voltage sensor with a resolution of 0.001V is selected to accurately measure the terminal voltage of the lithium-ion energy storage battery. The accuracy of the current sensor should be within ±0.1% to accurately measure the charging and discharging current. High-precision thermistor or thermocouple temperature sensor, the temperature measurement accuracy must reach ±0.5°C. The temperature sensors are evenly distributed on the battery surface or in key internal locations (such as near the electrodes) to fully monitor the temperature changes of the battery.

[0062] In the embodiment of the present application, a data acquisition card with a multi-channel synchronous acquisition function is selected, and the sampling rate should be determined according to the application scenario of the battery and the frequency of state changes that need to be captured. For general energy storage battery applications, the sampling rate can be set between 1Hz-10Hz. If it is necessary to analyze the transient characteristics of the battery during rapid charging and discharging, the sampling rate can be increased to 100Hz or even higher. The data acquisition card must be able to make a good electrical connection with the sensor and transmit the collected data to a computer or storage device for subsequent processing.

[0063] In the embodiment of the present application, the voltage sensor is connected to the positive and negative terminals of the battery, and the terminal voltage data of the battery is continuously collected at a set sampling frequency through a data acquisition card, and the acquisition timestamp is recorded. In a complete charging process, from the beginning to the end of charging, the voltage data is collected at a sampling rate of 1Hz, and a series of voltage value sequences V(t) that change with time will be obtained, where t represents the acquisition time point.

[0064] In the embodiment of the present application, the current sensor is connected in series in the charging and discharging circuit of the battery, and also collects current data according to the set sampling frequency. When charging, the current is positive, and when discharging, it is negative. s The collected current data sequence is recorded as I(t), and its unit is ampere (A).

[0065] In the embodiment of the present application, multiple temperature sensors are distributed at different positions on the surface or inside of the battery, and temperature data of each point is collected at the same time. These temperature data can reflect the heating conditions of different parts of the battery and the overall temperature distribution characteristics.

[0066] In the embodiments of the present application, the initial material composition collection: the detailed composition information of the positive and negative electrode materials of the battery is obtained from the battery manufacturer, including the type and content percentage of the active material (such as lithium cobalt oxide, lithium iron phosphate, etc.) in the positive electrode material, and the relevant information of the negative electrode material (such as graphite, etc.). These data are usually recorded in the form of mass percentage or molar percentage.

[0067] In the embodiments of the present application, electrode structure parameters are collected. Parameters such as the thickness and area of ​​the electrode are measured or obtained. The electrode thickness can be measured using a high-precision caliper or optical measuring instrument in millimeters (mm). The electrode area can be calculated based on the geometric shape and design size of the battery in square centimeters (cm 2 ). Record the relevant parameters of the diaphragm, such as diaphragm thickness, porosity, etc.

[0068] In an embodiment of the present application, the collected real-time operation data and static data are stored in a certain format.

[0069] In an optional embodiment, the collected data is cleaned to remove outliers and noise data, and the data is normalized to convert data of different dimensions to the same order of magnitude for subsequent analysis and calculation.

[0070] In an optional embodiment, outliers are identified for the data. Taking the current parameter as an example, the other parameters are treated in the same way. The mean (I) and standard deviation (σ) of the collected current data sequence I(t) are calculated. I ). The mean calculation formula is: Where n is the total number of data points; the standard deviation is calculated as: Data points that are outside the range of ±3 times the standard deviation of the mean are considered outliers. <I_-3*σ I Or I(i)>I_+3*σ I , then I(i) is judged to be an outlier.

[0071] In an optional embodiment, the identified outliers can be processed using a moving average smoothing method. Select an appropriate window size (m);

[0072] In the embodiment of the present application, the window size is determined according to the fluctuation frequency of the data and the degree of smoothing required, and is generally an odd number, and 5 is selected.

[0073] In an optional embodiment, the data sequence is smoothed: for the data sequence X(t), the smoothed data sequence Y(t) is calculated as follows:

[0074] When t<(m-1) / 2 (i.e. at the beginning of the data sequence, the window is not completely covered),

[0075] When (m-1) / 2≤t≤n-(m-1) / 2 (the middle part of the data sequence, the window covers the entire data sequence),

[0076]

[0077] When t>n-(m-1) / 2 (the end of the data sequence, the window is not completely covered),

[0078] In an optional embodiment, Z-score normalization (standardization normalization) is performed. The mean (X_) and standard deviation (σ) of the data series are calculated. X ), the mean calculation formula is: The formula for calculating standard deviation is: Where n is the total number of data points. Each data point is normalized and the calculation formula of the normalized data sequence X′(t) is: The normalized data will have the characteristics of a mean of 0 and a standard deviation of 1, achieving the purpose of standardizing data of different dimensions to ensure the quality of the data and lay a good foundation for accurately extracting the factors affecting the life of lithium-ion energy storage batteries.

[0079] It should be noted that obtaining the first target data and performing the first preprocessing on the first target data to obtain the second target data can effectively reduce the noise and outliers in the data and improve the accuracy and reliability of the data. Through the preprocessing step, it can be ensured that the subsequent analysis and model training will not be interfered by the irregular factors in the original data, thereby improving the accuracy of the life prediction of the lithium-ion energy storage battery. In addition, the preprocessed data is more suitable for feature extraction and pattern recognition, which is crucial for identifying and analyzing the key factors affecting the battery life.

[0080] S102, establishing a first feature parameter set, and performing a first analysis in combination with the second target data to obtain a second feature parameter set;

[0081] In the embodiment of the present application, the first feature parameter set includes at least a number of static feature parameters and a number of dynamic feature parameters.

[0082] In an optional embodiment, the first characteristic parameter set may include material structure characteristic parameters, battery design parameters, charging rate related parameters, discharge rate related parameters, charge and discharge depth related parameters, Coulomb efficiency related parameters, internal resistance change rate related parameters, voltage characteristic parameters, current integral parameters, temperature change parameters, energy characteristic parameters, time characteristic parameters, power characteristic parameters, polarization related parameters, electrode potential parameters and electrolyte related parameters.

[0083] Specifically, the material structure characteristic parameters include:

[0084] (1) Specific surface area of ​​positive electrode material (S positive ): surface area per unit mass of positive electrode material, measured by BET method.

[0085] (2) Specific surface area of ​​negative electrode material (S negative ): Specific surface area parameter of negative electrode material, measured by BET method.

[0086] (3) Active substance content active ): Calculate the mass proportion of active materials in the positive electrode and the negative electrode respectively.

[0087] Furthermore, battery design parameters include:

[0088] (1) Electrode thickness electrode ): Includes the thickness parameters of the positive and negative electrodes.

[0089] (2) Porosity separator ): The porosity of the separator affects the transfer rate of ions between the positive and negative electrodes and the pressure balance inside the battery.

[0090] Furthermore, the charging rate related parameters include:

[0091] (1) Average charge rate (C avgcharge ): The average value of the ratio of charging current to battery rated capacity during the entire charging process. The calculation formula is: in is the charging current at the i-th sampling moment, n is the number of sampling points during the charging process, C rated is the rated capacity of the battery.

[0092] (2) Maximum charge rate (C maxcharge ): The ratio of the maximum charging current to the rated capacity during the charging process, that is, in It is the maximum charging current during the charging process.

[0093] (3) Charging rate standard deviation (C stdchange ): It is used to measure the fluctuation degree of charging rate. The calculation formula is:

[0094]

[0095] Furthermore, the discharge rate related parameters include:

[0096] (1) Average discharge rate (C avgdischarge ): During the discharge process, in is the discharge current at the i-th sampling moment, and m is the number of sampling points during the discharge process.

[0097] (2) Maximum discharge rate (C maxdischarge ): The ratio of the maximum discharge current to the rated capacity during the discharge process,

[0098] (3) Discharge rate standard deviation (C stddischarge ): Calculated in the same way as the charge rate standard deviation, reflecting the fluctuation of the discharge rate.

[0099] Furthermore, the parameters related to charge and discharge depth include:

[0100] (1) Average charge and discharge depth (DOD) avg ): The average value of each charge and discharge depth in multiple charge and discharge cycles. The calculation formula is: in and are the depths of the jth charge and discharge, respectively, and k is the number of charge and discharge cycles.

[0101] (2) Maximum charge and discharge depth (DOD) max ): The maximum charge and discharge depth value that occurs in all charge and discharge cycles, that is:

[0102]

[0103] Furthermore, the parameters related to Coulombic efficiency include:

[0104] (1) Average Coulombic efficiency (CE avg ): The average value of Coulombic efficiency over multiple charge and discharge cycles, Among them, CE lis the Coulombic efficiency (ratio of discharge capacity to charge capacity) of the lth charge and discharge cycle, and p is the number of cycles.

[0105] (2) Coulomb efficiency standard deviation (CE std ): Measures the degree of fluctuation of Coulombic efficiency between different cycles,

[0106]

[0107] Furthermore, the parameters related to the internal resistance change rate include:

[0108] (1) Average change rate of internal resistance (IRR avg ): During the use of the battery, the internal resistance of the battery is measured regularly, and the average rate of change of the internal resistance over time or the number of cycles is calculated. The internal resistance measured at q time points or cycle points is R 1 ,R 2 ,…,R q ,but

[0109] (2) Standard deviation of internal resistance change rate (IRR std ): reflects the stability of internal resistance changes,

[0110] Furthermore, the voltage characteristic parameters include:

[0111] (1) Average charging voltage (V avgcharge ): The average value of the battery terminal voltage during the entire charging process. The calculation formula is: in is the charging voltage at the i-th sampling moment, and n is the number of sampling points during the charging process.

[0112] (2) Average discharge voltage (V avgdischarge ): During the discharge process, in is the discharge voltage at the i-th sampling moment, and m is the number of sampling points during the discharge process.

[0113] (3) Charging voltage standard deviation (V stdcharge ): Measures the fluctuation of charging voltage, and its calculation formula is:

[0114] (4) Discharge voltage standard deviation (V stddischarge ): used to evaluate the stability of the discharge voltage,

[0115]

[0116] Furthermore, the current integration parameters include:

[0117] (1) Cumulative charge capacity (Q chargeaccum ): The total amount of electricity that has passed through the battery during the charging process from the time the battery was first used to the current moment. This is obtained by integrating the charging current over time, that is, Among them I charge (t) is the charging current as a function of time, and t is the current time.

[0118] (2) Accumulated discharge capacity (Q_discharge_accum): Similar to the accumulated charge capacity, it is the total amount of electricity that passes through the battery during the discharge process. Among them I discharge (t) is the discharge current as a function of time.

[0119] (3) Net power (Q_net): The difference between the cumulative charge power and the cumulative discharge power, Q net =Q chargeaccum -Q discharge_accum .

[0120] Furthermore, the temperature variation parameters include:

[0121] (1) Temperature range (ΔT): The difference between the highest and lowest temperatures experienced during battery operation, i.e., ΔT = T max -T min , where T max is the maximum temperature, T min is the minimum temperature.

[0122] Furthermore, energy characteristic parameters include:

[0123] (1) Average charging energy (E avgcharge ): The average energy input into the battery per unit time during the charging process, obtained by multiplying the average charging voltage by the average charging current, E avgcharge =V avgcharge ×C avgcharge ×C rated .

[0124] (2) Average discharge energy (E avgdischarge ): is the average energy output from the battery per unit time during the discharge process, E avgdischarge =V avgdischarge ×C avgdischarge ×C rated .

[0125] (3) Charging energy standard deviation (E stdcharge ): Measures the stability of energy input during charging, calculated as:

[0126] (4) Discharge energy standard deviation (E stddischarge ): Evaluate the stability of discharge energy output,

[0127] Furthermore, the time characteristic parameters include:

[0128] (1) Average charging duration (T avgcharge ): The average value of charging duration during multiple charging processes. The calculation formula is: in is the duration of the jth charge, and k is the number of charges.

[0129] (2) Mean discharge duration (T avgdischarge ): is the average value of multiple discharge durations,

[0130]

[0131] (3) Standing time ratio (R rest ): The proportion of the rest time (time when neither charging nor discharging) in the total time during the entire battery life. The calculation formula is: where t rest is the resting time, t total is the total usage time.

[0132] Furthermore, the power characteristic parameters include:

[0133] (1) Average charging power (P avgcharge ): Average power during charging, P avgcharge =E avgcharge / T avgcharge .

[0134] (2) Average discharge power (P avgdischarge ): average power during discharge, P avgdischarge =E avgdischarge / T avgdischarge .

[0135] (3) Standard deviation of power change rate Calculate the standard deviation of the charge or discharge power rate of change, which is the change in power per unit time.

[0136] Furthermore, polarization related parameters include:

[0137] (1) Ohmic polarization resistance (R ohmic ): Resistance value measured in the high frequency band by electrochemical impedance spectroscopy (EIS) test.

[0138] (2) Charge transfer polarization resistance (R ct): The resistance corresponding to the mid-frequency region of the EIS test.

[0139] (3) Concentration polarization resistance (R con ): Calculated from EIS low-frequency data.

[0140] Furthermore, the electrode potential parameters include:

[0141] (1) Average positive electrode potential (E posavg ): The average value of the positive electrode potential during the battery charge and discharge cycle.

[0142] (2) Average potential of negative electrode (E negavg ): The average potential of the negative electrode reflects the electrochemical state of the negative electrode material during the charging and discharging process.

[0143] (3) Standard deviation of the positive and negative electrode potential difference Calculate the potential difference between the positive and negative electrodes (E diff =E pos -E neg ) during the charge and discharge process.

[0144] Furthermore, electrolyte related parameters include:

[0145] (1) Conductivity of electrolyte Directly affects the transfer rate of lithium ions in the electrolyte.

[0146] (2) Electrolyte lithium ion concentration (C_Li): The concentration of lithium ions in the electrolyte has a significant impact on the capacity and rate performance of the battery.

[0147] In the embodiment of the present application, the establishing of the first feature parameter set and performing the first analysis in combination with the second target data includes:

[0148] Presetting a first correlation algorithm and a first correlation threshold;

[0149] Solving the correlation of the first feature parameter set based on the second target data;

[0150] Comparing the solution result with the first correlation threshold;

[0151] The feature parameter set that meets the comparison result is recorded as the second feature parameter set.

[0152] In an optional embodiment, the first correlation algorithm may adopt a Pearson correlation coefficient algorithm, which can effectively measure the linear correlation between two variables. By calculating the correlation coefficient between each parameter in the first characteristic parameter set and the second target data, parameters with a high correlation with battery life can be screened out.

[0153] In an optional embodiment, the first correlation algorithm can also use the Spearman rank correlation coefficient algorithm. This algorithm is suitable for evaluating the monotonic relationship between two variables and can work effectively even if the data does not satisfy the normal distribution. In this way, it can be further ensured that the correlation between the screened feature parameter set and the battery life is robust. In addition, the Spearman rank correlation coefficient algorithm also shows its advantages in dealing with nonlinear relationships, which helps to capture complex correlations that may be ignored by the Pearson correlation coefficient algorithm.

[0154] In an optional embodiment, the first correlation algorithm can also use the Kendall rank correlation coefficient algorithm. This algorithm is a non-parametric statistical method for evaluating the correlation between two variables, and is particularly suitable for situations where the amount of data is small or the data distribution is unknown. Through the Kendall rank correlation coefficient algorithm, the correlation between battery life and characteristic parameters can be ranked, thereby identifying key parameters that have a greater impact on battery life. In addition, the algorithm is insensitive to outliers, so it can still provide reliable analysis results when there are outliers in the battery data. By comprehensively using these three correlation algorithms, characteristic parameters closely related to battery life can be comprehensively evaluated and screened out, providing solid data support for subsequent battery life prediction and optimization.

[0155] In the embodiment of the present application, the first correlation algorithm uses the Pearson correlation coefficient method to calculate the correlation between each characteristic parameter and the battery life. The value range of the Pearson correlation coefficient is between -1 and 1. The closer the absolute value is to 1, the stronger the correlation is. By analyzing the correlation coefficient, the characteristic parameters with significant correlation with the battery life are screened out, and the possible influencing factors are preliminarily determined.

[0156] In an optional embodiment, let the feature parameter matrix be X, whose dimension is n×p, where p is the number of feature parameters. The battery life data is a vector Y, whose dimension is n×1.

[0157] In an optional embodiment, the mean of the characteristic parameter and the battery life is calculated. j (j=1,2,…,p), whose mean is Where X ij is the value of the jth characteristic parameter of the i-th sample. The mean value of the battery life Y is where Y i is the battery life value of the i-th sample.

[0158] In an optional embodiment, the covariance and standard deviation are calculated for each feature parameter X j and battery life Y, covariance Cov(X j ,Y) is calculated as: Characteristic parameter X j Standard Deviation The calculation formula is: The standard deviation S of the battery life Y Y The calculation formula is:

[0159] In an optional embodiment, the Pearson correlation coefficient is calculated, and the characteristic parameter X j Pearson correlation coefficient with battery life Y The calculation formula is:

[0160] In an optional embodiment, the absolute value of the correlation coefficient is analyzed, and after the Pearson correlation coefficient between each characteristic parameter and the battery life is calculated, the absolute value thereof is analyzed. Close to 1, it means that the characteristic parameter has a strong correlation with battery life; Close to 0, indicating a weak correlation.

[0161] In the embodiment of the present application, the threshold value is set to |r|>0.5. When the characteristic parameter X j Significantly correlated with battery life: Feature parameters that are significantly correlated with battery life are selected, and these parameters are initially considered to be factors that may affect battery life.

[0162] It should be noted that a first characteristic parameter set is established, and a first analysis is performed in combination with the second target data to obtain a second characteristic parameter set. By establishing the first characteristic parameter set, various data related to battery life can be systematically collected and organized. By performing the first analysis in combination with the second target data, characteristic parameters that have a significant correlation with battery life can be further screened out to form a second characteristic parameter set. Such steps help narrow the scope of research and improve the efficiency and accuracy of subsequent analysis. In this way, factors that have a significant impact on battery life can be more effectively identified, providing a scientific basis for battery design and optimization.

[0163] S103, performing a second analysis on the second feature parameter set to obtain a third feature data set;

[0164] In an optional embodiment, the second analysis may adopt a multiple linear regression analysis method. Through this method, a mathematical model can be established to describe the relationship between each characteristic parameter in the second characteristic parameter set and the battery life. In the multiple linear regression model, the battery life is used as a dependent variable, and the parameters in the second characteristic parameter set are used as independent variables. Through mathematical means such as the least squares method, the model parameters can be solved to obtain a regression equation that can predict the battery life.

[0165] In another optional embodiment, when performing multiple linear regression analysis, it is necessary to consider the goodness of fit of the model, that is, the model's ability to explain the data. The coefficient of determination (R2) is usually used to measure the goodness of fit of the model. The closer the R2 value is to 1, the stronger the model's ability to explain the data. In addition, it is necessary to perform a significance test on the model, such as an F test, to ensure that the model is significant as a whole, that is, at least one independent variable has a significant effect on the dependent variable.

[0166] In an optional embodiment, the second analysis can also use the principal component analysis (PCA) method. This method can convert multiple variables into a few principal components that can capture most of the information in the original data. Through principal component analysis, the dimension of the data can be reduced while retaining the characteristic parameters that have the greatest impact on battery life. After extracting the principal components, these principal components can be used as new characteristic variables for further multivariate linear regression analysis to establish a prediction model for battery life. This method helps to simplify the complexity of the model and may improve the prediction accuracy of the model.

[0167] In an embodiment of the present application, performing a second analysis on the second feature parameter set includes:

[0168] Preset the first principal component analysis algorithm;

[0169] A second analysis is performed on the second feature parameter set according to the first principal component analysis algorithm.

[0170] In the embodiment of the present application, principal component analysis is performed on the screened characteristic parameters with significant correlation using a principal component analysis algorithm. Principal component analysis can convert multiple related characteristic parameters into a few unrelated principal components, which can retain most of the information of the original data. By determining the contribution rate of the principal component, the principal component with a larger contribution rate is selected as the representative of the key influencing factor, the data dimension is further reduced and the main influencing factors are highlighted.

[0171] Specifically, after the first analysis, p characteristic parameters with significant correlation are screened out, and there are n sample data in total. These data are organized into an n×p matrix X, where each row represents a sample and each column corresponds to a characteristic parameter.

[0172] In an optional embodiment, the data is subjected to Z-score normalization. ij (represents the jth characteristic parameter value of the i-th sample, i=1,2,…,n; j=1,2,…,p), the standardized element z ij The calculation method is:

[0173]

[0174] Among them, (x_ j ) is the mean value of the jth characteristic parameter, and the calculation formula is σ j is the standard deviation of the jth characteristic parameter, calculated as The standardized data matrix is ​​denoted as Z, and its dimension is also n×p.

[0175] In an optional embodiment, the covariance matrix C of the standardized data matrix Z is calculated. The covariance matrix is ​​a p×p symmetric matrix, and its element c jk (representing the covariance between the jth feature and the kth feature) is calculated as follows:

[0176]

[0177] That is, the element in the j-th row and the k-th column of the matrix C is obtained by summing the products of the j-th standardized feature and the k-th standardized feature corresponding elements in all samples and then dividing by n-1.

[0178] Furthermore, we solve the eigenvalue λ of the covariance matrix C l (l=1,2,…,p) and the corresponding eigenvector v l (Each eigenvector is a p-dimensional vector.) Eigenvalues ​​and eigenvectors satisfy the following relationship:

[0179] [Cv l =λ l v l ]

[0180] Furthermore, the p eigenvalues ​​obtained are sorted in descending order, that is, (λ 1 ≥λ 2 ≥…≥λ p ≥0). The corresponding eigenvectors are rearranged in the order of eigenvalues ​​and recorded as (v 1 ,v 2 ,…,v p ).

[0181] Furthermore, the principal component contribution rate and cumulative contribution rate are calculated.

[0182] In an optional embodiment, the contribution rate CR of the lth principal component is l The calculation formula is:

[0183]

[0184] It indicates the proportion of the original data information contained in the lth principal component to the total information.

[0185] Furthermore, the cumulative contribution rate CVR of the first m principal components m The calculation formula is:

[0186]

[0187] It reflects the degree to which the first m principal components can retain the original data information.

[0188] In an optional embodiment, the first m principal components whose cumulative contribution rate reaches a certain proportion of 80% are selected as representatives of key influencing factors. Assuming that the first m principal components are selected, the corresponding eigenvectors are (v 1 ,v 2 ,…,v m ).

[0189] In an optional embodiment, for each sample i (i=1, 2, ..., n), its score F on the lth principal component is il (l=1,2,…,m) can be calculated by the following formula:

[0190] [F il =z iT v l ]

[0191] Among them, z i is the i-th row vector of the standardized data matrix Z (representing the standardized feature parameter vector of the i-th sample), (v l is the eigenvector corresponding to the lth principal component, and the superscript T represents the transposition operation. The score matrix F of each sample on the selected principal component is obtained, and its dimension is n×m.

[0192] In an embodiment of the present application, the first feature parameter set includes the second feature parameter set, and the second feature parameter set includes the third feature parameter set.

[0193] It should be noted that the second characteristic parameter set is subjected to a second analysis to obtain a third characteristic data set. Through the analysis of the second characteristic parameter set, the key factors affecting the life of lithium-ion energy storage batteries can be more accurately identified. The extraction of the third characteristic data set helps to further refine and optimize the battery management system, thereby improving the performance and life of the battery. In addition, the analysis results of the third characteristic data set can provide guidance for the design and manufacture of batteries, help engineers predict and avoid potential failures at an early stage, and ensure the stability and reliability of the battery under various working conditions.

[0194] S104: Establish a first extraction model, and extract influencing factors according to the first extraction model.

[0195] In an embodiment of the present application, the first extraction model includes:

[0196] The first extraction model is any model that takes the second target data as input and outputs several features in the third feature data set or can directly or indirectly obtain relevant parameters of several features in the third feature data set.

[0197] In an optional embodiment, the first extraction model can use a machine learning algorithm, such as a support vector machine (SVM), a random forest, a neural network, etc., to achieve automatic feature extraction and learning. The model is trained with a training data set so that it can identify key features related to battery life. During the training process, techniques such as cross-validation can be used to optimize model parameters and improve the generalization ability of the model. Ultimately, the first extraction model can output a third feature data set closely related to battery life based on the input second target data, providing important decision support for the battery management system.

[0198] In an optional embodiment, the first extraction model can also use a deep learning algorithm, such as a convolutional neural network (CNN), a recurrent neural network (RNN), or a long short-term memory network (LSTM), to process and analyze time series data. These algorithms are particularly suitable for processing a large amount of time series data generated during the battery charging and discharging process, and can capture the timing characteristics and patterns in the data. Through the training of the deep learning model, the accuracy and efficiency of feature extraction can be further improved, thereby more accurately predicting the battery life. In addition, the deep learning model can also continuously optimize its performance through self-learning and adjustment to adapt to the complexity of battery performance changing over time.

[0199] In an optional embodiment, the first extraction model can also use an ensemble learning method, such as a gradient boosted decision tree (GBDT) or a random forest, to improve the stability and prediction accuracy of the model. Ensemble learning can effectively reduce the risk of overfitting and improve the model's generalization ability for new data by combining the prediction results of multiple models. In the scenario of extracting factors affecting battery life, the ensemble learning method can combine the advantages of different models to more comprehensively capture the various factors affecting battery life. In addition, ensemble learning models usually have good interpretability, which helps researchers understand the specific effects of different feature parameters on battery life. In this way, a more reliable and accurate basis can be provided for the prediction and optimization of battery life.

[0200] In the embodiment of the present application, a convolutional neural network (CNN) is used to extract characteristic parameters of lithium-ion energy storage batteries. The generated principal component has n different characteristic parameters, and each characteristic parameter has corresponding time series data, and the time series length is m. The data is organized into a batch shape (batch size ,m,n) tensor form, where batchsize Indicates the number of samples input for each training. Divide the data set into training set, validation set and test set according to the usual proportion. 70% is used as training set, 15% as validation set and 15% as test set. Assuming the total data volume is N, the number of training sets is 0.7N, the number of validation sets is 0.15N and the number of test sets is 0.15N.

[0201] Furthermore, we define the convolutional layer. The number of convolutional kernels (filters) is set to k 1 = 16. Convolution kernel size (kernel size ) is set to s 1 =3. The stride is set to str 1 = 1. The padding is set to 'same' to keep the input and output time series of the same length. The output shape of this layer is calculated as: Output time series length Use 'same' padding, L 1 =m. The output shape is (batch size ,m,k 1 ).

[0202] Furthermore, the number of convolution kernels in the second convolution layer is set to k 2 =32. The convolution kernel size is set to s 2 =3. The step length is set to str 2 = 1. Filling mode is set to 'same'. Output time series length L 2 =L 1 =m. The output shape is (batch size ,m,k 2 ).

[0203] Furthermore, we define the pooling layer and the pooling kernel size (pool size ) is set to p = 2. The step length is set to str p =2. After the pooling layer, the time series length becomes The number of feature maps remains unchanged, and the output shape is

[0204] Furthermore, the multi-dimensional output tensor is flattened into a one-dimensional vector to connect to the fully connected layer. The output shape of the pooling layer is The length after flattening is

[0205] Furthermore, one or more fully connected layers are defined to further process the features. The activation function ReLU function (ReLU(x)=max(0,x)) is used between the fully connected layers to increase the nonlinear expression ability of the network.

[0206] Furthermore, the output layer uses a linear activation function.

[0207] Furthermore, for each training batch, the data is input into the model to obtain the predicted output of the model. The loss value between the predicted output and the true label is calculated according to the loss function. The parameters of the model are updated according to the loss value using the stochastic gradient descent (SGD) optimizer. Using the stochastic gradient descent (SGD) optimizer, its parameter update formula is: Where θ is the model parameter, η is the learning rate, is the gradient of the loss function with respect to the model parameters. Repeat the above steps and traverse the entire training set, usually for multiple training rounds (epochs). After each epoch, the performance of the model is usually evaluated on the validation set to observe whether the model is overfitting or underfitting, and the hyperparameters of the model are adjusted based on the evaluation results.

[0208] Furthermore, we define the loss function and use the mean square error loss function (mean squarederror ), the calculation formula is: Where n is the number of samples, y i is the true target value, is the predicted value.

[0209] Furthermore, the trained model is evaluated using the test set. Calculate the root mean square error It is used to measure the error between the model prediction and the true value.

[0210] Furthermore, the convolutional layers and fully connected layers of the trained model can extract feature representations that have an important impact on battery life.

[0211] In summary, the present invention proposes a method for extracting factors affecting the life of energy storage batteries, obtaining first target data, and performing a first preprocessing on the first target data to obtain second target data; establishing a first feature parameter set, and performing a first analysis in combination with the second target data to obtain a second feature parameter set; performing a second analysis on the second feature parameter set to obtain a third feature data set; establishing a first extraction model, and extracting influencing factors according to the first extraction model. This method can effectively improve the service life of lithium-ion energy storage batteries, reduce maintenance costs, and ensure the safe and stable operation of energy storage systems, which has important practical significance for promoting the widespread application of renewable energy and the development of energy storage technology.

[0212] Example 2

[0213] In a preferred embodiment, rich and multi-angle characteristic parameters are constructed based on the working principle and electrochemical characteristics of the battery, covering static characteristic parameters (such as material structure characteristics, battery design parameters, etc.), dynamic characteristic parameters (involving charging rate, discharge rate, charge and discharge depth, Coulomb efficiency, internal resistance change rate and other related parameters) and voltage characteristic parameters, current integral parameters, temperature change parameters, energy characteristic parameters, time characteristic parameters, power characteristic parameters, polarization related parameters, electrode potential parameters, electrolyte related parameters and many other categories. These characteristic parameters can comprehensively and deeply reflect the potential connection between its operating state and life from different angles such as the internal structure of the battery, electrochemical process and external operating state, providing sufficient information basis for accurate analysis of factors affecting battery life.

[0214] The correlation analysis between characteristic parameters and battery life was conducted through the Pearson correlation coefficient method. Based on scientific and reasonable thresholds (such as setting |r|>0.5 to screen significant correlations), characteristic parameters closely related to battery life were screened out, and possible influencing factors were initially locked in, which narrowed the scope for further mining of key factors and improved analysis efficiency and accuracy. With the help of principal component analysis, multiple characteristic parameters with significant correlations were screened out and converted into a few unrelated principal components. While retaining most of the original data information (by determining the appropriate cumulative contribution rate, such as 80% to select the principal component), the data dimension was effectively reduced, the main influencing factors were highlighted, and subsequent analysis and modeling could focus on key factors, further optimizing the extraction process of factors affecting battery life.

[0215] Convolutional neural network (CNN) is used to extract the processed feature parameters. By reasonably constructing the network structure (such as setting the appropriate number, size, and step size of convolution kernels, defining pooling layers, fully connected layers, and selecting appropriate activation functions, etc.), CNN can automatically learn and mine the deep feature representations hidden in the data. These features have an important impact on battery life. Compared with traditional feature extraction methods, it can better adapt to complex battery data characteristics and improve the effect and efficiency of feature extraction.

[0216] During the model training process, the mean square error loss function is used to measure the difference between the predicted output and the true label, and the stochastic gradient descent (SGD) optimizer is used to update the model parameters according to the loss value. The hyperparameters are adjusted by multiple iterative training on the training set and evaluating the model performance on the validation set to avoid overfitting or underfitting of the model. Finally, the test set is used to evaluate the prediction error of the model (measured by the root mean square error RMSE), which ensures the accuracy and generalization ability of the model. Therefore, based on the extracted key features, the life of lithium-ion energy storage batteries can be predicted more effectively, the battery performance can be optimized, and a strong decision-making basis can be provided for the rational use and maintenance of batteries.

[0217] Example 3

[0218] This embodiment also provides a system for extracting factors affecting the life of an energy storage battery, including:

[0219] A data processing module, used for acquiring first target data, and performing first preprocessing on the first target data to obtain second target data;

[0220] A first analysis module, used for establishing a first characteristic parameter set, and performing a first analysis in combination with the second target data to obtain a second characteristic parameter set;

[0221] A second analysis module, used for performing a second analysis on the second feature parameter set to obtain a third feature data set;

[0222] The model building module is used to build a first extraction model and extract influencing factors according to the first extraction model.

[0223] The above-mentioned unit modules may be embedded in or independent of a processor in a computer device in the form of hardware, or may be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to the above-mentioned modules.

[0224] This embodiment also provides a computer device, which may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 2As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a method for extracting factors affecting the life of an energy storage battery is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse, etc.

[0225] This embodiment further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0226] Acquire first target data, and perform first preprocessing on the first target data to obtain second target data;

[0227] Establishing a first feature parameter set, and performing a first analysis in combination with the second target data to obtain a second feature parameter set;

[0228] Performing a second analysis on the second feature parameter set to obtain a third feature data set;

[0229] A first extraction model is established, and influencing factors are extracted according to the first extraction model.

[0230] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0231] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The scheme in the embodiments of the present application can be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0232] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0233] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0234] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0235] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0236] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A method for extracting factors affecting the life of an energy storage battery, characterized in that: include: Acquire first target data, and perform first preprocessing on the first target data to obtain second target data; Establishing a first feature parameter set, and performing a first analysis in combination with the second target data to obtain a second feature parameter set; Performing a second analysis on the second feature parameter set to obtain a third feature data set; A first extraction model is established, and influencing factors are extracted according to the first extraction model.

2. The method for extracting factors affecting the life of an energy storage battery according to claim 1, characterized in that: The first extraction model comprises: The first extraction model is any model that takes the second target data as input and outputs several features in the third feature data set or can directly or indirectly obtain relevant parameters of several features in the third feature data set.

3. The method for extracting factors affecting the life of an energy storage battery according to claim 2, characterized in that: The first characteristic parameter set includes at least a plurality of static characteristic parameters and a plurality of dynamic characteristic parameters.

4. The method for extracting factors affecting the life of an energy storage battery according to claim 3, characterized in that: The step of establishing a first feature parameter set and performing a first analysis in combination with the second target data includes: Presetting a first correlation algorithm and a first correlation threshold; Solving the correlation of the first feature parameter set based on the second target data; Comparing the solution result with the first correlation threshold; The feature parameter set that meets the comparison result is recorded as the second feature parameter set.

5. The method for extracting factors affecting the life of an energy storage battery according to claim 4, characterized in that: The performing a second analysis on the second characteristic parameter set comprises: Preset the first principal component analysis algorithm; A second analysis is performed on the second feature parameter set according to the first principal component analysis algorithm.

6. The method for extracting factors affecting the life of an energy storage battery according to claim 5, characterized in that: The first feature parameter set includes the second feature parameter set, and the second feature parameter set includes the third feature parameter set.

7. The method for extracting factors affecting the life of an energy storage battery according to claim 6, characterized in that: The performing first preprocessing on the first target data comprises: Cleaning the first target data and identifying outliers; performing outlier processing on the first target data after outlier identification; The first target data after outlier processing is normalized to obtain second target data.

8. A system for extracting factors affecting the life of energy storage batteries, characterized in that: include: A data processing module, used for acquiring first target data, and performing first preprocessing on the first target data to obtain second target data; A first analysis module, used for establishing a first characteristic parameter set, and performing a first analysis in combination with the second target data to obtain a second characteristic parameter set; A second analysis module, used for performing a second analysis on the second feature parameter set to obtain a third feature data set; The model building module is used to build a first extraction model and extract influencing factors according to the first extraction model.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images