Energy storage management method of ternary polymer lithium battery in shared charging equipment

Abstract

The present application relates to the technical field of lithium battery energy storage management, and discloses a method for energy storage management of a ternary polymer lithium battery in a shared charging device, the method comprising: performing multi-condition cycle charge and discharge tests on the ternary polymer lithium battery and fitting the test data to obtain a multivariate capacity decay function; constructing an adaptive composite energy storage system; performing mathematical model analysis to obtain a dynamic model of the energy storage system; generating an initial energy allocation strategy and performing multi-scale time-frequency decomposition and power allocation to obtain a multi-level adaptive power shunt strategy; constructing a neural network-optimized six-input and three-output fuzzy logic controller based on the multi-level adaptive power shunt strategy, and calculating the power regulation coefficients of the ternary polymer lithium battery, supercapacitor and shared charging device, thereby realizing efficient management of the energy storage system of the shared charging device, improving the use efficiency and life of the ternary polymer lithium battery, and meeting the complex load requirements of the shared charging device.

Description

Technical Field

[0001] The present application relates to the technical field of lithium battery energy storage management, and in particular to an energy storage management method for a ternary polymer lithium battery in a shared charging device. Background Art

[0002] Shared charging equipment faces problems such as battery performance degradation and inefficient energy management, which seriously affect the user experience and economic benefits of the equipment. Traditional lithium battery management systems are often unable to effectively cope with the complex usage scenarios and variable load requirements of shared charging equipment, resulting in rapid battery capacity decay, low charging and discharging efficiency, and shortened equipment life.

[0003] In addition, the usage characteristics of shared charging equipment determine that its load power has a high degree of volatility and uncertainty, which puts higher requirements on the dynamic response capability and power allocation strategy of the energy storage system. Most existing energy management methods use a single energy storage unit or a simple control strategy, which is difficult to meet the power requirements of shared charging equipment at different time scales, and cannot effectively balance multiple goals such as system efficiency, energy storage unit life and power balance. Summary of the invention

[0004] The present application provides a method for energy storage management of a ternary polymer lithium battery in a shared charging device, thereby achieving efficient management of the energy storage system of the shared charging device, improving the use efficiency and life of the ternary polymer lithium battery, and meeting the complex load requirements of the shared charging device.

[0005] In a first aspect, the present application provides a method for managing energy storage of a ternary polymer lithium battery in a shared charging device. The method for managing energy storage of a ternary polymer lithium battery in a shared charging device comprises:

[0006] The ternary polymer lithium battery was subjected to multi-condition cycle charge and discharge tests and test data fitting to obtain a multivariate capacity decay function.

[0007] Based on the multivariable capacity decay function, a hybrid energy storage topology structure having a ternary polymer lithium battery, a supercapacitor, a shared charging device, a bidirectional DC / DC converter and a DC bus is constructed to obtain an adaptive composite energy storage system;

[0008] Performing mathematical model analysis on each energy storage unit in the adaptive composite energy storage system to obtain a dynamic model of the energy storage system;

[0009] Generate an initial energy distribution strategy based on the dynamic model of the energy storage system, and perform multi-scale time-frequency decomposition and power distribution on the load power of the adaptive composite energy storage system to obtain a multi-level adaptive power diversion strategy;

[0010] Based on the multi-level adaptive power splitting strategy, a neural network optimized six-input three-output fuzzy logic controller is constructed, and the power regulation coefficients of ternary polymer lithium batteries, supercapacitors and shared charging equipment are calculated.

[0011] A second aspect of the present application provides an energy storage management system for a ternary polymer lithium battery in a shared charging device, wherein the energy storage management system for the ternary polymer lithium battery in the shared charging device comprises:

[0012] The test module is used to perform multi-condition cycle charge and discharge tests on ternary polymer lithium batteries and fit the test data to obtain a multivariate capacity decay function;

[0013] A construction module is used to construct a hybrid energy storage topology structure having a ternary polymer lithium battery, a supercapacitor, a shared charging device, a bidirectional DC / DC converter and a DC bus based on the multivariable capacity decay function to obtain an adaptive composite energy storage system;

[0014] An analysis module, used for performing mathematical model analysis on each energy storage unit in the adaptive composite energy storage system to obtain a dynamic model of the energy storage system;

[0015] A distribution module, used to generate an initial energy distribution strategy based on the dynamic model of the energy storage system, and perform multi-scale time-frequency decomposition and power distribution on the load power of the adaptive composite energy storage system to obtain a multi-level adaptive power diversion strategy;

[0016] The calculation module is used to construct a six-input and three-output fuzzy logic controller optimized by a neural network based on the multi-level adaptive power splitting strategy, and calculate the power regulation coefficients of the ternary polymer lithium battery, supercapacitor and shared charging equipment.

[0017] The third aspect of the present application provides an electronic device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the electronic device executes the above-mentioned energy storage management method of the ternary polymer lithium battery in a shared charging device.

[0018] The fourth aspect of the present application provides a computer-readable storage medium, which stores instructions. When the computer-readable storage medium is run on a computer, it enables the computer to execute the above-mentioned energy storage management method of the ternary polymer lithium battery in a shared charging device.

[0019] Compared with the prior art, the present application has the following beneficial effects: by performing multi-condition cycle charge and discharge tests and data fitting on ternary polymer lithium batteries, the multivariable capacity decay function obtained can accurately describe the relationship between battery capacity and cycle number, temperature and charge and discharge rate. The adaptive composite energy storage system constructed based on the multivariable capacity decay function combines the advantages of ternary polymer lithium batteries, supercapacitors and shared charging equipment, and can effectively respond to load requirements in different frequency ranges, thereby improving the overall performance and reliability of the system. The mathematical model analysis of each energy storage unit in the adaptive composite energy storage system obtains a dynamic model of the energy storage system that can accurately reflect the real-time state of the system. Through multi-scale time-frequency decomposition and power allocation, the multi-level adaptive power diversion strategy obtained can better match the characteristics of different energy storage units, improve energy utilization efficiency, and extend system life. The six-input three-output fuzzy logic controller optimized by neural network is used to realize the intelligent calculation of the power regulation coefficient of ternary polymer lithium batteries, supercapacitors and shared charging equipment, thereby improving the dynamic response capability and control accuracy of the system. The present invention realizes the efficient management of the energy storage system of shared charging equipment, improves the use efficiency and life of ternary polymer lithium batteries, and meets the complex load requirements of shared charging equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0021] The structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with this technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantive technical significance. Any structural modification, change in proportion or adjustment of size, without affecting the effects and purposes that can be achieved by the present invention, should still fall within the scope of the technical contents disclosed by the present invention.

[0022] Figure 1 It is a flow chart of the energy storage management method of the ternary polymer lithium battery in the shared charging device provided by the embodiment of the present invention;

[0023] Figure 2 It is a schematic block diagram of the structure of the energy storage management system of the ternary polymer lithium battery in the shared charging device provided by the embodiment of the present invention. DETAILED DESCRIPTION

[0024] 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 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.

[0025] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0026] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0027] It should be further understood that the term "and / or" used in the specification and appended claims of this application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations. Figure 1 In the embodiment of the present application, an embodiment of the energy storage management method of the ternary polymer lithium battery in the shared charging device includes:

[0028] Step 100, performing multi-condition cycle charge and discharge tests on the ternary polymer lithium battery and fitting the test data to obtain a multivariate capacity decay function;

[0029] It is understandable that the execution subject of the present application can be the energy storage management system of the ternary polymer lithium battery in the shared charging device, or it can be a terminal or a server, which is not limited here. The embodiment of the present application is described by taking the server as the execution subject as an example.

[0030] Specifically, the ternary polymer lithium battery is repeatedly tested for cycle charge and discharge under different temperature ranges and charge and discharge rate conditions, and the battery capacity, cycle number, temperature and charge and discharge rate data in each cycle are recorded to form a complete test data set. The test data set contains the performance changes of lithium batteries under different use conditions, especially the changes in capacity over time and environment. The test data set is normalized to eliminate the differences between different dimensions so that each data is within the same scale range to obtain a normalized data set. Based on this data set, the battery capacity is expressed as a function of the cycle number, temperature and charge and discharge rate. Based on the normalized data set, the battery capacity is expressed as a function of the cycle number, temperature and charge and discharge rate, and a polynomial regression model is constructed to capture the nonlinear relationship between the battery capacity and the cycle number, temperature and rate by introducing polynomial terms. The least squares method is used to estimate the parameters of the polynomial regression model. By minimizing the error between the model prediction value and the actual data, a set of optimal parameters is found to obtain the initial capacity decay function, which is used to preliminarily predict the change trend of the battery capacity. The initial capacity decay function is verified. The k-fold cross-validation method is used to reduce the risk of overfitting of the model by dividing the data into training sets and validation sets multiple times. In each validation, the root mean square error and the coefficient of determination are calculated. These two indicators reflect the average error between the model prediction value and the actual value and the explanatory power of the model, respectively. Based on the validation results, if it is found that there is still room for improvement in the prediction ability of the model, the initial capacity decay function is fine-tuned and the parameters of the model are adjusted to improve the fitting accuracy. Through the adjustment, a multivariate capacity decay function is obtained. The multivariate capacity decay function is used to describe the relationship between the battery capacity and the number of cycles, temperature and charge and discharge rate, and to construct a function model that can accurately describe the battery capacity decay process. The function combines the reference temperature Tref and the reference rate Rref, so that it can accurately predict the change of battery capacity under different conditions of use.

[0031] Step 200: construct a hybrid energy storage topology structure having a ternary polymer lithium battery, a supercapacitor, a shared charging device, a bidirectional DC / DC converter and a DC bus based on a multivariable capacity decay function to obtain an adaptive composite energy storage system;

[0032] Specifically, the remaining capacity of the ternary polymer lithium battery under different working conditions is calculated based on the multivariate capacity decay function. Through the decay function, the remaining capacity of the battery under different temperatures, charge and discharge rates and cycle times is predicted, and the battery capacity state matrix is ​​generated. The battery capacity state matrix is ​​subjected to singular value decomposition, and the matrix is ​​decomposed into multiple simplified parts to construct a battery capacity feature space to describe the change trend of the battery capacity in different dimensions. Based on the battery capacity feature space, the load characteristics of the shared charging device are analyzed to obtain the load power spectrum. Through the analysis of the power spectrum, the power demand of the device in different time periods is understood. Based on the load power spectrum, frequency segmentation is performed to determine the respective operating frequency ranges of the ternary polymer lithium battery, supercapacitor and shared charging device. Lithium batteries are suitable for energy supply at lower frequencies, while supercapacitors are good at processing high-frequency peak power. Through frequency segmentation, the working efficiency of each energy storage unit is optimized. The topology and control strategy of the bidirectional DC / DC converter are created according to the operating frequency range. The converter plays a key role in energy conversion in the entire system, and the selection of its parameters directly affects the stability and efficiency of the energy storage system. Based on the frequency range and energy demand, the key parameter set of the converter is defined. Based on this parameter set, a bus voltage stability controller is created to ensure that the entire system can maintain the voltage stability of the DC bus during operation and ensure that each energy storage unit can work normally. The capacity ratio of the ternary polymer lithium battery, supercapacitor and shared charging device is optimized to obtain the optimal target capacity configuration scheme. Based on the target capacity configuration scheme and the bus voltage stability controller, a hybrid energy storage topology structure including ternary polymer lithium battery, supercapacitor, shared charging device, bidirectional DC / DC converter and DC bus is constructed. In this structure, the ternary polymer lithium battery and the shared charging device are connected to the DC bus through a bidirectional DC / DC converter, and the supercapacitor is used as a peak power buffer unit to handle instantaneous high power demand.

[0033] Step 300: Perform mathematical model analysis on each energy storage unit in the adaptive composite energy storage system to obtain a dynamic model of the energy storage system;

[0034] It should be noted that a third-order Thevenin equivalent circuit model is constructed for the ternary polymer lithium battery in the system. The third-order Thevenin equivalent circuit model can accurately describe the dynamic characteristics of the battery under different loads and working conditions, especially the changes in voltage and internal resistance. In order to improve the accuracy of the model, the temperature coefficient is introduced to correct the open circuit voltage and internal resistance parameters, and the battery voltage equation is obtained to describe the voltage behavior of the battery under dynamic conditions. The charging characteristics of the shared charging device are analyzed, and a second-order RC model containing equivalent resistance and equivalent capacitance is constructed. Through the combination of equivalent resistance and capacitance, this model can reflect the relationship between voltage and current during the charging process of the shared charging device, obtain the voltage-current relationship equation of the device, and describe the dynamic characteristics of the shared charging device during charging and discharging. According to the bidirectional DC / DC converter, a converter transient model including a switching function is established. The transient model is processed by the state space averaging method to obtain a nonlinear average model of the converter, which describes the energy conversion process of the bidirectional DC / DC converter under different working conditions. The state space equation of the adaptive composite energy storage system is constructed by combining the battery voltage equation, the voltage-current relationship equation and the nonlinear average model of the converter, which systematically represents the dynamic characteristics and interactions of the battery, capacitor and converter. An extended Kalman filter is created based on the state space equation. The filter is used to estimate the state and unknown parameters of the system in real time. The state variables and unknown parameters of the adaptive composite energy storage system are combined into a vector to be estimated. Through the iterative calculation of the extended Kalman filter, these state variables and parameters are continuously estimated and updated to obtain the optimal estimated values ​​of the state and parameters of the system. The estimation error covariance matrix is ​​calculated based on the estimated values ​​of the system state and parameters. The adaptive weighting strategy is used to dynamically adjust the process noise covariance and the measurement noise covariance to obtain the optimal target Kalman gain. The target Kalman gain is used to correct the system state and parameters to obtain an accurate dynamic model of the energy storage system.

[0035] Step 400: Generate an initial energy allocation strategy based on the dynamic model of the energy storage system, and perform multi-scale time-frequency decomposition and power allocation on the load power of the adaptive composite energy storage system to obtain a multi-level adaptive power diversion strategy;

[0036] Specifically, the corresponding energy management objective function is constructed based on the dynamic model of the energy storage system. The objective function mainly includes three core elements: system efficiency, energy storage unit life, and power balance constraints. Through this objective function, it is ensured that while optimizing energy distribution, the service life of the energy storage unit is maximized and the power balance of the system is maintained. The energy management objective function is solved based on the dynamic programming algorithm to calculate the optimal energy distribution sequence under different time scales. The dynamic programming algorithm can effectively handle the time series problem in the energy storage system and ensure that the energy distribution strategy in each time period is optimal. The initial energy distribution strategy is obtained through calculation. The load power of the adaptive composite energy storage system is transformed by wavelet packet transform to decompose the signal into different frequency components. By selecting the optimal basis function and decomposition layer, the load power signal is decomposed into multiple wavelet coefficients. Threshold processing is performed based on the multi-scale wavelet coefficients to remove noise and irrelevant information, and then the wavelet coefficients are reconstructed into the initial low-frequency power component, the intermediate frequency power component, and the high-frequency power component. The initial low-frequency power component is subjected to empirical mode decomposition to decompose the low-frequency component into multiple eigenmode functions and a residual function. Based on the frequency characteristics of the intrinsic mode function and the residual function, they are recombined to obtain the target low-frequency power component, the target medium-frequency power component, and the target high-frequency power component. According to the initial energy allocation strategy, different power components are reasonably allocated to each energy storage unit. The target low-frequency power component is allocated to the ternary polymer lithium battery, because the low-frequency energy demand is usually relatively stable, which is suitable for being carried by a battery, an energy storage unit with a large capacity; the target medium-frequency power component is allocated to the supercapacitor, which can quickly respond to medium-frequency power demand and has a high charging and discharging efficiency; the target high-frequency power component is allocated to the shared charging device, which is mainly used to meet the power demand of the device at high frequency changes. Finally, a multi-level adaptive power diversion strategy is obtained to achieve optimized collaboration between ternary polymer lithium batteries, supercapacitors, and shared charging devices.

[0037] The energy management objective function is discretized, and the continuous state space and decision space are divided into a finite number of discrete points. The discrete state transfer equation is obtained to describe how the system state transitions from one discrete point to another discrete point. Based on the discrete state transfer equation, a state transfer matrix is ​​constructed. This matrix is ​​used to represent the possible transfer paths of the system from the current state to the next state, and the corresponding cost function value is calculated for each state transfer path. The cost function value reflects factors such as energy consumption, energy storage unit life impact, and system power balance during each state transfer process. By calculating all possible state transfer costs, a cost matrix is ​​obtained, which contains the corresponding costs of all state transfers. The cost matrix is ​​back-recursively calculated, and the optimal value function at each moment is calculated step by step from the termination moment to obtain the optimal value function sequence, which provides the optimal choice in each state from the termination moment to the initial moment. A forward search is performed based on the optimal value function sequence. In the forward search process, the optimal decision at each moment is determined according to the optimal value function, that is, the optimal energy allocation strategy at the current moment, and the search energy allocation strategy is obtained. The search energy allocation strategy is interpolated so that the strategies between discrete points can be continuous, generating a continuous energy allocation curve. Based on the continuous energy distribution curve, the system state variables at each moment are calculated to obtain the target operation trajectory, which describes the operating status of the entire energy storage system at different time points. In order to ensure the safety and stability of the energy storage system in actual operation, a constraint check is performed on the target operation trajectory to ensure that the state of charge and power of each energy storage unit operate within its limits. If it is found that the constraints are not met, corresponding adjustments and corrections are required. The energy distribution sequence at different time scales is extracted. This sequence describes the energy distribution of ternary polymer lithium batteries, supercapacitors and shared charging devices at various time points. After dynamic programming optimization processing, the initial energy distribution strategy is obtained, which can maximize the efficiency and life of the energy storage system while meeting the system operation constraints.

[0038] Step 500: construct a neural network optimized six-input three-output fuzzy logic controller based on a multi-level adaptive power splitting strategy, and calculate the power regulation coefficients of the ternary polymer lithium battery, supercapacitor and shared charging device.

[0039] Specifically, the state of charge of the ternary polymer lithium battery, the state of charge of the supercapacitor, the state of charge of the shared charging device, the load power, the first-order change rate and the second-order change rate of the load power are taken as six input variables. Through these six key variables, the dynamic characteristics of the energy storage unit and the change trend of the load are captured. The six input variables are fuzzified. A membership function is created for each input variable, and the membership function defines the correspondence between the variable value and the fuzzy set. Through the membership function, the original precise data is converted into a fuzzy set to obtain the fuzzy set of the input variable, and an initial fuzzy rule base is created based on the multi-level adaptive power diversion strategy. The initial fuzzy rule base contains the response rules of the system under different input conditions, and constructs the mapping relationship from the fuzzy set of the input variable to the fuzzy set of the output variable. Based on the initial fuzzy rule base, the maximum-minimum synthesis method is used for fuzzy reasoning. The maximum-minimum synthesis method is a fuzzy reasoning method that obtains the fuzzy set of the output variable by synthesizing the maximum and minimum values ​​of the input fuzzy set. The output fuzzy set is defuzzified, and the exact value of the output variable is calculated by the centroid method. The centroid method is a defuzzification technique that can convert fuzzy output into precise values ​​to obtain the three output variables of the initial fuzzy logic controller, namely, the power regulation coefficient of the ternary polymer lithium battery, supercapacitor and shared charging device. In order to improve the performance of the fuzzy logic controller, a three-layer feedforward neural network is constructed, and the input data and output data of the initial fuzzy logic controller are used as training samples of the neural network. In this way, the fuzzy logic controller is optimized by utilizing the adaptive learning ability of the neural network. The neural network is trained using the back propagation algorithm, which minimizes the error by continuously adjusting the weights of the neural network to obtain the optimized neural network model. Based on the trained neural network model, the initial fuzzy rule base is optimized, and the accuracy and response speed of the rule base are improved by adjusting the weights and structure of the rules. Based on the training results of the neural network, the membership function is fine-tuned to more accurately describe the relationship between the input variables and the fuzzy set. By optimizing the rule base and the membership function, a neural network-optimized six-input and three-output fuzzy logic controller is constructed. In the actual operation process, the state of charge data of the ternary polymer lithium battery, supercapacitor and shared charging device are collected in real time, and the real-time value of the load power is obtained at the same time. Based on these real-time data, the real-time first-order and second-order change rates of the load power are calculated, and the six input variables are input into the fuzzy logic controller optimized by the neural network. The controller processes the input data through the fuzzy rules and membership functions optimized by the neural network, outputs the power regulation coefficients of the ternary polymer lithium battery, supercapacitor and shared charging equipment, and realizes adaptive power allocation and optimization control between energy storage units.

[0040] Collect real-time state of charge data of ternary polymer lithium batteries, supercapacitors and shared charging devices. These data include the real-time state of charge information of each energy storage unit and the real-time value of the load power of the current shared charging device. Through these data, the current state of the system is accurately evaluated. The real-time first-order rate of change and second-order rate of change of load power are calculated to capture the trend of load power changes over time. The rate of change helps to determine the amplitude and speed of load fluctuations, thereby optimizing the response strategy of power distribution. After integrating all these data, standardized input data is obtained. The standardized input data is input into the six-input and three-output fuzzy logic controller optimized by the neural network. The controller performs fuzzification on each input variable. The membership value of the input variable is calculated using the target membership function. The membership function maps the input variable to the fuzzy set to obtain the fuzzy input set. The fuzzy input set reflects the degree of input data under different fuzzy rules for subsequent fuzzy reasoning. According to the fuzzy input set, rule matching and activation are performed based on the pre-built target fuzzy rule base. The process of rule matching is to compare the fuzzy set of the input variable with the rules in the fuzzy rule base to find the rules that meet the conditions. After activating these rules, the contribution of each rule to the final output is calculated according to the output strength of each rule to obtain the fuzzy output set. The fuzzy output set is defuzzified based on the centroid method. By calculating the centroid of the fuzzy output set, the fuzzy output value can be converted into an accurate value, and the power regulation coefficient of the ternary polymer lithium battery, supercapacitor and shared charging device can be calculated. The power regulation coefficient is used to adjust the power distribution between the energy storage units in real time to ensure that the system can operate efficiently under different load conditions, while optimizing the service life and performance of the energy storage unit.

[0041] Furthermore, the fuzzy output set is discretized. The continuous membership values ​​in the fuzzy output set are converted into a discrete membership value sequence to form a discretized fuzzy output set. Based on the discretized fuzzy output set, the product of the membership value of each discrete point and its corresponding output value is calculated. The product represents the weight of each discrete point in the fuzzy output, and a weighted membership value sequence is obtained by calculating the weighted membership value. The sequence links different fuzzy memberships with corresponding output values ​​to ensure that the influence of each discrete point is taken into account during the defuzzification process. The weighted membership value sequence is accumulated to calculate the sum of all weighted membership values. At the same time, the membership value sequence in the discretized fuzzy output set is accumulated to obtain the sum of membership values. The sum of membership values ​​represents the overall contribution of the membership of all fuzzy output points in the defuzzification process. The centroid value is calculated by dividing the sum of weighted membership values ​​by the sum of membership values. The centroid method is a defuzzification method used in fuzzy control. It obtains an accurate output value by calculating the weighted average of all discrete points in the fuzzy output set. The centroid value reflects the final output of the fuzzy controller, that is, the power regulation ratio between the ternary polymer lithium battery, supercapacitor and shared charging device. Data smoothing is performed based on the initial defuzzification result to eliminate sudden fluctuations and obtain a more stable target defuzzification result. The target defuzzification result is normalized to ensure that the sum of the power regulation coefficients of the ternary polymer lithium battery, supercapacitor and shared charging device is 1. Normalization can ensure that the system can maintain a balanced distribution of energy under any load conditions, so that different energy storage units can reasonably allocate power according to demand. The power regulation coefficients of the ternary polymer lithium battery, supercapacitor and shared charging device are obtained.

[0042] In the embodiment of the present application, by performing multi-condition cycle charge and discharge tests and data fitting on the ternary polymer lithium battery, the multivariable capacity decay function obtained can accurately describe the relationship between the battery capacity and the number of cycles, temperature and charge and discharge rate. The adaptive composite energy storage system constructed based on the multivariable capacity decay function combines the advantages of ternary polymer lithium batteries, supercapacitors and shared charging equipment, and can effectively respond to load requirements in different frequency ranges, thereby improving the overall performance and reliability of the system. The mathematical model analysis of each energy storage unit in the adaptive composite energy storage system is carried out, and the obtained dynamic model of the energy storage system can accurately reflect the real-time state of the system. Through multi-scale time-frequency decomposition and power allocation, the obtained multi-level adaptive power diversion strategy can better match the characteristics of different energy storage units, improve energy utilization efficiency, and extend the life of the system. The six-input three-output fuzzy logic controller optimized by neural network is adopted to realize the intelligent calculation of the power regulation coefficient of ternary polymer lithium batteries, supercapacitors and shared charging equipment, and improve the dynamic response capability and control accuracy of the system. The present invention realizes the efficient management of the energy storage system of shared charging equipment, improves the use efficiency and life of ternary polymer lithium batteries, and meets the complex load requirements of shared charging equipment.

[0043] In a specific embodiment, the process of executing step 100 may specifically include the following steps:

[0044] Perform cyclic charge and discharge tests and collect test data on ternary polymer lithium batteries at different temperature ranges and charge and discharge rates to obtain a test data set, which includes battery capacity, cycle number, temperature, and charge and discharge rate data;

[0045] The test data set is normalized to obtain a normalized data set, and based on the normalized data set, the battery capacity is expressed as a function of the number of cycles, temperature, and charge and discharge rate, and a polynomial regression model is constructed;

[0046] The least square method was used to estimate the parameters of the polynomial regression model to obtain the initial capacity decay function, and the initial capacity decay function was verified based on the k-fold cross-validation method, and the root mean square error and determination coefficient were calculated to obtain the verification results;

[0047] Based on the verification results, the initial capacity decay function is fine-tuned to obtain a multivariate capacity decay function. The multivariate capacity decay function is used to describe the relationship between battery capacity and cycle number, temperature and charge and discharge rate. The multivariate capacity decay function is:

[0048] ;

[0049] Among them, C (n, T, R) represents the capacity of the battery under target conditions, C0 represents the initial capacity of the battery, n represents the number of cycles, T represents the temperature, R represents the charge and discharge rate, e represents the base of the natural logarithm, a1, a2, b1, b2 are parameters that describe the capacity attenuation with the number of cycles, k1 is the temperature influence coefficient, k2 is the rate influence coefficient, k3 is the interaction influence coefficient of the number of cycles, temperature and rate, Tref is the reference temperature, and Rref is the reference rate.

[0050] Specifically, the ternary polymer lithium battery is subjected to cyclic charge and discharge tests at different temperature ranges and charge and discharge rates, and the test data are collected. The performance changes of the battery are observed by controlling the experimental conditions. Temperature, charge and discharge rate and number of cycles are key variables affecting the battery capacity. In the experiment, by setting different temperature ranges (such as 25°C, 40°C, 60°C, etc.) and different charge and discharge rates (such as 1C, 2C, 3C, etc.), the battery is repeatedly charged and discharged cycle tested, and the battery capacity after each cycle and the corresponding temperature and rate are recorded to form a test data set containing battery capacity, number of cycles, temperature and charge and discharge rate data. The test data set is normalized to obtain a normalized data set. Data of different scales are converted to the same scale to avoid excessive influence of some large-scale data on the model. A common normalization method is to scale the data according to its minimum and maximum values. After normalization, a normalized data set is obtained. Based on the normalized data set, the battery capacity is expressed as a function of the number of cycles, temperature and charge and discharge rate. In order to construct this function, a polynomial regression model is used. The polynomial regression model can effectively capture the nonlinear relationship between battery capacity and variables such as cycle number, temperature, and rate. The specific model form can be:

[0051] ;

[0052] Among them, C(n, T, R) represents the battery capacity under the conditions of cycle number n, temperature T and charge and discharge rate R, CO is the initial capacity of the battery, a1 and a2 are coefficients describing the capacity decay with the cycle number, b1 and b2 are decay rate parameters, k1 is the influence coefficient of temperature on capacity, k2 is the influence coefficient of rate on capacity, and k3 is the interaction coefficient between cycle number, temperature and rate. Tref and Rref are the reference temperature and reference charge and discharge rate, respectively, which are used to standardize the effects of temperature and rate. The polynomial regression model can fit a reasonable battery capacity change trend based on the experimental data, reflecting the performance changes of the battery under different environmental conditions and usage conditions. In order to estimate the parameters of the polynomial regression model, the least squares method is used. The least squares method is a standard method for fitting data. The optimal model parameters are found by minimizing the error between the model prediction value and the actual test data. Parameters such as a1, a2, b1, b2, k1, k2, and k3 are estimated to ensure that the model can reflect the decay trend of the battery capacity with the cycle number, temperature, and charge and discharge rate as accurately as possible. After obtaining the initial capacity decay function, the model is validated. The validation steps include using the k-fold cross validation method. The k-fold cross validation method divides the data set into k subsets, of which k-1 subsets are used to train the model and the remaining subset is used to test the performance of the model. By repeating this process many times, it is ensured that the performance of the model on different data subsets has good generalization ability. In each validation, the root mean square error (RMSE) and the coefficient of determination (R²) are calculated. RMSE is used to evaluate the error between the model's predicted value and the true value, and R² reflects the model's ability to interpret the data. Ideally, the smaller the RMSE, the better, and the closer R² is to 1, the better. If it is found that there is room for improvement in the performance of the model during the validation, the initial capacity decay function is fine-tuned based on the validation results. The accuracy of the model is improved by adjusting the parameters of the model, such as the attenuation coefficients such as a1, a2, b1, b2 and the environmental impact coefficients such as k1 and k2. Through the above steps, a multivariate capacity decay function is obtained to accurately describe the relationship between the battery capacity and the number of cycles, temperature and charge and discharge rate.

[0053] In a specific embodiment, the process of executing step 200 may specifically include the following steps:

[0054] Based on the multivariate capacity decay function, the remaining capacity of the ternary polymer lithium battery under different working conditions is calculated and the battery capacity state matrix is ​​generated. The battery capacity state matrix is ​​subjected to singular value decomposition to construct the battery capacity feature space.

[0055] The load characteristics of the shared charging device are analyzed based on the battery capacity feature space to obtain the load power spectrum, and the frequency is segmented based on the load power spectrum to determine the operating frequency range of the ternary polymer lithium battery, supercapacitor and shared charging device;

[0056] A topology and control strategy of a bidirectional DC / DC converter is created according to the operating frequency range, a converter parameter set is obtained, and a bus voltage stability controller is created based on the converter parameter set;

[0057] Optimize the capacity ratio of ternary polymer lithium batteries, supercapacitors and shared charging equipment to obtain the target capacity configuration plan;

[0058] Based on the target capacity configuration scheme and the bus voltage stability controller, a hybrid energy storage topology structure including a ternary polymer lithium battery, a supercapacitor, a shared charging device, a bidirectional DC / DC converter and a DC bus is constructed to obtain an adaptive composite energy storage system. The adaptive composite energy storage system connects the ternary polymer lithium battery and the shared charging device to the DC bus through a bidirectional DC / DC converter, and uses the supercapacitor as a peak power buffer unit.

[0059] Specifically, the remaining capacity of the ternary polymer lithium battery under different working conditions is calculated based on the multivariate capacity decay function, and a battery capacity state matrix is ​​generated. The rows and columns of the matrix correspond to different cycle times, temperatures, and charge and discharge rates, respectively. Each element in the matrix represents the remaining capacity of the battery under the corresponding conditions. The generated battery capacity state matrix is ​​subjected to singular value decomposition. Singular value decomposition is a matrix decomposition method that can decompose a complex multidimensional matrix into a number of basis vectors and singular values ​​with independent characteristics. Through singular value decomposition, the battery capacity feature space, that is, the characteristic mode of the battery capacity under different working conditions, is obtained. These characteristic modes reveal the main change trends of the battery in dimensions such as temperature, rate, and number of cycles. By analyzing these characteristics, capacity management is simplified and the key behaviors of the battery under different working conditions are revealed. Based on the battery capacity feature space, the load characteristics of the shared charging device are analyzed. The load characteristics of the shared charging device include the power demand fluctuations of the device under different time and environmental conditions. In order to quantify the load characteristics, the load power spectrum is obtained through power analysis. The load power spectrum reflects the distribution of the power demand of the device in the frequency domain. Typically, the high frequency part corresponds to the instantaneous power demand of the device, while the low frequency part represents the stable long-term load demand. By frequency segmenting the load power spectrum, the operating frequency range of ternary polymer lithium batteries, supercapacitors, and shared charging devices is determined. Ternary polymer lithium batteries are suitable for low-frequency loads because they have a large capacity but a slow response speed; while supercapacitors are suitable for handling high-frequency loads because they have a fast charging and discharging speed and are suitable for instantaneous power demand. Through this segmentation, the power distribution of each energy storage unit in the system is optimized so that each energy storage unit operates within its optimal operating frequency range. Based on the operating frequency range, a topology and control strategy are created for the bidirectional DC / DC converter. The bidirectional DC / DC converter is the core component that connects the energy storage unit to the shared charging device and is responsible for bidirectional energy conversion between different energy storage units. The optimal converter parameter set is designed through load frequency-based analysis. The parameter set includes switching frequency, input and output voltage range, power control strategy, etc. In order to ensure the stability of the entire energy storage system, a bus voltage stability controller is created based on the converter parameter set. The function of this controller is to maintain the voltage stability of the DC bus and avoid drastic changes in the bus voltage due to load fluctuations or energy exchange between energy storage units. The capacity ratio of ternary polymer lithium batteries, supercapacitors and shared charging equipment is optimized to ensure that the capacity of each energy storage unit is reasonably allocated so that it can operate under optimal conditions. Ternary polymer lithium batteries are suitable for providing large energy reserves, while supercapacitors are used to handle instantaneous high power demands. Therefore, in the optimization of capacity ratio, the working characteristics of the two need to be balanced. Through the optimized capacity configuration plan, the utilization efficiency of the energy storage system is maximized, while the service life of batteries and capacitors is extended.Based on the optimized capacity configuration scheme and bus voltage controller, a hybrid energy storage topology consisting of ternary polymer lithium batteries, supercapacitors, shared charging devices, bidirectional DC / DC converters and DC buses is constructed. In this topology, the ternary polymer lithium batteries and shared charging devices are connected to the DC bus through a bidirectional DC / DC converter. The supercapacitor acts as a peak power buffer unit, responsible for providing a fast response when the load has an instantaneous power demand. The battery is responsible for providing long-term, stable energy output, while the supercapacitor smoothes load fluctuations through rapid charging and discharging capabilities, avoiding accelerated aging of the battery due to frequent high-power output.

[0060] In a specific embodiment, the process of executing step 300 may specifically include the following steps:

[0061] The third-order Thevenin equivalent circuit model of the ternary polymer lithium battery in the adaptive composite energy storage system is constructed, and the temperature coefficient is introduced to correct the open circuit voltage and internal resistance parameters to obtain the battery voltage equation;

[0062] The charging characteristics of the shared charging equipment are analyzed, and a second-order RC model including equivalent resistance and equivalent capacitance is constructed to obtain the voltage-current relationship equation;

[0063] According to the bidirectional DC / DC converter, the converter transient model including the switching function is established, and the converter nonlinear average model is obtained by the state space averaging method.

[0064] The state space equation of the adaptive composite energy storage system is constructed by combining the battery voltage equation, the voltage-current relationship equation and the converter nonlinear average model.

[0065] An extended Kalman filter is created based on the state space equation, the state variables and unknown parameters of the adaptive composite energy storage system are combined into a vector to be estimated, and the vector to be estimated is iteratively calculated to obtain the estimated values ​​of the system state and parameters;

[0066] The estimated error covariance matrix is ​​calculated based on the estimated values ​​of the system state and parameters, and the process noise covariance and measurement noise covariance are dynamically adjusted using an adaptive weighting strategy to obtain the target Kalman gain. The system state and parameters are corrected based on the target Kalman gain to obtain the dynamic model of the energy storage system.

[0067] Specifically, a third-order Thevenin equivalent circuit model of the ternary polymer lithium battery is constructed. The Thevenin model is a classic circuit model used to simulate the internal dynamic characteristics of the battery, which can characterize the dynamic behavior of the battery through simple resistance, capacitance and voltage source. In the third-order Thevenin equivalent circuit, the internal structure of the battery is modeled as a combination of three RC circuits, each of which corresponds to a different dynamic response of the battery (short-term, medium-term and long-term characteristics). This model usually includes an open circuit voltage source (OCV), a dynamic internal resistance in series and a capacitor in parallel. In order to ensure that the model can accurately reflect the performance of the battery in actual use, a temperature correction coefficient is introduced to perform temperature compensation for the open circuit voltage and internal resistance. The effect of temperature on the performance of lithium batteries is very significant, so the temperature correction coefficient dynamically adjusts these parameters to ensure that the model can still accurately simulate the state of the battery under different temperature conditions. Through the equivalent circuit model, the voltage equation of the battery is obtained, which is expressed as:

[0068] ;

[0069] in, is the terminal voltage of the battery, is the open circuit voltage, which is a function of temperature T, is the discharge current of the battery, is the temperature-corrected internal resistance of the battery, and Represent the equivalent resistance and time constant respectively, reflecting the dynamic response inside the battery. The charging characteristics of the shared charging device are analyzed. In order to describe its charging and discharging process, a second-order RC model is constructed, which contains equivalent resistance and equivalent capacitance. This model is used to simulate the voltage-current relationship of the charging device during charging. Unlike batteries, the characteristics of shared charging devices are more reflected in their resistive and capacitive characteristics. Through the combination of RC circuits, the dynamic behavior of the device during charging and discharging is captured. Based on this analysis, the voltage-current relationship equation of the device is obtained:

[0070] ;

[0071] in, is the voltage of the device, is the charging and discharging current of the device, is the equivalent resistance of the device, is the equivalent capacitance of the device. A transient model is established for a bidirectional DC / DC converter. The DC / DC converter is used to convert energy between batteries, supercapacitors and shared charging devices. Since the working principle of the converter depends on the switching of high-speed switches, its transient model can be described by including the switching function. The state space averaging method is used to convert the transient model into a nonlinear average model. The state space averaging method is to obtain a simplified model by averaging the dynamic behavior of the converter over multiple switching cycles, thereby describing the overall performance of the converter without being affected by the instantaneous switching state. The battery voltage equation, the voltage-current relationship equation and the nonlinear average model of the converter are combined to construct the state space equation of the adaptive composite energy storage system. The state space equation is a system dynamic equation that can uniformly describe the dynamic changes of variables such as voltage, current, and power in the entire energy storage system. Through this equation, the state changes of batteries and shared charging devices are tracked in real time. An extended Kalman filter is created based on the state space equation. The extended Kalman filter is a state estimation method for nonlinear systems. By estimating and updating the system state variables and unknown parameters in real time, the influence of system noise and uncertainty can be effectively reduced. The state variables (such as voltage, power, etc.) and unknown parameters (such as the internal resistance of the battery, the capacitance of the shared charging device, etc.) of the adaptive composite energy storage system are combined into a vector to be estimated, and the vector is iteratively calculated through the Kalman filter. During the iteration process, the extended Kalman filter continuously updates the estimated values ​​of the system state and parameters according to the difference between the actual output of the system and the model prediction value. This process can greatly improve the accuracy of the model, enabling it to provide accurate state prediction and system control information under actual working conditions with large uncertainties. The extended Kalman filter introduces the estimation error covariance matrix, which is used to quantify the error between the estimated value and the true value. In order to optimize the performance of the Kalman filter, an adaptive weighting strategy is adopted to dynamically adjust the process noise covariance and measurement noise covariance. The process noise covariance reflects the uncertainty of the system during dynamic operation, while the measurement noise covariance reflects the possible errors in the measurement process. By adaptively adjusting these covariance matrices, the target Kalman gain of the Kalman filter is always in the optimal state. The Kalman gain is used to balance the weight between the model prediction value and the measurement value, so as to ensure that the final state estimate has the minimum mean square error. Based on the target Kalman gain, the system state and parameters are corrected to obtain the dynamic model of the adaptive composite energy storage system. This dynamic model can not only accurately describe the energy conversion process between the battery, shared charging equipment and DC / DC converter, but also adapt to the changes in the system state in real time. By extending the dynamic correction mechanism of the Kalman filter, it ensures that the system can operate in the optimal way under various working conditions.

[0072] In a specific embodiment, the process of executing step 400 may specifically include the following steps:

[0073] Based on the dynamic model of the energy storage system, the corresponding energy management objective function is constructed, which includes system efficiency, energy storage unit life and power balance constraints;

[0074] Based on the dynamic programming algorithm, the optimal energy allocation sequence under different time scales is calculated for the energy management objective function to obtain the initial energy allocation strategy;

[0075] The load power of the adaptive composite energy storage system is transformed by wavelet packet, the optimal basis function and decomposition layer number are selected to obtain multi-scale wavelet coefficients, and threshold processing is performed based on the multi-scale wavelet coefficients to reconstruct the wavelet coefficients into initial low-frequency power components, initial intermediate-frequency power components and initial high-frequency power components;

[0076] Performing empirical mode decomposition on the initial low-frequency power component to obtain multiple eigenmode functions and a residual function, and recombining them into a target low-frequency power component, a target intermediate-frequency power component and a target high-frequency power component based on the frequency characteristics of the eigenmode function and the residual function;

[0077] According to the initial energy allocation strategy, the target low-frequency power component is allocated to the ternary polymer lithium battery, the target medium-frequency power component is allocated to the supercapacitor, and the target high-frequency power component is allocated to the shared charging device, thus obtaining a multi-level adaptive power splitting strategy.

[0078] Specifically, the corresponding energy management objective function is constructed based on the dynamic model of the energy storage system to optimize the overall efficiency of the system, while maximizing the life of the energy storage unit and ensuring the power balance of the system under different loads. The energy management objective function usually consists of three main parts: system efficiency, energy storage unit life, and power balance constraints. System efficiency reflects the loss of the energy storage unit during energy conversion and storage, the life of the energy storage unit depends on its working state under different working conditions, and the power balance constraint ensures that the system can maintain energy supply and demand balance under various load conditions. In order to achieve the optimization goal, an energy management objective function is defined:

[0079] ;

[0080] in, represents the objective function value at time t, For system efficiency, and is the weight coefficient used to balance system efficiency, energy storage unit health status and power balance constraints, is the load power, is the total power supplied by the energy storage unit. The design purpose of the objective function is to maximize the system efficiency and extend the service life of the battery and supercapacitor under the premise of meeting the load demand. Based on the objective function, the dynamic programming algorithm is used to calculate the optimal energy allocation sequence at different time scales. The dynamic programming algorithm is suitable for multi-stage decision-making problems and can recursively calculate the optimal decision at each time point to optimize the overall objective function. The core of dynamic programming is to decompose complex optimization problems into a series of smaller sub-problems. By solving these sub-problems, the optimal energy allocation strategy for the entire system is gradually obtained. At each time scale, the algorithm calculates the optimal energy allocation sequence based on the system efficiency, the health status of the energy storage unit, and the constraints of power balance to obtain the initial energy allocation strategy. The load power of the adaptive composite energy storage system is transformed by wavelet packet transform, and the optimal basis function and decomposition layer number are selected to obtain the multi-scale wavelet coefficients of the load power. When selecting the basis function, the Daubechies wavelet function or Haar wavelet function suitable for the characteristics of the power signal can be selected. The number of decomposition layers determines the ability to capture details of the signal, usually 3 to 5 layers. By decomposition, the load power is decomposed into multiple frequency components. Threshold processing is performed based on multi-scale wavelet coefficients to remove noise and irrelevant high-frequency components. The wavelet coefficients are reconstructed into initial low-frequency power components, initial intermediate-frequency power components, and initial high-frequency power components. Low-frequency components usually represent stable, long-term load demands, while intermediate-frequency and high-frequency components correspond to load fluctuations and instantaneous power demands in a shorter period of time. Empirical mode decomposition is performed on the initial low-frequency power components. Empirical mode decomposition is an adaptive signal processing technique that can decompose complex signals into a series of intrinsic mode functions (IMFs), each of which represents a different frequency component in the signal. Through empirical mode decomposition, the low-frequency power component is decomposed into multiple IMFs and a residual function. IMFs are used to represent the main frequency characteristics in the signal, while the residual function reflects the low-frequency trend in the signal. Based on the frequency characteristics of the intrinsic mode function and the residual function, the optimized target low-frequency power component, target intermediate-frequency power component, and target high-frequency power component are recombined. The low-frequency power component mainly corresponds to the power supply requirements of the ternary polymer lithium battery, and the lithium battery is more suitable for handling stable long-term power output. The intermediate-frequency power component is allocated to the supercapacitor, which can quickly respond to medium-frequency power demands. The high-frequency power component is distributed to shared charging devices, where there may be instantaneous high power demands, and supercapacitors can discharge quickly to meet these demands.

[0081] In a specific embodiment, the execution step calculates the optimal energy allocation sequence under different time scales based on the energy management objective function using a dynamic programming algorithm, and the process of obtaining the initial energy allocation strategy may specifically include the following steps:

[0082] Discretize the energy management objective function, divide the continuous state space and decision space into a finite number of discrete points, and obtain the discrete state transfer equation;

[0083] Construct a state transfer matrix based on the discrete state transfer equation, calculate the cost function value of each state transfer, and obtain the cost matrix;

[0084] Perform backward recursion on the cost matrix, and gradually calculate the optimal value function at each moment starting from the termination moment to obtain the optimal value function sequence;

[0085] Based on the optimal value function sequence, forward search is performed to determine the optimal decision at each moment, and the search energy allocation strategy is obtained. The search energy allocation strategy is interpolated to make the energy allocation strategy between discrete points continuous to obtain a continuous energy allocation curve.

[0086] Based on the continuous energy allocation curve, the state variables at each moment are calculated to obtain the target operation trajectory, and the target operation trajectory is constrained to ensure that the charge state limit and power limit of each energy storage unit are met. The energy allocation sequence at different time scales is extracted to obtain the initial energy allocation strategy.

[0087] Specifically, the continuous state space and decision space are discretized. In the energy storage system, the state space represents the different energy states of the system, such as the state of charge of the battery, while the decision space represents the energy allocation decision of the system at different time points. Since the state space and decision space in practical problems are continuous, they are divided into a finite number of discrete points for calculation and optimization in discrete scenarios. Through discretization, the discrete state transfer equation is obtained to describe the behavior of the system transferring from one state to another. The basic form of the state transfer equation can be expressed as:

[0088] ;

[0089] in, is the state of the system at time t, is the decision variable of the system at time t (e.g., the power allocation of the battery), and Represents the state transfer function, which describes how the state of the system at time t+1 is determined by the current state and decision. Through discretization, the state space is divided into several discrete points. Based on the discrete state transfer equation, a state transfer matrix is ​​constructed. Each item in the state transfer matrix represents the cost of transferring from one state to another. In order to obtain the cost matrix, the corresponding cost function value is calculated for each state transition. The cost function is usually related to factors such as system efficiency, life of the energy storage unit, and power balance constraints. In each state, the cost function can be in the form of:

[0090] ;

[0091] in, Indicates from the state By decision The cost incurred, represents the efficiency of the system, SoH is the health status of the energy storage unit, is the load power, is the total power supplied by the system, and is the weight coefficient in the cost function, which is used to adjust the trade-off between efficiency, life and power balance. The cost matrix is ​​calculated backward recursively. Backward recursion is a process of deriving the optimal value function from the termination moment forward. Starting from the termination moment, the optimal value function of each state is calculated, and the cumulative cost from the current state to the termination moment is recorded. Through recursion, the optimal value function of each moment is calculated forward step by step. The optimal value function sequence is obtained, which represents the optimal value and its corresponding decision at different time points. Forward search is performed based on the optimal value function sequence. Through the optimal value function sequence, the optimal decision at each moment is gradually determined. At each moment, a search energy allocation strategy is constructed by selecting the decision that minimizes the cost according to the optimal value function. This strategy describes how the system allocates power between different energy storage units at each time point. The search energy allocation strategy is interpolated to ensure that the energy allocation strategy is continuous between discrete points. The interpolation process obtains a continuous energy allocation curve by smoothly connecting the values ​​between discrete points, which reflects the actual energy allocation of the system at different time points and is suitable for continuous working environments. Based on the continuous energy allocation curve, the state variables at each moment are calculated. State variables usually include the state of charge (SOC) of the battery, the power output of the energy storage unit, etc. By calculating the state variables, the target operation trajectory of the system is obtained, which describes the working state and energy change of the energy storage unit during the entire operation of the system. Constraint checks are performed on the target operation trajectory. Check whether the state of charge of each energy storage unit is within the allowable range and whether the power output exceeds its design limit. These constraints usually include that the SOC of the battery must be kept within a certain range (for example, 20% to 80%) to avoid overcharging or over-discharging, and the power output of supercapacitors and other energy storage units must also be kept within their maximum allowable power range. By checking the constraints, the operation safety of the system is ensured and excessive loss of the energy storage unit is avoided. After completing the constraint check, energy allocation sequences at different time scales are extracted from the target operation trajectory. These sequences represent the optimal energy allocation strategy of the system at different time points. Through these sequences, the energy storage unit can respond to load demand in an optimal way and ensure the optimization of power balance and system efficiency.

[0092] In a specific embodiment, the process of executing step 500 may specifically include the following steps:

[0093] The state of charge of the ternary polymer lithium battery, the state of charge of the supercapacitor, the state of charge of the shared charging device, the load power, the first-order rate of change of the load power and the second-order rate of change are taken as six input variables;

[0094] The six input variables are fuzzified, the membership function of each input variable is created, the fuzzy set of the input variables is obtained, and the initial fuzzy rule base is created based on the multi-level adaptive power diversion strategy to construct the mapping relationship from the fuzzy set of the input variables to the fuzzy set of the output variables;

[0095] Based on the initial fuzzy rule base, the maximum-minimum synthesis method is used for fuzzy reasoning to obtain the fuzzy set of output variables;

[0096] The fuzzy set of output variables is defuzzified, and the exact values ​​of three output variables are calculated using the centroid method to obtain the initial fuzzy logic controller.

[0097] A three-layer feedforward neural network is constructed, and the input data and output data of the initial fuzzy logic controller are used as training samples. The three-layer feedforward neural network is trained based on the back propagation algorithm to obtain a neural network model.

[0098] Based on the neural network model, the initial fuzzy rule base is optimized, the weight and structure of the rules are adjusted to obtain the target fuzzy rule base, and the membership function is fine-tuned based on the neural network model to obtain the target membership function;

[0099] Based on the target fuzzy rule base and the target membership function, a neural network optimized six-input three-output fuzzy logic controller is generated;

[0100] The real-time state of charge data of ternary polymer lithium batteries, supercapacitors and shared charging equipment are collected to obtain the real-time value of load power, and the first-order change rate and second-order change rate of the real-time value of load power are calculated. The data are input into the six-input and three-output fuzzy logic controller optimized by neural network to obtain the power regulation coefficient of ternary polymer lithium batteries, supercapacitors and shared charging equipment.

[0101] Specifically, six key input variables are defined, including the state of charge (SOC) of the ternary polymer lithium battery, the state of charge of the supercapacitor, the state of charge of the shared charging device, the load power, the first-order rate of change of the load power, and the second-order rate of change of the load power. These six input variables are fuzzified. The continuous input variables are converted into membership values ​​in the fuzzy set to facilitate fuzzy reasoning. In this process, a membership function is created for each input variable. The membership function defines how the input variable is mapped to a certain range in the fuzzy set. For example, the SOC of the battery can be divided into membership sets such as "low" and "medium", and each set corresponds to a certain range of SOC values. After creating the membership function, the fuzzy set of the input variables is obtained. Based on the multi-level adaptive power splitting strategy, an initial fuzzy rule base is constructed. These rules are used to describe how the system should allocate power under different input conditions. The basic form of each fuzzy rule is "if...then...", for example, "if the SOC of the battery is 'low', the SOC of the supercapacitor is 'high', and the load power change rate is 'large', then increase the power output of the supercapacitor." These rules define the mapping relationship between the input variable fuzzy set and the output variable fuzzy set. Based on the initial fuzzy rule base, the maximum-minimum synthesis method is used for fuzzy reasoning. The maximum-minimum synthesis method determines the activation degree of each fuzzy rule by synthesizing the input membership value, and thus obtains the fuzzy set of the output variable. At this time, the output variable is the membership value in the fuzzy set. In order to convert the fuzzy output into a specific numerical value, defuzzification is performed. The commonly used method for defuzzification is the centroid method, also known as the centroid method. The centroid method converts the fuzzy value into a specific numerical value by calculating the centroid of the output fuzzy set. For example, the membership value of the output variable is obtained by weighted averaging to obtain an accurate value, which is used to guide the power distribution of the battery and supercapacitor. The initial fuzzy logic controller is obtained, and power regulation is performed according to the real-time input data. In order to optimize the performance of the fuzzy controller, a three-layer feedforward neural network is constructed. The input of the neural network is the input data of the initial fuzzy logic controller, and the output is the output data of the fuzzy controller. These data are used as training samples of the neural network, and the purpose is to optimize the control rules and parameters of the fuzzy controller through the adaptive learning ability of the neural network. Through the back propagation algorithm, the neural network can adjust its weights and biases to minimize the output error of the model. The back propagation algorithm is an iterative optimization algorithm that updates the weights in the network by calculating the gradient of the output error and gradually improves the performance of the model. After the training is completed, an optimized neural network model is obtained. Based on this model, the initial fuzzy rule base is optimized. The neural network can automatically adjust the weight of each fuzzy rule and optimize the structure of the rules based on the actual data. For example, the weight of a rule may be reduced because it is not often activated, or the system finds that more complex rules need to be introduced to handle specific input situations. Fine-tune the membership function based on the neural network model.The neural network optimizes the shape of the membership function by learning a large amount of historical data to make it more consistent with the behavior of the actual system. For example, the membership range of some variables may need to be adjusted to better reflect their impact on power distribution. Based on the optimized fuzzy rule base and membership function, a neural network optimized six-input three-output fuzzy logic controller is generated. The controller can not only process real-time input data, but also has adaptive capabilities and can continuously learn and optimize based on historical data. The input variables include the real-time state of charge data of the ternary polymer lithium battery, supercapacitor and shared charging device, as well as the real-time value of the load power and its first-order and second-order change rate. The real-time data is input into the neural network optimized fuzzy logic controller. The controller outputs the power regulation coefficients of the ternary polymer lithium battery, supercapacitor and shared charging device through fuzzy reasoning and defuzzification processes. These coefficients reflect how each energy storage unit should distribute power under the current load and energy storage state. For example, if the rate of change of the load power is large, the supercapacitor will be instructed to assume a larger power output, while the ternary polymer lithium battery is mainly used for stable, low-frequency energy supply.

[0102] In a specific embodiment, the execution step collects real-time state of charge data of the ternary polymer lithium battery, the supercapacitor and the shared charging device, obtains the real-time value of the load power, and calculates the first-order change rate and the second-order change rate of the real-time value of the load power, and inputs the six-input and three-output fuzzy logic controller optimized by the neural network to obtain the power regulation coefficient of the ternary polymer lithium battery, the supercapacitor and the shared charging device. The process can specifically include the following steps:

[0103] Collect real-time state of charge data of ternary polymer lithium batteries, supercapacitors and shared charging equipment, obtain real-time value of load power, calculate first-order and second-order change rates of load power, and integrate them to obtain standardized input data;

[0104] The standardized input data is input into the six-input and three-output fuzzy logic controller optimized by the neural network, each input variable is fuzzified, and the membership value is calculated based on the target membership function to obtain the fuzzy input set;

[0105] According to the fuzzy input set, rule matching and activation are performed based on the target fuzzy rule base to obtain an activated rule set, and based on the activated rule set, the output strength of the rule is calculated to obtain a fuzzy output set;

[0106] The fuzzy output set is defuzzified based on the centroid method, and the power regulation coefficients of ternary polymer lithium batteries, supercapacitors and shared charging equipment are calculated.

[0107] Specifically, the real-time state of charge data of ternary polymer lithium batteries, supercapacitors and shared charging devices are collected to obtain the load power value of the current system to reflect the current energy demand of the system. The first-order change rate and second-order change rate of the load power are calculated. The first-order change rate indicates the rate at which the load power changes over time, and the second-order change rate reflects the acceleration of the load power change. These two parameters can capture the dynamic characteristics of load fluctuations and help the system better predict future power demand. For example, if the load power changes rapidly and the acceleration is large, the system needs to respond quickly to avoid the impact of excessive load fluctuations on the energy storage unit. The state of charge data, load power and its first-order and second-order change rates are integrated into a set of standardized input data to eliminate the scale differences between different input variables. The standardized input data is input into a six-input and three-output fuzzy logic controller optimized by a neural network. Inside the controller, each input variable is fuzzified. The target membership function is used to convert the standardized input variables into membership values ​​in the fuzzy set. The membership function is used to map each input variable to several fuzzy sets (for example, "low", "medium", and "high") to form a fuzzy input set. Assuming that the SOC of the ternary polymer lithium battery is 0.6, through fuzzification, its membership function may classify it as a "medium" set, and the membership value is 0.7, which means that the state of charge of the battery is at a high level in the "medium" level. Similarly, variables such as the rate of change of load power will also be fuzzified through their respective membership functions to obtain a series of fuzzy input values. According to the fuzzy input set, rule matching and activation are performed based on the target fuzzy rule base. The fuzzy rule base contains initial rules on how to allocate power, which usually exist in the form of "if-then". The rule matching process finds the rules that meet the conditions by comparing the current fuzzy input with the rules in the fuzzy rule base. Each rule in the activated rule set is activated according to the membership value of the input variable, and the activation degree is determined by the membership of the fuzzy input. For example, if the membership value of the input variable is high, the corresponding rule activation degree is also high. Based on the activated rule set, the output strength of each rule is calculated. The output strength reflects the impact of each rule on the final power allocation. In this process, the maximum-minimum synthesis method is used to determine the membership value of the rule output. The maximum-minimum synthesis method is a common fuzzy reasoning method that calculates the output fuzzy set by taking the minimum membership value in the rule, and takes the maximum membership value to represent the contribution of the rule to the output. The fuzzy output set is defuzzified and the fuzzy output is converted into a specific value. The centroid method is used to calculate the centroid of the fuzzy output set to obtain the accurate output value. The formula of the centroid method is as follows:

[0108] ;

[0109] in, is the output value after defuzzification, is the membership value of the fuzzy output set, is the specific value of the fuzzy output. Through this process, the power regulation coefficients of the ternary polymer lithium battery, supercapacitor and shared charging device are obtained. The power regulation coefficient determines how much power each energy storage unit should bear at the current moment.

[0110] In a specific embodiment, the execution step performs defuzzification processing on the fuzzy output set based on the centroid method, and the process of calculating the power regulation coefficient of the ternary polymer lithium battery, the supercapacitor and the shared charging device can specifically include the following steps:

[0111] Discretize the fuzzy output set, convert the fuzzy output set into a discrete membership value sequence, and obtain a discretized fuzzy output set;

[0112] Based on the discretized fuzzy output set, the product of the membership value of each discrete point and its corresponding output value is calculated to obtain a weighted membership value sequence;

[0113] The weighted membership value sequence is accumulated to obtain the weighted membership value sum, and the membership value sequence in the discretized fuzzy output set is accumulated to obtain the membership value sum;

[0114] Based on the sum of weighted membership values ​​divided by the sum of membership values, the centroid value is calculated to obtain the initial defuzzification result;

[0115] Perform data smoothing processing based on the initial defuzzification result to obtain the target defuzzification result;

[0116] The target defuzzification results are normalized to ensure that the sum of the power regulation coefficients of the ternary polymer lithium battery, supercapacitor and shared charging equipment is 1, and the power regulation coefficients of the ternary polymer lithium battery, supercapacitor and shared charging equipment are obtained.

[0117] Specifically, the fuzzy output set is discretized to convert the continuous membership value into discrete point values. The continuous membership value sequence in the fuzzy output set is divided into several discrete points to obtain a discretized fuzzy output set. For example, assuming that the membership value range contained in the fuzzy output set is between 0 and 1, during the discretization process, this range is divided into several discrete points at equal intervals, such as 0.1, 0.2, 0.3, etc. The continuous membership values ​​are assigned to these discrete points, and each point represents a specific output value and its corresponding membership. Based on the discretized fuzzy output set, the product of the membership value of each discrete point and its corresponding output value is calculated. The product is used to weight the membership value of each discrete point, thereby reflecting the contribution of the point to the overall output. The specific formula is:

[0118] ;

[0119] in, represents the output value of the i-th discrete point, Represents the membership value of the point, the product Indicates the weighted contribution of the point to the final output result. By calculating the weighted membership value of each discrete point, a weighted membership value sequence is obtained. The weighted membership value sequence is accumulated to obtain the sum of the weighted membership values:

[0120] Sum of weighted membership values ​​= ;

[0121] At the same time, the membership value sequence in the discretized fuzzy output set is accumulated to obtain the sum of the membership values:

[0122] The total membership value = ;

[0123] The sum of the membership values ​​represents the total contribution of all discrete points in the fuzzy output set. The centroid value is calculated based on the sum of the weighted membership values ​​divided by the sum of the membership values. The formula for the centroid value is:

[0124] Center of gravity value = ;

[0125] The centroid value calculated by this formula is the initial defuzzification result, which reflects the weighted average of the discrete fuzzy output set at each membership value. Perform data smoothing on the initial defuzzification result. Use the moving average method or the exponential weighted average method to reduce sharp changes in the data, obtain a more continuous and stable output value, and obtain the target defuzzification result. Normalize the target defuzzification result to ensure that the sum of the power regulation coefficients of the ternary polymer lithium battery, supercapacitor, and shared charging device is 1. The formula for normalization is:

[0126] ;

[0127] in, represents the normalized power regulation coefficient, is the original power regulation coefficient of each energy storage unit, and the normalized sum ensures that the total power allocation of each energy storage unit is 1. Through this process, the power regulation coefficients of ternary polymer lithium batteries, supercapacitors and shared charging devices are obtained. These regulation coefficients are used to control the response of different energy storage units under load power changes. For example, assuming that the current load power is 100W, after defuzzification and normalization, the power regulation coefficient of the ternary polymer lithium battery is 0.5, the power regulation coefficient of the supercapacitor is 0.3, and the power regulation coefficient of the shared charging device is 0.2. In this case, the load power of 100W is distributed as follows: 50W is provided by the ternary polymer lithium battery, 30W is provided by the supercapacitor, and the remaining 20W is provided by the shared charging device.

[0128] The above describes the energy storage management method of the ternary polymer lithium battery in the shared charging device in the embodiment of the present application. The following describes the energy storage management system 10 of the ternary polymer lithium battery in the shared charging device in the embodiment of the present application. Figure 2 In the embodiment of the present application, an energy storage management system 10 of a ternary polymer lithium battery in a shared charging device includes:

[0129] The test module 11 is used to perform multi-condition cycle charge and discharge tests on the ternary polymer lithium battery and test data fitting to obtain a multivariate capacity decay function;

[0130] A construction module 12 is used to construct a hybrid energy storage topology structure having a ternary polymer lithium battery, a supercapacitor, a shared charging device, a bidirectional DC / DC converter and a DC bus based on a multivariable capacity decay function to obtain an adaptive composite energy storage system;

[0131] The analysis module 13 is used to perform mathematical model analysis on each energy storage unit in the adaptive composite energy storage system to obtain a dynamic model of the energy storage system;

[0132] The allocation module 14 is used to generate an initial energy allocation strategy based on the dynamic model of the energy storage system, and perform multi-scale time-frequency decomposition and power allocation on the load power of the adaptive composite energy storage system to obtain a multi-level adaptive power diversion strategy;

[0133] The calculation module 15 is used to construct a six-input three-output fuzzy logic controller optimized by a neural network based on a multi-level adaptive power splitting strategy, and calculate the power regulation coefficients of the ternary polymer lithium battery, supercapacitor and shared charging equipment.

[0134] Through the cooperation of the above-mentioned components, the multivariable capacity decay function obtained by multi-condition cycle charge and discharge test and data fitting of the ternary polymer lithium battery can accurately describe the relationship between the battery capacity and the number of cycles, temperature and charge and discharge rate. The adaptive composite energy storage system constructed based on the multivariable capacity decay function combines the advantages of ternary polymer lithium batteries, supercapacitors and shared charging equipment, and can effectively respond to load requirements in different frequency ranges, thereby improving the overall performance and reliability of the system. The mathematical model analysis of each energy storage unit in the adaptive composite energy storage system can accurately reflect the real-time state of the system. Through multi-scale time-frequency decomposition and power allocation, the multi-level adaptive power diversion strategy can better match the characteristics of different energy storage units, improve energy utilization efficiency, and extend the life of the system. The six-input and three-output fuzzy logic controller optimized by neural network is used to realize the intelligent calculation of the power regulation coefficient of ternary polymer lithium batteries, supercapacitors and shared charging equipment, and improve the dynamic response capability and control accuracy of the system. The present invention realizes the efficient management of the energy storage system of shared charging equipment, improves the use efficiency and life of ternary polymer lithium batteries, and meets the complex load requirements of shared charging equipment.

[0135] The present application also provides an electronic device, which includes a memory and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the energy storage management method of the ternary polymer lithium battery in a shared charging device in the above-mentioned embodiments.

[0136] The present application also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein instructions are stored in the computer, and when the instructions are executed on a computer, the computer executes the steps of the energy storage management method of the ternary polymer lithium battery in a shared charging device.

[0137] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0138] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable an electronic device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0139] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for managing energy storage of a ternary polymer lithium battery in a shared charging device, characterized in that: The method comprises: The invention relates to a method for performing multi-condition cycle charge and discharge test and test data fitting on a ternary polymer lithium battery to obtain a multivariate capacity decay function; specifically comprising: performing cycle charge and discharge test and test data collection on the ternary polymer lithium battery at different temperature ranges and charge and discharge rates to obtain a test data set, wherein the test data set includes battery capacity, cycle number, temperature and charge and discharge rate data; normalizing the test data set to obtain a normalized data set, and expressing the battery capacity as a function of cycle number, temperature and charge and discharge rate based on the normalized data set to construct a polynomial regression model; using the least squares method to perform parameter estimation on the polynomial regression model to obtain an initial capacity decay function, and validating the initial capacity decay function based on the k-fold cross validation method, calculating the root mean square error and the determination coefficient, and obtaining a validation result; fine-tuning the initial capacity decay function based on the validation result to obtain a multivariate capacity decay function, wherein the multivariate capacity decay function is used to describe the relationship between the battery capacity and the cycle number, temperature and charge and discharge rate, and the multivariate capacity decay function is: C(n, T, R)=C0*(a1*e （-b1*n） +a2*e （-b2*n） )*(1+k1*(T-Tref))*(1+k2*(R-Rref)) 2 +k3*n*T*R; where C (n, T, R) represents the capacity of the battery under target conditions, C0 represents the initial capacity of the battery, n represents the number of cycles, T represents the temperature, R represents the charge and discharge rate, e represents the base of the natural logarithm, a1, a2, b1, b2 are parameters that describe the capacity attenuation with the number of cycles, k1 is the temperature influence coefficient, k2 is the rate influence coefficient, k3 is the interaction influence coefficient of the number of cycles, temperature and rate, Tref is the reference temperature, and Rref is the reference rate; Based on the multivariable capacity decay function, a hybrid energy storage topology structure having a ternary polymer lithium battery, a supercapacitor, a shared charging device, a bidirectional DC / DC converter and a DC bus is constructed to obtain an adaptive composite energy storage system; Performing mathematical model analysis on each energy storage unit in the adaptive composite energy storage system to obtain a dynamic model of the energy storage system; Generate an initial energy allocation strategy based on the dynamic model of the energy storage system, and perform multi-scale time-frequency decomposition and power allocation on the load power of the adaptive composite energy storage system to obtain a multi-level adaptive power diversion strategy; Based on the multi-level adaptive power splitting strategy, a neural network optimized six-input three-output fuzzy logic controller is constructed, and the power regulation coefficients of ternary polymer lithium batteries, supercapacitors and shared charging equipment are calculated.

2. The energy storage management method of the ternary polymer lithium battery in a shared charging device according to claim 1, characterized in that: The hybrid energy storage topology structure having a ternary polymer lithium battery, a supercapacitor, a shared charging device, a bidirectional DC / DC converter and a DC bus is constructed based on the multivariable capacity decay function to obtain an adaptive composite energy storage system, including: Based on the multivariate capacity decay function, the remaining capacity of the ternary polymer lithium battery under different working conditions is calculated and a battery capacity state matrix is ​​generated, and the battery capacity state matrix is ​​subjected to singular value decomposition to construct a battery capacity feature space; Performing load characteristic analysis on the shared charging device based on the battery capacity feature space to obtain a load power spectrum, and performing frequency segmentation based on the load power spectrum to determine the operating frequency range of the ternary polymer lithium battery, the supercapacitor and the shared charging device; Creating a topology and a control strategy of a bidirectional DC / DC converter according to the operating frequency range, obtaining a converter parameter set, and creating a bus voltage stabilization controller based on the converter parameter set; Optimizing the capacity ratio of the ternary polymer lithium battery, the supercapacitor and the shared charging device to obtain a target capacity configuration scheme; Based on the target capacity configuration scheme and the bus voltage stabilization controller, a hybrid energy storage topology structure including a ternary polymer lithium battery, a supercapacitor, a shared charging device, a bidirectional DC / DC converter and a DC bus is constructed to obtain an adaptive composite energy storage system. The adaptive composite energy storage system connects the ternary polymer lithium battery and the shared charging device to the DC bus through a bidirectional DC / DC converter, and uses the supercapacitor as a peak power buffer unit.

3. The energy storage management method of the ternary polymer lithium battery in a shared charging device according to claim 2, characterized in that: The method of performing mathematical model analysis on each energy storage unit in the adaptive composite energy storage system to obtain a dynamic model of the energy storage system includes: A third-order Thevenin equivalent circuit model is constructed for the ternary polymer lithium battery in the adaptive composite energy storage system, and a temperature coefficient is introduced to correct the open circuit voltage and internal resistance parameters to obtain a battery voltage equation; Analyze the charging characteristics of the shared charging device, construct a second-order RC model including an equivalent resistor and an equivalent capacitor, and obtain a voltage-current relationship equation; According to the bidirectional DC / DC converter, a converter transient model including a switching function is established, and a converter nonlinear average value model is obtained by a state space averaging method; Combining the battery voltage equation, the voltage-current relationship equation and the converter nonlinear average value model to construct a state space equation of the adaptive composite energy storage system; Creating an extended Kalman filter based on the state space equation, combining the state variables and unknown parameters of the adaptive composite energy storage system into a vector to be estimated, and iteratively calculating the vector to be estimated to obtain estimated values ​​of the system state and parameters; The estimated error covariance matrix is ​​calculated based on the estimated values ​​of the system state and parameters, and the process noise covariance and the measurement noise covariance are dynamically adjusted using an adaptive weighting strategy to obtain a target Kalman gain. The system state and parameters are corrected based on the target Kalman gain to obtain a dynamic model of the energy storage system.

4. The energy storage management method of the ternary polymer lithium battery in a shared charging device according to claim 1, characterized in that: The initial energy allocation strategy is generated based on the dynamic model of the energy storage system, and the load power of the adaptive composite energy storage system is subjected to multi-scale time-frequency decomposition and power allocation to obtain a multi-level adaptive power diversion strategy, including: Constructing a corresponding energy management objective function based on the energy storage system dynamic model, wherein the energy management objective function includes system efficiency, energy storage unit life and power balance constraints; Based on the dynamic programming algorithm, the optimal energy allocation sequence under different time scales is calculated for the energy management objective function to obtain an initial energy allocation strategy; Performing wavelet packet transform on the load power of the adaptive composite energy storage system, selecting the best basis function and decomposition layer number, obtaining multi-scale wavelet coefficients, and performing threshold processing based on the multi-scale wavelet coefficients to reconstruct the wavelet coefficients into an initial low-frequency power component, an initial intermediate-frequency power component, and an initial high-frequency power component; Performing empirical mode decomposition on the initial low-frequency power component to obtain a plurality of eigenmode functions and a residual function, and recombining them into a target low-frequency power component, a target intermediate-frequency power component, and a target high-frequency power component based on the frequency characteristics of the eigenmode function and the residual function; According to the initial energy allocation strategy, the target low-frequency power component is allocated to the ternary polymer lithium battery, the target medium-frequency power component is allocated to the supercapacitor, and the target high-frequency power component is allocated to the shared charging device, thereby obtaining a multi-level adaptive power splitting strategy.

5. The energy storage management method of the ternary polymer lithium battery in a shared charging device according to claim 4, characterized in that: The method of calculating the optimal energy allocation sequence under different time scales for the energy management objective function based on the dynamic programming algorithm to obtain the initial energy allocation strategy includes: Discretizing the energy management objective function, dividing the continuous state space and decision space into a finite number of discrete points, and obtaining a discrete state transfer equation; Constructing a state transfer matrix based on the discrete state transfer equation, calculating the cost function value of each state transfer, and obtaining a cost matrix; Performing backward recursion on the cost matrix, and gradually calculating the optimal value function at each moment starting from the termination moment, to obtain an optimal value function sequence; Based on the optimal value function sequence, a forward search is performed to determine the optimal decision at each moment, and a search energy allocation strategy is obtained; and the search energy allocation strategy is interpolated to make the energy allocation strategy between discrete points continuous to obtain a continuous energy allocation curve; Based on the continuous energy allocation curve, the state variables at each moment are calculated to obtain the target operation trajectory, and the target operation trajectory is subjected to constraint checking to ensure that the state of charge limit and power limit of each energy storage unit are met, and the energy allocation sequence at different time scales is extracted to obtain the initial energy allocation strategy.

6. The energy storage management method of the ternary polymer lithium battery in a shared charging device according to claim 1, characterized in that: The method of constructing a six-input three-output fuzzy logic controller optimized by a neural network based on the multi-level adaptive power splitting strategy and calculating the power regulation coefficient of the ternary polymer lithium battery, the supercapacitor and the shared charging device includes: The state of charge of the ternary polymer lithium battery, the state of charge of the supercapacitor, the state of charge of the shared charging device, the load power, the first-order rate of change of the load power and the second-order rate of change are taken as six input variables; Performing fuzzification processing on six input variables, creating a membership function for each input variable, obtaining a fuzzy set of input variables, creating an initial fuzzy rule base based on the multi-level adaptive power splitting strategy, and constructing a mapping relationship from the fuzzy set of input variables to the fuzzy set of output variables; Based on the initial fuzzy rule base, a maximum-minimum synthesis method is used to perform fuzzy reasoning to obtain a fuzzy set of output variables; Defuzzification is performed on the output variable fuzzy set, and the exact values ​​of the three output variables are calculated using the centroid method to obtain an initial fuzzy logic controller; Constructing a three-layer feedforward neural network, using the input data and output data of the initial fuzzy logic controller as training samples, and training the three-layer feedforward neural network based on a back propagation algorithm to obtain a neural network model; Optimizing the initial fuzzy rule base based on the neural network model, adjusting the weights and structures of the rules to obtain a target fuzzy rule base, and fine-tuning the membership function based on the neural network model to obtain a target membership function; Based on the target fuzzy rule base and the target membership function, a neural network optimized six-input three-output fuzzy logic controller is generated; The real-time state of charge data of the ternary polymer lithium battery, supercapacitor and shared charging device are collected to obtain the real-time value of the load power, and the first-order change rate and the second-order change rate of the real-time value of the load power are calculated and input into the six-input and three-output fuzzy logic controller optimized by the neural network to obtain the power regulation coefficient of the ternary polymer lithium battery, supercapacitor and shared charging device.

7. The energy storage management method of the ternary polymer lithium battery in a shared charging device according to claim 6, characterized in that: The real-time state of charge data of the ternary polymer lithium battery, supercapacitor and shared charging device are collected to obtain the real-time value of the load power, and the first-order change rate and the second-order change rate of the real-time value of the load power are calculated and input into the six-input and three-output fuzzy logic controller optimized by the neural network to obtain the power regulation coefficient of the ternary polymer lithium battery, supercapacitor and shared charging device, including: Collect real-time state of charge data of ternary polymer lithium batteries, supercapacitors and shared charging equipment, obtain real-time value of load power, and calculate the first-order change rate and second-order change rate of the load power, and integrate them to obtain standardized input data; Inputting the standardized input data into the six-input three-output fuzzy logic controller optimized by the neural network, performing fuzzification processing on each input variable, calculating the membership value based on the target membership function, and obtaining a fuzzy input set; According to the fuzzy input set, rule matching and activation are performed based on the target fuzzy rule base to obtain an activated rule set, and based on the activated rule set, the output strength of the rule is calculated to obtain a fuzzy output set; The fuzzy output set is defuzzified based on the centroid method to calculate the power regulation coefficients of the ternary polymer lithium battery, the supercapacitor and the shared charging device.

8. The energy storage management method of the ternary polymer lithium battery in a shared charging device according to claim 7, characterized in that: The defuzzification process of the fuzzy output set based on the centroid method to calculate the power regulation coefficient of the ternary polymer lithium battery, the supercapacitor and the shared charging device includes: Discretizing the fuzzy output set, converting the fuzzy output set into a discrete membership value sequence, and obtaining a discretized fuzzy output set; Calculate the product of the membership value of each discrete point and its corresponding output value based on the discretized fuzzy output set to obtain a weighted membership value sequence; Performing a cumulative operation on the weighted membership value sequence to obtain a weighted membership value sum, and performing a cumulative operation on the membership value sequence in the discretized fuzzy output set to obtain a membership value sum; Based on the weighted membership value sum divided by the membership value sum, a centroid value is calculated to obtain an initial defuzzification result; Performing data smoothing processing based on the initial defuzzification result to obtain a target defuzzification result; The target defuzzification result is normalized to ensure that the sum of the power regulation coefficients of the ternary polymer lithium battery, the supercapacitor and the shared charging device is 1, and the power regulation coefficients of the ternary polymer lithium battery, the supercapacitor and the shared charging device are obtained.

9. An energy storage management system for ternary polymer lithium batteries in shared charging equipment, characterized in that: A method for managing energy storage of a ternary polymer lithium battery in a shared charging device according to any one of claims 1 to 8, the system comprising: The test module is used to perform multi-condition cycle charge and discharge tests and test data fitting on a ternary polymer lithium battery to obtain a multivariate capacity decay function; specifically comprising: performing cycle charge and discharge tests and test data collection on a ternary polymer lithium battery at different temperature ranges and charge and discharge rates to obtain a test data set, wherein the test data set includes battery capacity, cycle number, temperature and charge and discharge rate data; normalizing the test data set to obtain a normalized data set, and based on the normalized data set, expressing the battery capacity as a function of the cycle number, temperature and charge and discharge rate to construct a polynomial regression model; using the least squares method to perform parameter estimation on the polynomial regression model to obtain an initial capacity decay function, and validating the initial capacity decay function based on the k-fold cross validation method, calculating the root mean square error and the determination coefficient to obtain a validation result; fine-tuning the initial capacity decay function based on the validation result to obtain a multivariate capacity decay function, wherein the multivariate capacity decay function is used to describe the relationship between the battery capacity and the cycle number, temperature and charge and discharge rate, and the multivariate capacity decay function is: C(n, T, R)=C0*(a1*e （-b1*n） +a2*e （-b2*n） )*(1+k1*(T-Tref))*(1+k2*(R-Rref)) 2 +k3*n*T*R; where C (n, T, R) represents the capacity of the battery under target conditions, C0 represents the initial capacity of the battery, n represents the number of cycles, T represents the temperature, R represents the charge and discharge rate, e represents the base of the natural logarithm, a1, a2, b1, b2 are parameters that describe the capacity attenuation with the number of cycles, k1 is the temperature influence coefficient, k2 is the rate influence coefficient, k3 is the interaction influence coefficient of the number of cycles, temperature and rate, Tref is the reference temperature, and Rref is the reference rate; A construction module is used to construct a hybrid energy storage topology structure having a ternary polymer lithium battery, a supercapacitor, a shared charging device, a bidirectional DC / DC converter and a DC bus based on the multivariable capacity decay function to obtain an adaptive composite energy storage system; An analysis module, used for performing mathematical model analysis on each energy storage unit in the adaptive composite energy storage system to obtain a dynamic model of the energy storage system; A distribution module, used to generate an initial energy distribution strategy based on the dynamic model of the energy storage system, and perform multi-scale time-frequency decomposition and power distribution on the load power of the adaptive composite energy storage system to obtain a multi-level adaptive power diversion strategy; The calculation module is used to construct a six-input and three-output fuzzy logic controller optimized by a neural network based on the multi-level adaptive power splitting strategy, and calculate the power regulation coefficients of the ternary polymer lithium battery, supercapacitor and shared charging equipment.

Images