Intelligent optimization method and system for optimizing charging and discharging process of lithium battery

Abstract

The invention discloses an intelligent optimization method and system for optimizing the charging and discharging process of a lithium battery, and particularly relates to the technical field of lithium battery management, the lithium battery is divided into a plurality of monitoring areas according to the design structure of the lithium battery, and the monitoring areas are independently monitored and integrally monitored at the same time; carrying out charging and discharging inertia analysis on the basis of the overall monitoring information set, judging whether independent processing is needed or not, carrying out feature extraction on the monitoring information of each monitoring area, and carrying out fatigue analysis and degradation analysis; inputting an analysis result into a pre-trained convolutional neural network model, dividing a monitoring area into a high-degree disorder area, a medium-degree disorder area and a low-degree disorder area, performing dynamic adjustment on each area, and selecting a refined or integral dynamic optimization strategy to improve the performance and the safety of the battery; according to the invention, dynamic adjustment can be carried out according to degradation conditions of different areas, overuse of degraded areas is reduced, battery aging acceleration caused by overfatigue is avoided, the overheating problem is prevented, and the use safety of the lithium battery is enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of lithium battery management, and more specifically, to an intelligent optimization method and system for optimizing the charging and discharging process of a lithium battery. Background Art

[0002] With the widespread application of lithium batteries in new energy vehicles, energy storage equipment, smart devices and other fields, the safety, stability and service life of lithium batteries have gradually become key technical issues. During the charging and discharging process of lithium batteries, the battery parameters such as current, voltage and temperature may fluctuate violently, especially when working at high power output or in harsh environments, the performance of different areas inside the lithium battery may vary significantly. This differentiated performance can easily lead to overheating, capacity loss and even safety hazards in certain areas of the battery, such as thermal runaway caused by overheating.

[0003] Traditional battery management systems (BMS) mainly control the charge and discharge status of batteries by monitoring the overall parameters of the battery, but ignore the differentiated performance of each area inside the battery and cannot achieve refined management. Therefore, when individual battery cells overheat or current or voltage fluctuates, the system often cannot respond in time, resulting in excessive battery loss or safety problems. In addition, traditional BMS systems have difficulty dealing with phenomena such as lithium battery performance degradation and thermal fatigue, especially the inability to adopt differentiated dynamic optimization strategies for different areas, which limits the overall efficiency and life of the battery. Summary of the invention

[0004] To achieve the above object, the present invention provides the following technical solutions: The intelligent optimization method for optimizing the charging and discharging process of a lithium battery comprises the following steps: When the lithium battery is charging and discharging, it is divided into multiple monitoring areas according to its design structure, and each monitoring area is independently monitored to obtain multiple independent monitoring information sets, and the lithium battery is monitored as a whole at the same time to obtain an overall monitoring information set; Based on the overall monitoring information set, the charge and discharge inertia analysis is performed and it is decided whether to make an independent judgment. When making an independent judgment, feature extraction is performed on each independent monitoring information set, and fatigue analysis and degradation analysis are performed based on the extracted features; The results of fatigue analysis and degradation analysis are imported into the pre-trained convolutional neural network model, and the monitoring area corresponding to each independent monitoring information set is divided into a high disorder area, a medium disorder area, and a low disorder area; According to the division results, all highly disordered areas and moderately disordered areas are dynamically adjusted, and the overall monitoring information set obtained by overall monitoring of all moderately disordered areas and lithium batteries is analyzed for performance consistency. Based on the analysis results, it is decided to adopt refined dynamic optimization or holistic dynamic optimization operations, and refined dynamic optimization is adopted for all highly disordered areas.

[0005] In a preferred embodiment, performing charge and discharge inertia analysis based on the overall monitoring information set refers to: In a fixed time window, obtain each preset charging and discharging inertia parameter in the overall monitoring information set, take the charging and discharging inertia parameter corresponding to one type as the target inertia parameter, and calculate the change rate of the target inertia parameter at adjacent time points. ; i represents the type of inertia parameter, Indicates the sampling time interval corresponding to the inertial parameter type i, Represents the difference of the sampling data of adjacent time points of inertial parameter type i, and obtains a set of change rates The coefficient of variation of the time series Si is calculated for all the charging and discharging inertia parameters of type i.

[0006] In a preferred embodiment, the basis for determining whether to make an independent judgment is: If the coefficient of variation of the time series Si is greater than the preset inertia threshold of the charging and discharging inertia parameter of type i, a fluctuation signal is generated; if the coefficient of variation of the time series Si is less than or equal to the preset inertia threshold of the charging and discharging inertia parameter of this type, a normal signal is generated; when there is a fluctuation signal, an independent judgment is made for each monitoring area.

[0007] When feature extraction is performed on each independent monitoring information set, a thermal fatigue information set and a performance degradation information set are extracted, fatigue analysis is performed based on the thermal fatigue information set to generate a fatigue index, and degradation analysis is performed based on the performance degradation information set to generate a degradation index.

[0008] In a preferred embodiment, the logic for obtaining the fatigue index is: Obtain the thermal fatigue information set of each monitoring area and obtain the collected temperature time series and the corresponding sampling time t, and then the temperature difference between all adjacent data points Sum and average to get the average temperature change rate, the formula is: ; Calculate the second-order derivative of temperature change to reflect the acceleration of temperature fluctuation. Sum the absolute values ​​of acceleration at all times to get the average value of acceleration term. The formula is: ; Set the temperature fluctuation threshold and count the number of times the temperature fluctuation amplitude exceeds the threshold. Indicates the frequency of violent temperature fluctuations in the area; Introducing the time decay factor , giving lower weight to older data to ensure that current data contributes more to the fatigue index, combining the temperature change rate, acceleration term and fluctuation frequency to generate the fatigue index: ; The base time representing the start of time decay, , are preset proportional factors. is the weight coefficient of temperature change acceleration, which is used to adjust the contribution of temperature acceleration to fatigue index. represents the time interval between adjacent data points, N is the amount of sampled data, represents the i-th data point in the temperature time series, Represents the fatigue index. The fatigue index calculation formula performs boundary processing, that is, the last acceleration value calculated at the N-2th point is extended to the N-1th point.

[0009] In a preferred embodiment, the logic for obtaining the degradation index is: Obtain the performance degradation information set of each monitoring area, obtain the battery data at the current time point, including the internal resistance Rt, voltage Vt and capacity Ct, as well as the reference data R0, V0, C0 at the initial moment, and obtain the corresponding internal resistance Ri and voltage Vi from the battery at the past M time points; Calculate the change rate of the internal resistance at the current moment relative to the initial moment, and perform logarithmic mapping after squaring it. The formula is: ; Calculate the battery voltage attenuation rate and add the smoothing term k to avoid distortion in calculations at extremely low voltages. The formula is: ; Calculate the capacity decay rate at the current moment and add the power exponent p to adjust the impact of capacity decay on the overall degradation index. The formula is: ; Perform weighted processing on the battery's past M sampling data and calculate the ratio of internal resistance to voltage for each sampling. , and weighted average it by weight factor wi, the formula is: ; The internal resistance change, voltage decay, capacity decay and multiple sampling data are integrated into the final degradation index. The complete calculation formula is: ; f1, f2, and f3 are all preset influence coefficients. is the degradation index.

[0010] In a preferred embodiment, the results of fatigue analysis and degradation analysis, namely the fatigue index and degradation index, are imported into a pre-trained convolutional neural network model as input data, and the monitoring area corresponding to each independent monitoring information set is divided into a high disorder area, a moderate disorder area and a low disorder area.

[0011] In a preferred embodiment, the refined dynamic optimization is used for all highly disordered areas, which means that the battery management module is used to dynamically optimize all highly disordered areas respectively, and the temperature safety threshold of the monomer is dynamically set, and the formula is: ; is the preset temperature basic safety threshold, The temperature safety threshold of the monomer is set dynamically. , , Respectively represent the current value, voltage value and temperature change value, , They represent the maximum current and maximum voltage allowed by lithium batteries. , , All are preset safety factors.

[0012] In a preferred embodiment, all the moderately disordered areas and the lithium battery are subjected to overall monitoring to obtain an overall monitoring information set for performance consistency analysis, and the use of refined dynamic optimization or overall dynamic optimization operation is determined according to the analysis results. The current current value, voltage value and temperature change value in the overall monitoring information set obtained by overall monitoring of the moderately disordered area and the lithium battery are respectively obtained and used as a set of corresponding vectors, and the Euclidean distance of the corresponding vectors is calculated. When the standard deviation of all Euclidean distances is greater than the preset performance consistency threshold, all moderately disordered areas are subjected to the same refined dynamic optimization as the highly disordered areas. When the standard deviation of all Euclidean distances is less than or equal to the preset performance consistency threshold, all moderately disordered areas are subjected to overall dynamic optimization. Overall dynamic optimization refers to using the battery management module to dynamically optimize all moderately disordered areas and dynamically set the overall temperature safety threshold. The formula is: ; is the preset temperature basic safety threshold, The overall temperature safety threshold is set dynamically. , , They respectively represent the average current value, average voltage value and average temperature change value of all moderately disordered areas.

[0013] In a preferred embodiment, an intelligent optimization system for optimizing the charging and discharging process of a lithium battery includes: The regional monitoring module is divided into multiple monitoring areas according to the design structure of the lithium battery, and each monitoring area is independently monitored to obtain multiple independent monitoring information sets. At the same time, the entire lithium battery is monitored as a whole to obtain an overall monitoring information set; The inertia analysis module performs charge and discharge inertia analysis based on the overall monitoring information set and decides whether to make independent judgments on each monitoring area based on the inertia analysis results; The feature extraction module extracts features from each independent monitoring information set when making independent judgments. After feature extraction, fatigue analysis and degradation analysis are performed. The results of fatigue analysis and degradation analysis are imported into a pre-trained convolutional neural network model to divide each monitoring area into a high disorder area, a medium disorder area, and a low disorder area. The dynamic adjustment module dynamically adjusts all highly disordered areas and moderately disordered areas according to the regional division results, performs performance consistency analysis on the moderately disordered areas and the overall monitoring information set of the lithium battery, and decides whether to adopt refined dynamic optimization or overall dynamic optimization operations; The battery management module manages the batteries in each monitoring area according to the dynamic optimization conditions of the dynamic adjustment module.

[0014] Technical effects and advantages of the present invention: The present invention effectively improves the management accuracy of lithium batteries by dividing lithium batteries into multiple monitoring areas and independently monitoring and analyzing each area. Unlike the traditional battery management system that only monitors overall parameters, the present invention can perform differentiated dynamic optimization according to the specific status of each area, dynamically set the safety threshold of each area for fluctuations in temperature, current, and voltage, and avoid over-response or failure of the overall system due to local problems, thereby improving the safety and stability of lithium batteries.

[0015] By performing thermal fatigue analysis and performance degradation analysis on each monitoring area, the present invention can effectively identify the aging degree and fatigue state of lithium batteries in different areas. During charging and discharging, the system can dynamically adjust the degradation conditions of different areas to reduce excessive use of degraded areas and avoid accelerated battery aging due to excessive fatigue. This method can slow down the rate of battery performance degradation, thereby extending the service life of lithium batteries.

[0016] The consistency analysis in the present invention can accurately determine whether the working status of these areas deviates from the overall trend by comparing the moderately disordered areas with the overall performance. Through methods such as Euclidean distance and standard deviation, the system can flexibly decide whether to perform refined dynamic optimization or overall dynamic optimization, avoid taking overly drastic adjustment measures when there are local fluctuations, and ensure the stability of the system. In addition, using the average value as the adjustment benchmark can balance the relationship between local fluctuations and overall performance, reducing unnecessary over-adjustment of the system.

[0017] The present invention can monitor the current, voltage and temperature changes of each monitoring area in real time, and prevent overheating problems in certain areas by dynamically adjusting the dynamic temperature safety threshold set by the module, effectively avoiding the risk of thermal runaway of lithium batteries. For highly disordered areas, the system adopts refined dynamic optimization to ensure that these areas can be quickly responded and accurately regulated when there is a high risk, further ensuring the safety of battery use. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings; Figure 1 This is a schematic diagram of the intelligent optimization method for optimizing the charging and discharging process of a lithium battery in the present invention.

[0019] Figure 2 This is a schematic diagram of the intelligent optimization system for optimizing the charging and discharging process of a lithium battery in the present invention. DETAILED DESCRIPTION

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

[0021] Reference Figure 1 - Figure 2 The following embodiments are obtained: Embodiment 1: An intelligent optimization method for optimizing the charging and discharging process of a lithium battery comprises the following steps: When a lithium battery is charging and discharging, it is divided into regions according to its design structure to obtain multiple monitoring regions. Each monitoring region is monitored independently to obtain multiple independent monitoring information sets, and the lithium battery is monitored as a whole at the same time to obtain an overall monitoring information set. The internal structure of a lithium battery usually contains multiple cells, and the physical position, heat dissipation conditions and load status of the cells may be different. Therefore, in actual applications, different regions (i.e. different cells) in the battery pack may have different temperature, voltage and current change trends. By dividing the battery into regions, it is possible to achieve independent monitoring of each region, and to master the working status of each region in a refined manner, so as to discover potential problems and optimize them in time. For example: For example, a lithium battery pack containing 12 cells can be divided into 4 monitoring regions according to its cell arrangement and heat dissipation characteristics, and each region consists of 3 cells. These regions may be divided based on the physical position of the cell, the contact condition of the radiator, or other cell characteristics. Assuming that the first region is located in the center of the battery pack and has poor heat dissipation conditions, it may be more prone to overheating than other regions, so it needs to be monitored independently. In addition to independently monitoring each region, the entire lithium battery pack also needs to be monitored as a whole. Overall monitoring can provide the global status of the battery, such as overall voltage, current, temperature, etc. These data can be used in conjunction with regional monitoring data to provide a comprehensive battery health status assessment. For example: in a 12-cell battery pack, the overall monitoring information can include the total voltage, current and temperature of the battery pack. For example, the overall voltage may be 48V, while the voltages of each region are slightly different (such as 12V per group, but with slight fluctuations). By comparing these differences, it can be determined whether the overall performance matches the health status of the local area.

[0022] Based on the overall monitoring information set, the charge and discharge inertia analysis is performed and it is decided whether to make an independent judgment. When making an independent judgment, feature extraction is performed on each independent monitoring information set. Based on the extracted features, fatigue analysis and degradation analysis are performed; the main purpose of the charge and discharge inertia analysis is to detect the stability of the system by analyzing the changes in various parameters during the battery charge and discharge process. When the battery is charging and discharging, if the change rate of various parameters (such as voltage, current, temperature, etc.) is large, it may cause battery performance degradation or safety hazards. Through inertia analysis, the system can judge the volatility and inertia of these parameters, thereby early warning of potential problems. By extracting features from each independent monitoring information set, relevant features can be extracted from parameters such as voltage, current, and temperature to reflect the fatigue state (such as thermal fatigue caused by temperature fluctuations) and degradation state (such as battery capacity attenuation, increased internal resistance, etc.) of the battery. Fatigue analysis and degradation analysis can effectively evaluate the health of the battery and help optimize the battery life and safety. For example: If the temperature in a certain monitoring area fluctuates frequently, it may indicate that the area is experiencing greater thermal fatigue. Through feature extraction and fatigue analysis, a fatigue index can be obtained to indicate the degree of thermal fatigue in the area. Similarly, by extracting features such as internal resistance and voltage, the degree of degradation in the area can be analyzed.

[0023] The results of fatigue analysis and degradation analysis are imported into the pre-trained convolutional neural network model, and the monitoring area corresponding to each independent monitoring information set is divided into a high disorder area, a moderate disorder area and a low disorder area; based on the results of fatigue analysis and degradation analysis, the data of each area is imported into the pre-trained convolutional neural network model, and the model can automatically divide each area into a high disorder area, a moderate disorder area and a low disorder area according to the input data. The high disorder area represents that the area has obvious abnormalities, the moderate disorder area indicates that there may be potential problems, and the low disorder area represents a relatively normal operating state. For example: Assume that after analysis, area 1 and area 2 are divided into high disorder areas because their temperature change rate, internal resistance change rate and voltage fluctuation range are obviously beyond the normal range; area 3 and area 4 are divided into moderate, indicating that their operating state is less abnormal, but further optimization is required. The ultimate goal is to prevent the lithium battery from triggering the independent judgment mechanism or all areas are divided into low disorder areas after the trigger, indicating that the lithium battery is in normal operation during charging and discharging.

[0024] According to the division results, all highly disordered areas and moderately disordered areas are dynamically adjusted, and the overall monitoring information set obtained by overall monitoring of all moderately disordered areas and lithium batteries is analyzed for performance consistency. According to the analysis results, it is decided to adopt refined dynamic optimization or overall dynamic optimization operation, and refined dynamic optimization is adopted for all highly disordered areas. After dividing the high, moderate and low disordered areas, the system will dynamically adjust the high and moderate disordered areas to optimize the working conditions of these areas. For example, the system can prevent excessive temperature or battery damage by reducing the charging current and adjusting the operating temperature threshold. This can effectively avoid further deterioration of the battery and extend its service life. For example: If an area is divided into a highly disordered area, the system will immediately start dynamic adjustment, such as reducing the charging rate of the area, or reducing the temperature of the area through external heat dissipation measures. The performance consistency analysis is to determine whether the working state of the moderately disordered area is consistent with the overall working state of the lithium battery. By comparing the current, voltage and temperature parameters of these areas with the overall battery, it can be determined whether there are potential problems in these areas. If the consistency is poor, these areas need to be refined and dynamically optimized; if the consistency is good, overall dynamic optimization can be used. For example: If the current and voltage fluctuations in a moderately disordered area are significantly higher than the overall situation of the battery pack, the system may adopt more refined optimization methods for this area, such as lowering the current or performing local cooling; if the fluctuations in this area are small, the overall optimization operation can be performed directly. Refined dynamic optimization is required in highly disordered areas due to obvious abnormalities. Refined dynamic optimization means that the system will control the working status of these areas more accurately, adjust parameters such as temperature, current, and voltage in real time, and ensure that the working status of these areas returns to the normal range to avoid safety hazards. For example: If the temperature of a highly disordered area exceeds the safe range, the system will dynamically adjust the temperature threshold of the area and reduce the charging current until the temperature returns to a safe range.

[0025] Assume that a battery pack consists of 12 cells arranged in a 4x3 structure, located in different physical locations, and have different heat dissipation characteristics. Based on the physical structure and heat dissipation characteristics, the battery pack can be divided into 4 monitoring areas, each including 3 cells. For example: Area 1: Located at the top of the battery pack, close to the radiator, usually with a lower temperature. Area 2: Located in the middle of the battery pack, affected by the surrounding cells, the temperature fluctuates greatly. Area 3: Located at the bottom of the battery pack, the heat dissipation is poor and the temperature is more likely to rise. Area 4: Located on the side of the battery pack, the temperature changes are relatively stable.

[0026] The charge and discharge inertia analysis based on the overall monitoring information set means: in a fixed time window, each preset charge and discharge inertia parameter in the overall monitoring information set is obtained, and the charge and discharge inertia parameter corresponding to one type is used as the target inertia parameter, and the change rate of the target inertia parameter at adjacent time points is calculated. ; i represents the type of inertia parameter, Indicates the sampling time interval corresponding to the inertial parameter type i, Represents the difference of the sampling data of adjacent time points of inertial parameter type i, and obtains a set of change rates The time series Si composed of all types i of charging and discharging inertia parameters is calculated to calculate the coefficient of variation of the time series Si. Within a fixed time window, the system calculates the rate of change of each charging and discharging inertia parameter (such as voltage and current change rate) at adjacent time points, and combines these change rates into a time series Si. By calculating the coefficient of variation of the time series, the coefficient of variation calculation formula is the prior art, which will not be described in detail here, and the system can effectively identify abnormal fluctuations. If the coefficient of variation exceeds the preset inertia threshold, it means that the parameter fluctuates greatly, which may mean that the battery is abnormal or about to have problems. Through inertia analysis, the battery management system (BMS) in the prior art corresponds to the battery management module in the present invention, which can identify the fluctuation of the battery state in advance and make appropriate responses. This can not only improve the system's response speed to emergencies, but also avoid unnecessary optimization and adjustment operations. Only when obvious fluctuations are detected will the system enter the independent judgment stage to further analyze the conditions of the specific area, thereby improving the efficiency and accuracy of the system operation.

[0027] The basis for deciding whether to make an independent judgment is: if the coefficient of variation of the time series Si is greater than the preset inertia threshold of the charge and discharge inertia parameter of the type i, a fluctuation signal is generated; if the coefficient of variation of the time series Si is less than or equal to the preset inertia threshold of the charge and discharge inertia parameter of the type i, a normal signal is generated; when there is a fluctuation signal, an independent judgment is made for each monitoring area. The system determines whether an independent judgment is required based on the coefficient of variation of the time series. If the coefficient of variation of a certain charge and discharge inertia parameter exceeds the preset inertia threshold, the system generates a fluctuation signal and starts independent judgment. Conversely, if the coefficient of variation of all parameters is within the normal range, the system generates a normal signal and no further independent judgment is required. Specific example: Assume that during the battery charging process, the system detects that the voltage change fluctuates greatly at different time points, which may mean that the health of a certain battery cell is abnormal. If the coefficient of variation of the voltage change rate is found to exceed the preset threshold in the inertia analysis, the system generates a fluctuation signal, indicating that the charging process is unstable. At this time, the system will independently analyze each monitoring area, further detect its temperature, current and other parameters, and decide whether to take optimization measures.

[0028] When feature extraction is performed on each independent monitoring information set, the thermal fatigue information set and the performance degradation information set are extracted, fatigue analysis is performed based on the thermal fatigue information set to generate a fatigue index, and degradation analysis is performed based on the performance degradation information set to generate a degradation index. The fatigue index is mainly used to measure the thermal fatigue experienced by the battery during the charging and discharging process, reflecting the cumulative damage caused by temperature changes to the battery materials or systems. A larger fatigue index means that the battery has experienced more temperature fluctuations during the charging and discharging process, especially rapid and drastic temperature changes. A higher fatigue index indicates that the battery has greater thermal stress during operation, and the battery material may suffer microstructural damage due to frequent temperature changes, thereby shortening the battery life. The degradation index is used to evaluate the overall degradation of the battery, which is usually related to the increase in the internal resistance, voltage drop and capacity loss of the battery. A larger degradation index means that the internal performance of the battery has undergone significant degradation, and there may be a decrease in capacity, an increase in internal resistance or voltage decay. A higher degradation index is usually a sign of battery aging, meaning that the battery is in poor health and is approaching or has exceeded its design life.

[0029] The logic for obtaining the fatigue index is: Obtain the thermal fatigue information set of each monitoring area and obtain the collected temperature time series By collecting the data of battery temperature changes over time and the corresponding sampling time t, the thermal load on the battery during the charging and discharging process can be reflected. Temperature change is an important factor affecting battery performance, especially when the temperature fluctuates greatly, it will aggravate the fatigue damage of the battery. Therefore, the temperature time series is the basic data for analyzing thermal fatigue.

[0030] The temperature difference between all adjacent data points Sum and average to get the average temperature change rate, the formula is: The average temperature change rate reflects the fluctuation of the battery temperature during the monitoring period. By calculating the temperature difference between adjacent data points, the temperature fluctuation can be judged. If the temperature change rate is large, it means that the battery has experienced severe temperature fluctuations during the charge and discharge process, which may cause thermal fatigue. By taking the absolute value to measure the temperature change between each two time points and calculating the average value, the overall trend of temperature fluctuation can be obtained.

[0031] Calculate the second-order derivative of temperature change to reflect the acceleration of temperature fluctuation. Sum the absolute values ​​of acceleration at all times to get the average value of acceleration term. The formula is: ; The second derivative of temperature change, that is, the acceleration of temperature change, reflects the severity of temperature fluctuations. If the acceleration of temperature change is large, it means that the temperature change is not only fast, but also the severity of temperature fluctuations increases, which may accelerate battery damage. The second-order derivative is used to measure the change in the speed of temperature change, that is, the acceleration of temperature. By taking the absolute value of the second-order derivative and averaging it, the positive and negative directions of temperature fluctuations can be avoided from canceling each other out, thereby more accurately measuring the severity of the fluctuations.

[0032] Set the temperature fluctuation threshold and count the number of times the temperature fluctuation amplitude exceeds the threshold. Indicates the frequency of severe temperature fluctuations in the area; by setting a temperature fluctuation threshold, counting the number of times the temperature fluctuation amplitude is greater than the threshold, reflecting the frequency of severe temperature fluctuations in the area. Frequent temperature fluctuations are one of the important manifestations of battery thermal fatigue. The number of times above a certain temperature fluctuation threshold reflects the frequency of temperature fluctuations. Frequent large fluctuations usually indicate that the battery is in an unstable state or overheated, which may cause thermal fatigue of the battery. Therefore, counting the frequency of large fluctuations helps to quantify this phenomenon.

[0033] Introducing the time decay factor , giving lower weights to older data to ensure that current data contributes more to the fatigue index, and the time decay factor is used to weighted decay the older data and give it lower weights. This can highlight the contribution of current temperature changes to the fatigue index and ensure that the fatigue analysis results are closer to the current state of the battery. The time decay factor is used to avoid excessive impact of temperature fluctuations in the distant past on the current fatigue state, so that more focus can be placed on recent data to ensure that the evaluation results are more timely.

[0034] The temperature change rate, acceleration term, and fluctuation frequency are combined to generate a fatigue index: ; The base time representing the start of time decay, , are preset proportional coefficients, which respectively control the contribution of temperature change rate and acceleration to fatigue index, and the contribution of temperature fluctuation frequency to fatigue index. is the weight coefficient of temperature change acceleration, which is used to adjust the contribution of temperature acceleration to fatigue index. represents the time interval between adjacent data points, N is the amount of sampled data, represents the i-th data point in the temperature time series, Represents the fatigue index. The fatigue index calculation formula performs boundary processing, that is, the last acceleration value calculated at point N−2 is generalized to point N−1. The fatigue index FI combines factors such as the temperature change rate, the temperature change acceleration, the temperature fluctuation frequency, and performs time decay on the historical data. This formula fully reflects the thermal fatigue condition of the battery. Boundary processing is used to solve the problem of insufficient data points when calculating the temperature change acceleration (because three consecutive points are required to calculate the second-order derivative). By generalizing the acceleration value of point N−2, the consistency of the entire data set can be guaranteed. When calculating the second-order derivative, the last two data points are not enough to calculate the acceleration value, so the acceleration result of point N−2 is usually applied to point N−1 to complete the calculation of the entire data set.

[0035] The first part of the formula combines the temperature change rate and the temperature change acceleration, which together measure the magnitude of temperature fluctuations and the rate of change. , the contribution of temperature change acceleration to fatigue index can be adjusted according to actual needs. The combination of temperature change rate and temperature change acceleration reflects the simultaneous effects of temperature fluctuation and rapid change. The acceleration term is weighted by Control, to ensure that the fatigue index is more sensitive to severe temperature fluctuations. The second part is the frequency of temperature fluctuations. Through the exponential decay function Adjusted so that older data is less weighted, ensuring that the fatigue index focuses more on the current state.

[0036] The logic for obtaining the degradation index is: Obtain the performance degradation information set of each monitoring area, obtain the battery data at the current time point, including the internal resistance Rt, voltage Vt and capacity Ct, as well as the reference data R0, V0, C0 at the initial moment, and obtain the corresponding internal resistance Ri and voltage Vi from the battery at the past M time points; Calculate the change rate of the internal resistance at the current moment relative to the initial moment, and perform logarithmic mapping after squaring it. The formula is: ; The use of logarithmic and square forms avoids excessive influence of linear growth on the exponential, and when the internal resistance changes slightly, the formula can more sensitively reflect small degradation. The square can increase the penalty for larger changes, while the introduction of the logarithmic function ensures nonlinear and smooth changes in the degradation rate.

[0037] Calculate the battery voltage attenuation rate and add the smoothing term k to avoid distortion in calculations at extremely low voltages. The formula is: The voltage decay term is processed in the form of a square root, which increases the decay control of the high voltage interval during the voltage decay process, making the impact of voltage changes on degradation in different intervals more balanced. The constant k prevents calculation instability when the voltage is close to zero.

[0038] Calculate the capacity decay rate at the current moment and add the power exponent p to adjust the impact of capacity decay on the overall degradation index. The formula is: The capacity decay term introduces a power exponent p to enhance the impact of capacity decay when the battery enters the aging stage. When p>1, the impact of capacity decay on the degradation index will accelerate; when p<1, the capacity decay has little impact.

[0039] Perform weighted processing on the battery's past M sampling data and calculate the ratio of internal resistance to voltage for each sampling. , and weighted average it by weight factor wi, the formula is: ; The internal resistance and voltage ratio of the past M samples are integrated to quantify the comprehensive performance changes of the battery cell over the past period of time. wi is a weight factor. Data at different sampling times can be assigned different weights to smooth the impact of noise and better reflect the overall trend. The weight factor wi is introduced to distinguish sampling data from different time periods, so that more important or more recent data has a greater impact on the final degradation index. By weighted averaging of historical data, the overall degradation trend of the battery can be reflected more smoothly, avoiding the randomness of a single measurement result affecting the overall judgment.

[0040] The internal resistance change, voltage decay, capacity decay and multiple sampling data are integrated into the final degradation index. The complete calculation formula is: ; f1, f2, and f3 are all preset influence coefficients, which control the contribution of internal resistance change, voltage decay, and capacity decay in the degradation index respectively. f1 determines the weight of the influence of internal resistance change on degradation, reflecting the severity of internal resistance growth. f2 controls the influence of voltage decay, and the contribution of voltage fluctuation to battery performance degradation is reflected by this coefficient. f3 reflects the importance of capacity decay, which is adjusted by this coefficient. is the degradation index. The change of internal resistance is an important indicator of battery aging. Its influence is enhanced through logarithmic and smoothing processing. Voltage decay and capacity decay reflect the energy output capacity and energy storage capacity of the battery respectively. Smoothing terms and power exponential processing are used in the formula to avoid calculation anomalies and enhance sensitivity to decay. Multiple sampling weighted average combines historical data to further enhance the degradation index's comprehensive evaluation ability of battery health status over a long period of time. Through these processing steps, the degradation index can comprehensively and accurately evaluate the degradation of the battery and provide data support for subsequent optimization operations.

[0041] The results of fatigue analysis and degradation analysis, namely fatigue index and degradation index, are imported into the pre-trained convolutional neural network model as input data, and the monitoring area corresponding to each independent monitoring information set is divided into high disorder area, medium disorder area and low disorder area. After completing fatigue analysis and degradation analysis in each monitoring area, two important analysis results, namely fatigue index and degradation index, are obtained respectively. These two results are provided to the convolutional neural network model as the main input data. Through multi-layer convolution operations, the convolutional neural network model can extract features from the input fatigue index and degradation index. These features are determined based on the pre-trained weights and parameters of the model, with the aim of identifying abnormal conditions in the battery area. The features of different areas are expressed differently, and the model can classify different areas into different risk levels by learning and identifying these differences.

[0042] The model classifies each monitoring area into risk levels based on the extracted features. Highly disordered areas usually indicate that both fatigue and degradation indices are at high levels, indicating that the battery status in this area is very unstable and there is a greater risk. Moderately disordered areas indicate that there are certain problems with the battery status, but it has not yet reached a high-risk state. Low-disordered areas mean that the battery is in a relatively stable working state with a lower risk. The convolutional neural network model is pre-trained based on a large amount of historical data and battery monitoring information, so it can identify complex patterns related to battery status. The model learns the performance of the battery under different fatigue and degradation states through a large amount of training data, and makes accurate judgments in actual use through this learning. The trained model does not need to be re-learned during application, but directly uses the previous learning results to classify real-time data.

[0043] Fine-grained dynamic optimization is used for all highly disordered areas. This means using the battery management module to dynamically optimize all highly disordered areas and dynamically set the temperature safety threshold of the monomer. The formula is: ; It is a preset temperature-based safety threshold, usually provided by the battery manufacturer based on the battery characteristics. It is the recommended safety temperature of the battery under normal working conditions. The temperature safety threshold of the cell is dynamically set, that is, the new upper safety temperature limit is set according to the actual working conditions of the battery. , , Respectively represent the current value, voltage value and temperature change value, , They represent the maximum current and maximum voltage allowed by lithium batteries. , , These are preset safety factors used to adjust the impact of current, voltage, and temperature change rate on the temperature safety threshold. By calculating the ratio of the current current to the maximum current, the impact of the current current on the battery thermal load can be determined. The ratio of the current voltage value to the maximum voltage allowed by the battery is used to evaluate the impact of voltage changes on the battery working state. The current temperature change rate reflects the temperature change amplitude per unit time and is used to evaluate the severity of temperature fluctuations.

[0044] When the current operating conditions of the battery (current, voltage and temperature change rate) are at a high level, the system will dynamically adjust the battery's temperature safety threshold in real time based on these parameters to keep it below the preset basic safety temperature. By dynamically setting the temperature safety threshold, the battery management module can ensure that the battery will not be damaged or dangerous due to overheating during operation. This dynamic adjustment mechanism can improve the safety of the battery and extend its service life.

[0045] All moderately disordered areas and lithium batteries are monitored as a whole to obtain a set of overall monitoring information for performance consistency analysis. Based on the analysis results, it is decided to adopt refined dynamic optimization or overall dynamic optimization operations, which means: The current current value, voltage value and temperature change value in the overall monitoring information set obtained by overall monitoring of the moderately disordered area and the lithium battery are respectively obtained and used as a set of corresponding vectors, and the Euclidean distance of the corresponding vectors is calculated. When the standard deviation of all Euclidean distances is greater than the preset performance consistency threshold, all moderately disordered areas are subjected to the same refined dynamic optimization as the highly disordered areas. When the standard deviation of all Euclidean distances is less than or equal to the preset performance consistency threshold, all moderately disordered areas are subjected to overall dynamic optimization. Overall dynamic optimization refers to using the battery management module to dynamically optimize all moderately disordered areas and dynamically set the overall temperature safety threshold. The formula is: ; is the preset temperature basic safety threshold, The overall temperature safety threshold is set dynamically. , , Respectively represent the average current value, average voltage value and average temperature change value of all moderately disordered areas. By adopting the average value, the system can better reflect the overall trend of all moderately disordered areas rather than the extreme performance of a single area. This processing method can ensure that the optimization process is smoother on a global scale, which helps to improve the stability and consistency of the system. If the maximum or minimum value of each area is used directly, it may cause the system to over-adjust. For example, if the current or temperature change rate in a certain area is abnormally high, the system may make a large adjustment to the area, which will affect the performance of other normal areas. Using the average value can ensure that the system is more robust when adjusting and avoid unnecessary adjustments. In the overall optimization process, the use of the average value can simplify the calculation complexity, especially when monitoring multiple areas. By calculating the average value, the system can respond and adjust quickly, avoiding the complexity of processing each individual area, thereby improving the adjustment efficiency.

[0046] For moderately disordered areas, by comparing the current values, voltage values, and temperature change rates of these areas with the data of the overall monitoring information set, it is possible to identify whether there are inconsistent areas. If the characteristics of these areas (such as current, voltage, and temperature fluctuations) differ greatly from the trend of the overall monitoring data, it indicates that the battery cells in these areas are performing abnormally and require refined dynamic optimization. If the moderately disordered areas are relatively consistent with the overall monitoring data, a holistic dynamic optimization strategy can be used. When the system finds that the consistency between the moderately disordered areas and the overall performance is poor, refined dynamic optimization is necessary. Refined dynamic optimization can adjust the parameters of these inconsistent areas more accurately to ensure that they are restored to a safe range. Areas with better consistency can reduce system complexity and adjustment costs through holistic optimization, which reflects the flexibility of the dynamic adjustment strategy.

[0047] Euclidean distance is a common method for calculating the difference between different vectors (here, current values, voltage values, and temperature change values). It reflects the degree of difference between the data in these areas and the monitoring data of the entire battery. By calculating the Euclidean distance between each moderately disordered area and the overall battery data, the system can quantify the deviation of these areas from the overall performance. The larger the Euclidean distance, the greater the difference between the current, voltage, and temperature changes in the area and the overall monitoring information, indicating that there may be obvious instability or potential risks in the area. A smaller distance means that these parameters are close to the overall monitoring data, indicating that the performance of the area is consistent with the overall. Using the Euclidean distance can accurately measure the difference between the moderately disordered area and the overall battery state, which is the basis for judging whether refined dynamic optimization is needed. Through this distance, the system can quantitatively evaluate the health status of each area, so as to carry out targeted management. The role of the standard deviation in this process is to quantify the volatility between the data in the moderately disordered area and the overall data. By calculating the standard deviation of all Euclidean distances, the system can judge the consistency of each area with the overall monitoring data. If the standard deviation is large, it means that these areas have large fluctuations and there may be local performance instability. It is necessary to take refined optimization measures for the moderately disordered areas. On the other hand, a small standard deviation indicates that these areas are consistent with the overall data fluctuations and can be optimized as a whole.

[0048] Embodiment 2: An intelligent optimization system for optimizing the charging and discharging process of a lithium battery, comprising: The regional monitoring module is divided into multiple monitoring areas according to the design structure of the lithium battery, and each monitoring area is independently monitored to obtain multiple independent monitoring information sets. At the same time, the entire lithium battery is monitored as a whole to obtain an overall monitoring information set; The inertia analysis module performs charge and discharge inertia analysis based on the overall monitoring information set and decides whether to make independent judgments on each monitoring area based on the inertia analysis results; The feature extraction module extracts features from each independent monitoring information set when making independent judgments. After feature extraction, fatigue analysis and degradation analysis are performed. The results of fatigue analysis and degradation analysis are imported into a pre-trained convolutional neural network model to divide each monitoring area into a high disorder area, a medium disorder area, and a low disorder area. The dynamic adjustment module dynamically adjusts all highly disordered areas and moderately disordered areas according to the regional division results, performs performance consistency analysis on the moderately disordered areas and the overall monitoring information set of the lithium battery, and decides whether to adopt refined dynamic optimization or overall dynamic optimization operations; The battery management module manages the batteries in each monitoring area according to the dynamic optimization conditions of the dynamic adjustment module.

[0049] It should be noted that in the prior art, the battery management system (BMS) is mainly responsible for real-time monitoring of the working status of the battery, including parameters such as voltage, current, and temperature. The BMS collects this data to monitor the health status of the battery and regulates the battery according to preset safety thresholds or alarm conditions, such as adjusting the charge and discharge current, limiting the operating voltage range, or activating a protection mechanism to prevent dangerous situations such as overheating and over-discharge.

[0050] In the present invention, the battery management module is optimized and expanded based on the existing BMS functions. It can manage different monitoring areas of the battery in real time according to the conditions of the dynamic adjustment module. Compared with the traditional BMS, the management module in the present invention can not only monitor the basic state parameters of the battery, but also combine the dynamic adjustment conditions to adopt differentiated management strategies for the highly disordered areas, moderately disordered areas and low disordered areas of the battery.

[0051] Data collection and monitoring: The battery management module first collects real-time data from each monitoring area, including key parameters such as current, voltage, temperature, and temperature change rate. By monitoring this data, the system can keep track of the battery's working status at any time.

[0052] Dynamic optimization conditions are temperature safety thresholds that are adjusted in real time: The battery management module calculates the dynamic temperature safety threshold of the area according to the formula, and resets the new temperature safety upper limit based on the values ​​of current, voltage and temperature change rate. The purpose of this is to avoid thermal runaway of the battery due to excessive operating temperature. Once the new dynamic temperature safety threshold is set, the battery management module will ensure that the operating conditions of the battery do not exceed the temperature threshold by adjusting parameters such as charge and discharge current and voltage. For high-risk, highly disordered areas, the battery management module may also take more cautious adjustment measures, such as significantly reducing the charge and discharge current or taking additional heat dissipation measures. After the dynamic optimization operation, the system will continue to monitor the status of each area to ensure that the adjusted battery status is within a safe range. If any further abnormalities are detected, the system will make dynamic adjustments again to maintain the health of the battery.

[0053] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0054] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0055] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0056] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0057] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An intelligent optimization method for optimizing the charging and discharging process of a lithium battery, characterized in that: The following steps are involved: When the lithium battery is charging and discharging, it is divided into multiple monitoring areas according to its design structure, and each monitoring area is independently monitored to obtain multiple independent monitoring information sets, and the lithium battery is monitored as a whole at the same time to obtain an overall monitoring information set; Based on the overall monitoring information set, the charge and discharge inertia analysis is performed and it is decided whether to make an independent judgment. When making an independent judgment, feature extraction is performed on each independent monitoring information set, and fatigue analysis and degradation analysis are performed based on the extracted features; The results of fatigue analysis and degradation analysis are imported into the pre-trained convolutional neural network model, and the monitoring area corresponding to each independent monitoring information set is divided into a high disorder area, a medium disorder area, and a low disorder area; According to the division results, all highly disordered areas and moderately disordered areas are dynamically adjusted, and the overall monitoring information set obtained by overall monitoring of all moderately disordered areas and lithium batteries is analyzed for performance consistency. Based on the analysis results, it is decided to adopt refined dynamic optimization or holistic dynamic optimization operations, and refined dynamic optimization is adopted for all highly disordered areas.

2. The intelligent optimization method for optimizing the charging and discharging process of a lithium battery according to claim 1, characterized in that: Charging and discharging inertia analysis based on the overall monitoring information set refers to: In a fixed time window, obtain each preset charging and discharging inertia parameter in the overall monitoring information set, take the charging and discharging inertia parameter corresponding to one type as the target inertia parameter, and calculate the change rate of the target inertia parameter at adjacent time points. ; i represents the type of inertia parameter, Indicates the sampling time interval corresponding to the inertial parameter type i, Represents the difference of the sampling data of adjacent time points of inertial parameter type i, and obtains a set of change rates The coefficient of variation of the time series Si is calculated for all the charging and discharging inertia parameters of type i.

3. The intelligent optimization method for optimizing the charging and discharging process of a lithium battery according to claim 2, characterized in that: The basis for deciding whether to exercise independent judgment is: If the coefficient of variation of the time series Si is greater than the preset inertia threshold of the charging and discharging inertia parameter of type i, a fluctuation signal is generated; if the coefficient of variation of the time series Si is less than or equal to the preset inertia threshold of the charging and discharging inertia parameter of this type, a normal signal is generated; when there is a fluctuation signal, an independent judgment is made for each monitoring area.

4. The intelligent optimization method for optimizing the charging and discharging process of a lithium battery according to claim 3, characterized in that: When feature extraction is performed on each independent monitoring information set, a thermal fatigue information set and a performance degradation information set are extracted, fatigue analysis is performed based on the thermal fatigue information set to generate a fatigue index, and degradation analysis is performed based on the performance degradation information set to generate a degradation index.

5. The intelligent optimization method for optimizing the charging and discharging process of a lithium battery according to claim 4, characterized in that: The logic for obtaining the fatigue index is: Obtain the thermal fatigue information set of each monitoring area and obtain the collected temperature time series and the corresponding sampling time t, and then the temperature difference between all adjacent data points Sum and average to get the average temperature change rate, the formula is: ; Calculate the second-order derivative of temperature change to reflect the acceleration of temperature fluctuation. Sum the absolute values ​​of acceleration at all times to get the average value of acceleration term. The formula is: ; Set the temperature fluctuation threshold and count the number of times the temperature fluctuation amplitude exceeds the threshold. Indicates the frequency of violent temperature fluctuations in the area; Introducing the time decay factor , giving lower weight to older data to ensure that current data contributes more to the fatigue index, combining the temperature change rate, acceleration term and fluctuation frequency to generate the fatigue index: ; The base time representing the start of time decay, , are preset proportional factors. is the weight coefficient of temperature change acceleration, which is used to adjust the contribution of temperature acceleration to fatigue index. represents the time interval between adjacent data points, N is the amount of sampled data, represents the i-th data point in the temperature time series, Represents the fatigue index. The fatigue index calculation formula performs boundary processing, that is, the last acceleration value calculated at the N-2th point is extended to the N-1th point.

6. The intelligent optimization method for optimizing the charging and discharging process of a lithium battery according to claim 5, characterized in that: The logic for obtaining the degradation index is: Obtain the performance degradation information set of each monitoring area, obtain the battery data at the current time point, including the internal resistance Rt, voltage Vt and capacity Ct, as well as the reference data R0, V0, C0 at the initial moment, and obtain the corresponding internal resistance Ri and voltage Vi from the battery at the past M time points; Calculate the change rate of the internal resistance at the current moment relative to the initial moment, and perform logarithmic mapping after squaring it. The formula is: ; Calculate the battery voltage attenuation rate and add the smoothing term k to avoid distortion in calculations at extremely low voltages. The formula is: ; Calculate the capacity decay rate at the current moment and add the power exponent p to adjust the impact of capacity decay on the overall degradation index. The formula is: ; Perform weighted processing on the battery's past M sampling data and calculate the ratio of internal resistance to voltage for each sampling. , and weighted average it by weight factor wi, the formula is: ; The internal resistance change, voltage decay, capacity decay and multiple sampling data are integrated into the final degradation index. The complete calculation formula is: ； f1, f2, and f3 are all preset influence coefficients. is the degradation index.

7. The intelligent optimization method for optimizing the charging and discharging process of a lithium battery according to claim 6, characterized in that: The results of fatigue analysis and degradation analysis, namely the fatigue index and degradation index, are imported into the pre-trained convolutional neural network model as input data, and the monitoring area corresponding to each independent monitoring information set is divided into high disorder area, moderate disorder area and low disorder area.

8. The intelligent optimization method for optimizing the charging and discharging process of a lithium battery according to claim 7, characterized in that: Fine-grained dynamic optimization is used for all highly disordered areas. This means using the battery management module to dynamically optimize all highly disordered areas and dynamically set the temperature safety threshold of the monomer. The formula is: ; is the preset temperature basic safety threshold, The temperature safety threshold of the monomer is set dynamically. , , Respectively represent the current value, voltage value and temperature change value, , They represent the maximum current and maximum voltage allowed by lithium batteries. , , All are preset safety factors.

9. The intelligent optimization method for optimizing the charging and discharging process of a lithium battery according to claim 8, characterized in that: All moderately disordered areas and lithium batteries are monitored as a whole to obtain a set of overall monitoring information for performance consistency analysis. Based on the analysis results, it is decided to adopt refined dynamic optimization or overall dynamic optimization operations, which means: The current current value, voltage value and temperature change value in the overall monitoring information set obtained by overall monitoring of the moderately disordered area and the lithium battery are respectively obtained and used as a set of corresponding vectors, and the Euclidean distance of the corresponding vectors is calculated. When the standard deviation of all Euclidean distances is greater than the preset performance consistency threshold, all moderately disordered areas are subjected to the same refined dynamic optimization as the highly disordered areas. When the standard deviation of all Euclidean distances is less than or equal to the preset performance consistency threshold, all moderately disordered areas are subjected to overall dynamic optimization. Overall dynamic optimization refers to using the battery management module to dynamically optimize all moderately disordered areas and dynamically set the overall temperature safety threshold. The formula is: ; is the preset temperature basic safety threshold, The overall temperature safety threshold is set dynamically. , , They respectively represent the average current value, average voltage value and average temperature change value of all moderately disordered areas.

10. An intelligent optimization system for optimizing the charging and discharging process of a lithium battery, which is implemented based on the intelligent optimization method for optimizing the charging and discharging process of a lithium battery according to any one of claims 1 to 8, characterized in that: include: The regional monitoring module is divided into multiple monitoring areas according to the design structure of the lithium battery, and each monitoring area is independently monitored to obtain multiple independent monitoring information sets. At the same time, the entire lithium battery is monitored as a whole to obtain an overall monitoring information set; The inertia analysis module performs charge and discharge inertia analysis based on the overall monitoring information set and decides whether to make independent judgments on each monitoring area based on the inertia analysis results. The feature extraction module extracts features from each independent monitoring information set when making independent judgments. After feature extraction, fatigue analysis and degradation analysis are performed. The results of fatigue analysis and degradation analysis are imported into a pre-trained convolutional neural network model to divide each monitoring area into a high disorder area, a medium disorder area, and a low disorder area. The dynamic adjustment module dynamically adjusts all highly disordered areas and moderately disordered areas according to the regional division results, performs performance consistency analysis on the moderately disordered areas and the overall monitoring information set of the lithium battery, and decides whether to adopt refined dynamic optimization or overall dynamic optimization operations; The battery management module manages the batteries in each monitoring area according to the dynamic optimization conditions of the dynamic adjustment module.

Images