Energy storage energy management system for coordinating power distribution and battery optimization

Abstract

The invention relates to an energy storage energy management system for coordinating power distribution and battery optimization, and belongs to the technical field of energy storage energy management. The system comprises a coordinated scheduling layer, a power distribution layer and a unit execution layer. The coordinated scheduling layer receives an original power signal, generates smooth power through smoothing and frequency domain decomposition, and decouples the smooth power; the power distribution layer outputs a final power instruction comprehensively considering economy, loss and SOC balance through sequence optimization; and the unit execution layer converts an instruction into an energy storage unit switching state and a power instruction through mixed integer programming and high-frequency management. According to the method, power frequency domain separation is realized through adaptive VMD decomposition, balance is improved based on dynamic SOC initial distribution, economic operation cost is optimized through resistance coefficient weighting, loss and balance are balanced through multi-objective optimization, and finally execution precision is guaranteed through MILP switching control. The problems of rough power distribution, accelerated battery aging, difficulty in multi-target cooperation and the like in traditional energy storage management are effectively solved.

Description

Technical Field

[0001] This invention belongs to the field of energy storage management technology, specifically relating to an energy storage management system that coordinates power distribution and battery optimization. Background Technology

[0002] With the increasing penetration rate of renewable energy, the demand for energy storage systems in the power grid is growing. Energy storage management is a core technology to ensure the safe, stable, and economical operation of the system, and its core task lies in coordinating power distribution and battery optimization.

[0003] Traditional energy management methods typically employ fixed filtering frequencies for power decomposition, failing to adequately consider the differences in dynamic response characteristics between batteries and supercapacitors, thus hindering the full realization of the advantages of energy storage components. Within battery systems, existing power allocation strategies often neglect the differences in initial states (such as SOC and SOH) between cells, simply allocating power according to capacity ratios. This can easily lead to overcharging and over-discharging of some cells, accelerating the deterioration of battery pack inconsistencies. Furthermore, multi-objective optimization problems such as system operating costs, energy loss, and battery balancing are often addressed in isolation, lacking a collaborative optimization mechanism, making it difficult to achieve a balance between economic efficiency and lifespan extension. Summary of the Invention

[0004] To address the aforementioned problems in existing technologies, this invention provides an energy storage management system that coordinates power allocation and battery optimization. It achieves power frequency domain separation through adaptive VMD decomposition, improves balance based on dynamic SOC initial allocation, optimizes economic operating costs through resistance coefficient weighting, balances losses and balances through multi-objective optimization, and finally ensures execution accuracy through MILP switching control. This effectively solves problems such as coarse power allocation, accelerated battery aging, and difficulty in multi-objective coordination in traditional energy storage management.

[0005] The objective of this invention can be achieved through the following technical solutions: This disclosure provides an energy storage management system that coordinates power distribution and battery optimization, including a coordination scheduling layer, a power distribution layer, and a unit execution layer; The coordination and scheduling layer is used as the interface between the system and the external environment. It is responsible for receiving the original power signal, generating smooth power that meets the grid connection requirements through smoothing and frequency domain decomposition algorithms, and using frequency domain decomposition technology to decouple the total power of hybrid energy storage into high-frequency power components and low-frequency power components. Then, it generates a preliminary power allocation scheme for battery cells based on the low-frequency power components. The power allocation layer is used to receive the low-frequency power components and preliminary power allocation schemes of the battery cells issued by the coordination and scheduling layer. With the economic operation of the system and multi-objective coordination as the core, it outputs the final power allocation instruction of the battery system that comprehensively considers economic cost, system loss and SOC balance through sequence optimization solution. The unit execution layer is used to receive the final power allocation command and high-frequency power component of the battery system sent down from the upper layer. Through mixed integer programming and high-frequency multi-layer management strategy, it transforms the final power allocation command and high-frequency power component of the battery system into specific energy storage unit switching states and precise power commands. The coordination and scheduling layer includes a battery power dynamic initial allocation module. Based on the low-frequency power components obtained by decomposition and combined with the real-time status data of each battery cell, the battery power dynamic initial allocation module performs dynamic power pre-allocation based on capacity and with SOC balance as the goal, and generates a preliminary power allocation scheme for the battery cells. The preliminary power allocation scheme for generating battery cells includes the following steps: Calculate the basic power allocation value: Calculate the capacity factor based on the rated capacity of each battery cell, and then allocate the low-frequency power component according to the capacity factor to obtain the basic power allocation value; Calculate the dynamic adjustment factor: Calculate the system average SOC, and based on the deviation of each unit's SOC from the average value, calculate the dynamic adjustment factor through a nonlinear function; Generate initial allocation scheme: Combine the basic allocation power value with the dynamic adjustment factor to calculate the initial battery allocation power, perform power limit check, and output the initial allocation scheme.

[0006] As a preferred embodiment of the present invention, the coordination and scheduling layer further includes a power smoothing and decomposition module; The power smoothing and decomposition module is responsible for processing the original power signal, generating smooth power that meets grid connection requirements through an adaptive smoothing algorithm, and decomposing the total power of the hybrid energy storage that needs to be carried by energy storage according to frequency characteristics using signal decomposition technology. The process of decomposing the total power of the hybrid energy storage system that needs to be stored according to frequency characteristics includes the following steps: Power smoothing: An adaptive moving average algorithm is used to dynamically adjust the window length L and calculate the smoothed power until the grid connection standard is met or the maximum window limit is reached. The formula for calculating the smoothing power is as follows: ; In the formula, P smooth [t] represents the smoothing power at time t; L represents the sliding window length; The original power sample value at historical time ti; L max This is the preset maximum window length limit; Hybrid power calculation: Calculate the difference between the original power signal and the smoothed power to obtain the total hybrid energy storage power that the energy storage system needs to bear; Power allocation and frequency domain allocation: Optimize VMD parameters to decompose the total power of hybrid energy storage with the goal of minimizing composite entropy; then analyze the instantaneous frequency of each modal component through Hilbert transform, merge the modal components according to the spectral characteristics, and finally output the high-frequency power component and the low-frequency power component.

[0007] As a preferred embodiment of the present invention, the formula for calculating the capacity factor is as follows: ; In the formula, m i C is the capacity factor of the i-th battery cell; i The rated capacity (Ah) of the i-th battery cell; ∑ j C j The total capacity of all battery cells; The formula for calculating the basic power allocation value is as follows: P base,i =P low ×m i ; In the formula, P base,i Assign power value based on the base, P low Low-frequency power components; The formula for calculating the dynamic adjustment factor is as follows: ; In the formula, α i α is the dynamic adjustment factor for the i-th battery cell; base is the basic adjustment factor; k is the adjustment coefficient, controlling the degree of influence of SOC difference on power allocation; SOC i The current state of charge (SOC) of the i-th battery cell. avg The average SOC of all battery cells; The formula for calculating the initial power allocation of the battery is as follows: ; In the formula, Initially allocate power to the battery.

[0008] As a preferred embodiment of the present invention, the power allocation layer includes an upper-layer economic optimization module and a lower-layer multi-objective coordination optimization module, wherein: The upper-level economic optimization module is used to solve a more economical power allocation scheme by constructing an optimization problem with the objective function of minimizing the total weighted operating cost of the system, starting from the initial power allocation of the battery. The process of finding a more economical power allocation scheme includes the following steps: Calculate the resistance coefficient: comprehensively evaluate the degree of deviation of the SOC from the ideal value and the SOH state of each battery cell, calculate the original resistance coefficient, and perform normalization processing; Constructing the economic objective function: Establish an objective function with the goal of minimizing the total weighted cost, expressed as: ; In the formula, C loss,i For internal loss costs, C aging,i For aging cost, N is the total number of battery cells, R is the cost of aging. i This is the normalized resistance coefficient; Solving the optimization problem: Under the constraints of power balance, SOC safety and unit power limits, the objective function is solved using a sequential quadratic programming algorithm, and the economically optimized power is output.

[0009] As a preferred technical solution of the present invention, the original resistance coefficient The calculation formula is: ; In the formula, SOC ref β is the SOC reference value; β is the SOC deviation penalty coefficient; γ is the SOH penalty coefficient; The formulas for calculating the internal damage cost and the aging cost are as follows: ; ; In the formula, This is the unit internal loss cost coefficient; These are the coefficients of the quadratic and linear terms in the internal loss power model, respectively. For scheduling time steps; Investment i Let be the investment cost of the i-th battery cell; Let be the rated capacity of the i-th battery cell; Let be the cycle life of the i-th battery cell; In order to time step Internal energy throughput For the first The power of each battery cell.

[0010] As a preferred technical solution of the present invention, the lower-level multi-objective coordination and optimization module is used to coordinate the two conflicting objectives of total system loss and SOC balance on the basis of economically optimized power, find the Pareto optimal solution set through multi-objective optimization algorithm, and select the final compromise power allocation scheme according to the real-time operating conditions of the system. The selection of the final compromise power allocation scheme based on the real-time operating conditions of the system includes the following steps: Construct a multi-objective optimization problem: Define two objective functions: Minimize total charge / discharge loss. Minimize SOC variance And set constraints for power balance, SOC safety and cell power limits; Solving the Pareto front: A multi-objective genetic algorithm is used to generate a set of non-dominated solutions through population evolution, and the objective function value of each solution is normalized. Choosing a final compromise: Based on the current total power demand and the system's average health status, dynamically calculate the weights of the two objectives. and The weighted composite score of each solution in the Pareto front is calculated, and the solution with the highest score is selected as the final multi-objective optimized power of the battery.

[0011] As a preferred embodiment of the present invention, the expressions for the two objective functions are as follows: ; ; In the formula, The target is the total charge and discharge loss. The goal is to achieve SOC balance. Let be the power loss of the i-th battery cell; Let be the overall efficiency of the i-th battery cell; The predicted SOC value for the next time step. ; The predicted average SOC value for the next time step; The formula for calculating the weighted composite score of each solution is as follows: ; ; ; In the formula, w1 and w2 are the loss targets, respectively. and balance goals The weights; The basic weight for the loss target; This is the power demand weighting coefficient; This refers to the rated total power of the battery pack. Total power requirement; For health status weighting coefficients; SOH i The health state of the i-th battery cell; the score is the weighted composite score for each Pareto solution; , The objective function value is the normalized value; N is the total number of battery cells; The battery's multi-objective optimized power P final,i for: .

[0012] As a preferred embodiment of the present invention, the unit execution layer includes a low-frequency scale battery MILP control module and a high-frequency scale supercapacitor multi-layer management module, wherein: The low-frequency scale battery MILP control module, based on the battery's multi-objective power optimization, optimizes and determines the switching state of each battery cell by establishing a mixed integer linear programming model, and obtains the optimal combination of battery switching states and corresponding power. The optimal combination of battery switching states and the corresponding power includes the following steps: Problem modeling: Define decision variables and construct the objective function. ; In the formula, J is the objective function value; N is the total number of battery cells; This represents the total power command obtained from the lower-level multi-objective optimization; Represents the weighting coefficients of the SOC equilibrium term; SOC i Indicates the first The current state of charge of each battery cell; This represents the average SOC of all battery cells in the active state. This indicates that the SOC balance penalty is calculated only for the input units; Among them, the decision variables include the switching state. and power ; Set constraints; Switching logic and system operation constraints: At least one unit in any series branch must be switched on; Solution and Output: The solution is obtained using the MILP solver, and the optimal combination of battery switching states is output. And the corresponding power value after fine-tuning.

[0013] As a preferred embodiment of the present invention, the constraints include: Power balance guiding constraints: In the formula For the allowable power tracking error; SOC security constraints: In the formula, SOC i The current state of charge of the i-th battery cell, Δt bat E represents the time step of the battery control layer. rated,i This represents the rated capacity of the i-th battery cell; Power limit constraints: In the formula P max,iThis represents the absolute value of the maximum allowable charging and discharging power of the i-th battery cell.

[0014] As a preferred technical solution of the present invention, the high-frequency scale supercapacitor multi-layer management module is used to dynamically adjust the power of each supercapacitor unit for high-frequency power components through a four-layer management strategy including constraint, selection, optimization and compensation. The dynamic adjustment of the power of each supercapacitor unit includes the following steps: Constraint layer: Real-time checks whether the SOC and power of each supercapacitor unit exceed safety limits, and forcibly prunes any power commands that exceed the limits; Selection layer: Based on the preset SOC range rules, select the cells that need to be optimized and adjusted from all supercapacitor cells; Optimization layer: For the selected cells, the power command is dynamically adjusted by applying a charge / discharge power correction factor based on the SOC level, so as to guide the SOC back to the normal range. Compensation layer: Calculates the difference in total system power before and after optimization and adjustment, and the unselected units in the healthy SOC range share the power difference according to the capacity or SOC margin ratio.

[0015] The beneficial effects of this invention are as follows: This invention effectively separates high- and low-frequency power components through adaptive moving average and VMD decomposition, allowing the supercapacitor to handle high-frequency pulsating power while the battery handles low-frequency energy-type power, thereby reducing the charging and discharging stress and aging rate of the battery. Dynamic initial allocation based on real-time SOC introduces an equalization mechanism at the beginning of power allocation, avoiding deep charging and discharging of some battery cells and improving the consistency of the battery pack. Introducing a resistance coefficient in the upper-level economic optimization transforms cell state differences into economic weights, guiding power to be preferentially handled by cells in better states, reducing the total system operating cost. Lower-level multi-objective coordinated optimization solves the Pareto front and makes dynamic decisions, achieving an optimal trade-off between system losses and SOC equalization, improving the overall system performance. MILP switching control and multi-layer supercapacitor management ensure accurate execution of total power commands and that each cell operates within safe boundaries, ultimately achieving a synergistic improvement in the lifespan, economy, and operational reliability of the energy storage system. Attached Figure Description

[0016] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0017] Figure 1 A schematic diagram of the structure of an energy storage management system for coordinating power distribution and battery optimization provided in an embodiment of the present invention; Figure 2This is a schematic diagram illustrating the steps of an energy storage management method for coordinating power distribution and battery optimization, as provided in an embodiment of the present invention. Detailed Implementation

[0018] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0019] This embodiment provides an energy storage management system that coordinates power distribution and battery optimization, such as... Figure 1 As shown, it includes a coordination and scheduling layer, a power allocation layer, and a unit execution layer; The coordination and scheduling layer serves as the interface between the system and the external environment. It is responsible for receiving the raw power signal and allocating the raw power to two different types of energy storage devices, batteries and supercapacitors, through smoothing and frequency domain decomposition algorithms. At the same time, based on the preliminary equalization of the internal state of the battery pack, it outputs high-frequency power components and low-frequency power components, and then generates a preliminary power allocation scheme for the battery cells based on the low-frequency power components.

[0020] The coordination and scheduling layer includes a power smoothing and decomposition module and a battery power dynamic initial allocation module, wherein: The power smoothing and decomposition module is responsible for processing the original power signal, generating smooth power that meets grid connection requirements through an adaptive smoothing algorithm, and decomposing the total power of the hybrid energy storage to be carried by energy storage according to frequency characteristics using signal decomposition technology.

[0021] The process of decomposing the total power of the hybrid energy storage system that needs to be stored according to frequency characteristics includes the following steps: Power smoothing: An adaptive moving average algorithm is used to dynamically adjust the window length L and calculate the smoothed power until the grid connection standard is met or the maximum window limit is reached.

[0022] The formula for calculating the smoothing power is as follows: ; In the formula, P smooth [t] represents the smoothed power at time t; L is the sliding window length, determined by an adaptive algorithm (1≤L≤L). max ); The original power sample value at historical time ti; L max This is the preset maximum window length limit.

[0023] Hybrid power calculation: Calculate the difference between the original power signal and the smoothed power to obtain the total hybrid energy storage power that the energy storage system needs to bear.

[0024] Power allocation and frequency domain allocation: Optimize the parameters of VMD (number of modes K, penalty factor α) to decompose the total power of hybrid energy storage with the goal of minimizing composite entropy; then analyze the instantaneous frequency of each mode component through Hilbert transform, and merge the mode components according to the spectral characteristics to finally output high-frequency power components (pointing to supercapacitors) and low-frequency power components (pointing to batteries).

[0025] The battery power dynamic initial allocation module, based on the low-frequency power components obtained from the decomposition and combined with the real-time status data of each battery cell, performs dynamic power pre-allocation based on capacity and with SOC equalization as the goal, and generates a preliminary power allocation scheme for the battery cells.

[0026] The preliminary power allocation scheme for generating battery cells includes the following steps: Calculate the basic power allocation value: Calculate the capacity factor based on the rated capacity of each battery cell, and then allocate the low-frequency power component according to the capacity factor to obtain the basic power allocation value; The formula for calculating the capacity factor is as follows: ; In the formula, m i C is the capacity factor of the i-th battery cell; i The rated capacity (Ah) of the i-th battery cell; ∑ j C j This represents the total capacity of all battery cells.

[0027] The formula for calculating the basic power allocation value is: P base,i =P low ×m i ; In the formula, P base,i Assign power value based on the base, P low This represents the low-frequency power component.

[0028] Calculate the dynamic adjustment factor: Calculate the system average SOC, and based on the deviation of each unit's SOC from the average value, calculate the dynamic adjustment factor through a nonlinear function. The larger the deviation, the smaller the factor, in order to limit the power of that unit and prevent it from exceeding the limit.

[0029] The formula for calculating the dynamic adjustment factor is as follows: ; In the formula, α i α is the dynamic adjustment factor for the i-th battery cell; base is the base adjustment factor (usually set to 1); k is the adjustment coefficient, controlling the degree of influence of SOC difference on power allocation; SOC i The current state of charge (SOC) of the i-th battery cell. avgThe average SOC of all battery cells.

[0030] Generate initial power allocation scheme: Combine the basic power allocation value with the dynamic adjustment factor to calculate the initial power allocation of the battery. It also performs power limiting checks and outputs a preliminary allocation plan.

[0031] The formula for calculating the initial power allocation of the battery is as follows: ; In the formula, Initially allocate power to the battery.

[0032] The power allocation layer is used to receive the low-frequency power components and preliminary power allocation schemes of the battery cells issued by the coordination and scheduling layer. With the economic operation of the system and multi-objective coordination as the core, it outputs the final power allocation command of the battery system by solving the sequence optimization problem, which comprehensively considers economic cost, system loss and SOC balance.

[0033] The power allocation layer includes an upper-layer economic optimization module and a lower-layer multi-objective coordinated optimization module, wherein: The upper-level economic optimization module is used to solve a more economical power allocation scheme by constructing an optimization problem with the objective function of minimizing the total weighted operating cost of the system (internal loss cost and aging cost) starting from the initial power allocation of the battery.

[0034] The process of finding a more economical power allocation scheme includes the following steps: Calculate the resistance coefficient: comprehensively evaluate the degree of deviation of the SOC of each battery cell from the ideal value (0.5) and the SOH state, calculate the original resistance coefficient, and perform normalization processing.

[0035] Among them, the original resistance coefficient The calculation formula is: ; In the formula, SOC ref β is the SOC reference value, representing the battery's optimal operating point; β is the SOC deviation penalty coefficient, controlling the degree to which SOC deviation affects the resistance coefficient; γ is the SOH penalty coefficient, controlling the degree to which the state of health affects the resistance coefficient.

[0036] Constructing the economic objective function: Establish an objective function with the goal of minimizing the total weighted cost, expressed as: ; In the formula, C loss,i For internal loss costs, C aging,i For aging cost, N is the total number of battery cells, R is the cost of aging. i This is the normalized resistance coefficient.

[0037] The formulas for calculating internal damage cost and aging cost are as follows: ; ; In the formula, The unit internal loss cost coefficient (yuan / kWh); These are the coefficients of the quadratic and linear terms in the internal loss power model, respectively. The scheduling time step (hours); Investment i The investment cost (in yuan) for the i-th battery cell; The rated capacity (kWh) of the i-th battery cell; Let be the cycle life (times) of the i-th battery cell. In order to time step Energy throughput (kWh) within the system. For the first The power of each battery cell.

[0038] Solving the optimization problem: Under the constraints of power balance, SOC safety and unit power limit, algorithms such as sequential quadratic programming (SQP) are used to solve the objective function and output the economically optimized power.

[0039] The lower-level multi-objective coordination and optimization module is used to further coordinate the two conflicting objectives of total system loss and SOC balance on the basis of economically optimized power. It finds the Pareto optimal solution set through multi-objective optimization algorithm and selects the final compromise power allocation scheme according to the real-time operating conditions of the system.

[0040] The selection of the final compromise power allocation scheme based on the real-time operating conditions of the system includes the following steps: Construct a multi-objective optimization problem: Define two objective functions: Minimize total charge / discharge loss. Minimize SOC variance Constraints such as power balance, SOC safety, and cell power limits are set.

[0041] The expressions for the two objective functions are as follows: ; ; In the formula, The target for total charge / discharge loss is to minimize it; To achieve SOC balance, the goal is to minimize the variance. Let be the power loss of the i-th battery cell; Let be the overall efficiency of the i-th battery cell, which is a function of power; The predicted SOC value for the next time step. ; This is the predicted average SOC value for the system at the next time step.

[0042] Solving the Pareto front: Using a multi-objective genetic algorithm (MOGA), a set of non-dominated solutions (Pareto fronts) is generated through population evolution, and the objective function value of each solution is normalized.

[0043] Choosing a final compromise: Based on the current total power demand and the system's average health status, dynamically calculate the weights of the two objectives. and The weighted composite score of each solution in the Pareto front is calculated, and the solution with the highest score is selected as the final multi-objective optimized power of the battery.

[0044] The formula for calculating the weighted composite score for each solution is as follows: ; ; ; In the formula, w1 and w2 are the loss targets, respectively. and balance goals The weights; The basic weight for the loss target; This is the power demand weighting factor; when the total power demand The larger the value, the more it tends to optimize efficiency and reduce losses. This refers to the rated total power of the battery pack. This is a weighting coefficient for health status. The worse the mean health status (SOH), the more attention needs to be paid to degradation in order to extend lifespan. SOH i This represents the health status of the i-th battery cell. The weighted composite score for each Pareto solution; , This is the normalized objective function value.

[0045] Multi-objective power optimization P of the battery final,i for: .

[0046] The unit execution layer is used to receive the final power allocation command and high-frequency power component of the battery system sent down from the upper layer. Through mixed integer programming and high-frequency multi-layer management strategy, it transforms the final power allocation command and high-frequency power component of the battery system into specific energy storage unit switching states and precise power commands, so as to ensure accurate power tracking and real-time safety of the system.

[0047] The unit execution layer includes a low-frequency scale battery MILP control module and a high-frequency scale supercapacitor multi-layer management module, wherein: The low-frequency scale battery MILP control module, based on the battery's multi-objective power optimization, optimizes and determines the switching state of each battery cell by establishing a mixed integer linear programming model. While accurately tracking the total power, it further promotes the SOC balance among the activated cells, and obtains the optimal combination of battery switching states and corresponding power.

[0048] The optimal combination of battery switching states and the corresponding power includes the following steps: Problem modeling: Defining decision variables (switching states) and power Construct the objective function: ; In the formula, J is the value of the objective function. This represents the total power command obtained from the lower-level multi-objective optimization, i.e. ; The weighting coefficients of the SOC equalization term are used to adjust the relative importance between power point tracking (PPT) and SOC equalization; SOC i Indicates the first The current state of charge of each battery cell; This represents the average SOC of all battery cells in the active state. This indicates that this item applies only to the input unit ( ) Calculate the SOC equilibrium penalty.

[0049] Set constraints, including: Power balance guiding constraints: In the formula This represents the allowable power tracking error.

[0050] SOC security constraints (next cycle): In the formula, SOC i Let Δt be the state of charge of the i-th battery cell at the current moment. bat E represents the time step of the battery control layer. rated,i This represents the rated capacity of the i-th battery cell.

[0051] Power limit constraints: In the formula P max,i This represents the absolute value of the maximum allowable charging and discharging power of the i-th battery cell.

[0052] Switching logic and system operation constraints: At least one unit in any series branch must be engaged.

[0053] Solution and Output: The solution is obtained using a MILP solver (such as Gurobi or Cplex), and the optimal combination of battery switching states is output. And the corresponding power value after fine-tuning.

[0054] The high-frequency scale supercapacitor multi-layer management module is used to dynamically adjust the power of each supercapacitor unit for high-frequency power components through a four-layer management strategy that includes constraints, selection, optimization and compensation, so as to ensure that it operates within a safe range while accurately tracking power commands.

[0055] The dynamic adjustment of the power of each supercapacitor unit includes the following steps: Constraint layer: Real-time checks whether the SOC and power of each supercapacitor unit exceed safety limits, and forcibly prunes any power commands that exceed the limits; Selection layer: Based on the preset SOC range rules (e.g., SOC outside [0.4, 0.6]), select the cells that need to be optimized and adjusted from all supercapacitor cells; Optimization layer: For the selected cell, apply a charge and discharge power correction factor according to the SOC level (e.g., reduce the discharge factor and increase the charge factor if the SOC is low) to dynamically adjust its power command in order to guide the SOC back to the normal range. Compensation layer: Calculates the difference in total system power before and after optimization and adjustment, and the unselected units in the healthy SOC range share the power difference according to the capacity or SOC margin ratio to ensure that the final total power output is consistent with the command.

[0056] This embodiment also provides an energy storage management method that coordinates power distribution and battery optimization, such as... Figure 2 As shown, it includes the following steps: S1. Power Demand Smoothing and Decomposition: The original power signal is received, and a smoothed power that meets grid connection requirements is generated using an adaptive moving average algorithm. The total power of the hybrid energy storage system that needs to be borne by the energy storage system is calculated. Then, the variational mode decomposition (VMD) algorithm is used to decompose the total power of the hybrid energy storage system into multiple modal components. The frequency composition is analyzed using Hilbert transform, and finally, the modal components are merged based on the spectral characteristics, outputting high-frequency power components and low-frequency power components pointing to the supercapacitor and battery, respectively. This includes the following steps: S11 Power Smoothing Processing: An adaptive moving average algorithm is adopted. The initial window length L is set to 1, and L is gradually increased. The smoothed power under each L value is calculated. When the smoothed power meets the grid connection standard or L reaches the preset upper limit, the iteration stops and the final smoothed power is output.

[0057] The formula for calculating the smoothing power is as follows: ; In the formula, P smooth [t] represents the smoothed power at time t; L is the sliding window length, determined by an adaptive algorithm (1≤L≤L). max ); The original power sample value at historical time ti; L max This is a preset maximum window length limit to prevent excessive smoothing.

[0058] S12. Hybrid Power Calculation: Calculate the difference between the original power and the smoothed power as the total hybrid energy storage power.

[0059] S13. Power Allocation and Frequency Domain Allocation: The improved predator algorithm (IHPO) is used to optimize the number of modes K and the penalty factor α of VMD with the goal of minimizing the composite entropy. Then, VMD decomposition and Hilbert transform are performed. Based on the instantaneous frequency distribution, high-frequency modes are merged to generate the total high-frequency power, and low-frequency modes are merged to generate the total low-frequency power.

[0060] The minimized composite entropy is expressed as: ; Among them, E composite The composite entropy combines permutation entropy and mutual information entropy, and is used to measure the decomposition effect; u k Let K be the k-th intrinsic mode function (IMF); K is the number of modes, obtained through optimization using the IHPO algorithm (K... min ≤K≤K max ); α is the quadratic penalty factor, obtained through optimization using the IHPO algorithm (α) min ≤α≤α max ).

[0061] S2. Initial Dynamic Allocation of Battery Power: Based on low-frequency power components and combined with real-time status data (capacity, SOC) of each battery cell, the basic allocated power value is first calculated according to the capacity ratio; then, the difference between the SOC of each cell and the system average SOC is calculated, and a dynamic adjustment factor is calculated based on the difference; finally, the basic allocated power value and the dynamic adjustment factor are combined to generate the initial battery power allocation, including the following steps: S21. Calculate the basic power allocation value: Calculate the weight of each battery cell in the total capacity based on its rated capacity as the capacity factor, and then allocate the low-frequency power component according to the capacity factor to obtain the basic power allocation value based only on the capacity. The formula for calculating the capacity factor is as follows: ; In the formula, m i C is the capacity factor of the i-th battery cell; i The rated capacity (Ah) of the i-th battery cell; ∑ j Cj This represents the total capacity of all battery cells.

[0062] The formula for calculating the basic power allocation value is: P base,i =P low ×m i ; In the formula, P base,i Assign power value based on the base, P low This represents the low-frequency power component.

[0063] S22. Calculate the dynamic adjustment factor: Calculate the global SOC average value, then obtain the SOC deviation of each unit, and determine the dynamic adjustment factor nonlinearly based on the magnitude of the deviation.

[0064] The formula for calculating the dynamic adjustment factor is as follows: ; In the formula, α i α is the dynamic adjustment factor for the i-th battery cell; base is the base adjustment factor (usually set to 1); k is the adjustment coefficient, controlling the degree of influence of SOC difference on power allocation; SOC i The current state of charge (SOC) of the i-th battery cell. avg The average SOC of all battery cells.

[0065] S23. Generate initial allocation scheme: Combine the basic allocation power value with the dynamic adjustment factor, and perform power limiting check to generate the final initial battery allocation power. The formula for calculating the initial power allocation of the battery is as follows: ; In the formula, Initially allocate power to the battery.

[0066] It should be noted that the dynamic adjustment factor is set to a smaller value when the SOC difference is large to prevent power over-limit, and a larger value when the SOC difference is small to accelerate the balancing process.

[0067] S3. Upper-level economic optimization: Taking the initial power allocation of the battery as the initial solution, the optimization aims to minimize the total operating cost of the system. First, the resistance coefficient is calculated based on the SOC and state of health (SOH) of each unit. Then, an economic objective function is constructed that includes internal damage cost and aging cost, and the resistance coefficient is incorporated as a weight. Finally, the solution is obtained under the constraint of satisfying the total power balance, and the economically optimized power is output. This includes the following steps: S31. Calculate the resistance coefficient: comprehensively assess the deviation of the SOC from the ideal value (usually taken as...). The original resistance coefficients were calculated based on the degree of resistance and the state of state of somatic oxidase (SOH), and the original resistance coefficients of all units were normalized to obtain normalized resistance coefficients.

[0068] The formula for calculating the original resistance coefficient is as follows: ; In the formula, SOCref is the SOC reference value, usually set to 0.5, representing the optimal operating point of the battery; β is the SOC deviation penalty coefficient, which controls the degree of influence of SOC deviation on the resistance coefficient; γ is the SOH penalty coefficient, which controls the degree of influence of the health state on the resistance coefficient. The original resistance coefficient reflects the degree of "unwillingness" of a single unit.

[0069] Understandably, the normalized resistance coefficient is used for comparison and weighting between units.

[0070] S32. Construct and solve the economic objective function: Establish an objective function with the goal of minimizing the total weighted cost, expressed as: ; In the formula, C loss,i For internal loss costs, C aging,i For aging cost, N is the total number of battery cells, R is the cost of aging. i This is the normalized resistance coefficient.

[0071] The formulas for calculating internal damage cost and aging cost are as follows: ; ; In the formula, The unit internal loss cost coefficient (yuan / kWh); These are the coefficients of the quadratic and linear terms in the internal loss power model, respectively. The scheduling time step (hours); Investment i The investment cost (in yuan) for the i-th battery cell; The rated capacity (kWh) of the i-th battery cell; Let be the cycle life (times) of the i-th battery cell. In order to time step Energy throughput (kWh) within the system. For the first The power of each battery cell.

[0072] S33. Solving optimization problems: Set constraints and use mathematical programming algorithms (such as sequential quadratic programming, SQP) to solve the objective function; the solver will find a specific set of... The value that minimizes the objective function and satisfies all constraints is the optimal economic power.

[0073] The constraints include: Power balance constraints: ; SOC security constraints: ; Unit power limit constraints: ; in, , These are the lower and upper safety limits of the SOC, respectively; C i The capacity (kWh) of the i-th battery cell; , These are the maximum discharge and charging power of the i-th battery cell, respectively.

[0074] Specifically, the solver sets the initial power allocation of the battery as the decision variable. The initial value is used to iterate according to its algorithm (such as calculating the gradient, constructing an approximate model, etc.), and finally converges to the optimal solution that satisfies all constraints and minimizes the objective function.

[0075] It should be noted that the resistance coefficient reflects the "willingness" to participate in frequency modulation (the worse the SOC or SOH, the greater the resistance coefficient).

[0076] S4. Lower-level multi-objective coordinated optimization: Based on economically optimized power, a multi-objective function is constructed that includes minimizing losses and minimizing SOC variance. A series of non-dominated solutions (Pareto fronts) are obtained using multi-objective optimization algorithms (such as MOGA). Then, based on the current system operating conditions (such as the total power demand and SOC dispersion), decision rules are formulated to select the multi-objective optimized power of the battery from the Pareto front, including the following steps: S41. Construct a multi-objective optimization problem: Define two objective functions and establish constraints that include system dynamics. The two objective functions include minimizing total charge and discharge losses and maximizing SOC balance (i.e., minimizing SOC variance). Specifically, minimizing total charge and discharge losses aims to minimize the total energy loss of all battery cells during power conversion, including battery internal resistance losses and power conversion system losses. Maximizing SOC balance aims to make the state of charge of all battery cells as close as possible at the end of the optimization cycle, so as to avoid individual cells prematurely exiting operation due to excessively high or low SOC, thereby improving the system's available capacity and lifespan.

[0077] The expressions for the two objective functions are as follows: ; ; In the formula, The target for total charge / discharge loss is to minimize it; To achieve SOC balance, the goal is to minimize the variance. Let be the power loss of the i-th battery cell; Let be the overall efficiency of the i-th battery cell, which is a function of power; The predicted SOC value for the next time step. ; This is the predicted average SOC value for the system at the next time step.

[0078] The constraints include: Power balance constraints: ; SOC security constraints: ; Unit power limit constraints: .

[0079] S42. Solving the Pareto Front: Using a multi-objective genetic algorithm (MOGA), a set of non-dominated solutions is generated through population evolution (selection, crossover, mutation), forming the Pareto front. Each solution represents a power allocation scheme and cannot be surpassed by any other solution simultaneously on both objectives. The objective function value of each solution in the Pareto front is normalized.

[0080] S43. Select the final compromise solution: Based on the real-time operating status of the system, dynamically calculate the weights of the two objectives and select the solution with the highest comprehensive score.

[0081] The formula for calculating the weights of the two objectives is as follows: ; ; ; In the formula, w1 and w2 are the loss targets, respectively. and balance goals The weights; The basic weight for the loss target; This is the power demand weighting factor; when the total power demand The larger the value, the more it tends to optimize efficiency and reduce losses. This refers to the rated total power of the battery pack. This is a weighting coefficient for health status. The worse the mean health status (SOH), the more attention needs to be paid to degradation in order to extend lifespan. SOH i The health state of the i-th battery cell; the score is the weighted composite score for each Pareto solution; , This is the normalized objective function value.

[0082] The final determined number Multi-objective power optimization P of each battery cell final,i for: .

[0083] S5. Multi-timescale collaborative control: Based on the multi-objective power optimization of the battery, mixed integer linear programming (MILP) is used to optimize the switching state of the battery system. The goal is to achieve SOC equalization while tracking the total power and output the battery switching state. Based on the high-frequency power component of the supercapacitor system, a management strategy including constraints, selection, optimization, and compensation is used to dynamically adjust the power of each capacitor unit and output the supercapacitor switching state, including the following steps: S51. Low-frequency scale battery MILP control: Establish and solve a mixed-integer linear programming problem to determine the optimal switching state of each battery cell, achieving power distribution and SOC balance, including: Definition of decision variables: Indicates the first The switching status of each battery cell, 1 for engaged and 0 for disengaged; Indicates the first Power of each battery cell (kW); Note that when At that time, forced .

[0084] Objective function construction: The objective function is established by comprehensively considering power tracking accuracy and SOC equalization. ; In the formula, J is the value of the objective function. This represents the total power command obtained from the lower-level multi-objective optimization, i.e. ; The weighting coefficients of the SOC equalization term are used to adjust the relative importance between power tracking and SOC equalization. Indicates the first The current state of charge of each battery cell; This represents the average SOC of all battery cells in the active state. This indicates that this item applies only to the input unit ( ) Calculate the SOC equilibrium penalty.

[0085] The constraints include: Power balance guiding constraints: In the formula This represents the allowable power tracking error.

[0086] SOC security constraints (next cycle): In the formula Let Δt be the state of charge of the i-th battery cell at the current moment. bat E represents the time step of the battery control layer. rated,i This represents the rated capacity of the i-th battery cell.

[0087] Power limit constraints: In the formula P max,i This represents the absolute value of the maximum allowable charging and discharging power of the i-th battery cell.

[0088] Switching logic and system operation constraints: At least one unit in any series branch must be engaged.

[0089] Use a MILP solver (such as Gurobi or Cplex) to solve the objective function and obtain the optimal combination of battery switching states and the corresponding power.

[0090] S52. Multi-layer Management of High-Frequency Scale Supercapacitors: The constraint layer checks whether the supercapacitor's state of charge (SOC) and power are within safe ranges, and removes those exceeding the limits. The selection layer selects cells to be optimized according to SOC range rules. For example, cells with an SOC outside [0.4, 0.6] are selected. The optimization layer applies charge / discharge power correction coefficients to the selected cells to adjust their power. For example, when SOC < 0.4, the charging coefficient linearly decreases from 1.2 to 1.0, and the discharging coefficient linearly increases from 0.1 to 1.0; when SOC > 0.6, the charging coefficient linearly decreases from 1.0 to 0.1, and the discharging coefficient linearly increases from 1.0 to 1.2. The compensation layer calculates the power difference ΔP before and after optimization, and the unselected cells share the difference according to their capacity or SOC margin ratio, keeping the total power constant.

[0091] This invention effectively separates high- and low-frequency power components through adaptive moving average and VMD decomposition, allowing the supercapacitor to handle high-frequency pulsating power while the battery handles low-frequency energy-type power, thereby reducing the charging and discharging stress and aging rate of the battery. Dynamic initial allocation based on real-time SOC introduces an equalization mechanism at the beginning of power allocation, avoiding deep charging and discharging of some battery cells and improving the consistency of the battery pack. Introducing a resistance coefficient in the upper-level economic optimization transforms cell state differences into economic weights, guiding power to be preferentially handled by cells in better states, reducing the total system operating cost. Lower-level multi-objective coordinated optimization solves the Pareto front and makes dynamic decisions, achieving an optimal trade-off between system losses and SOC equalization, improving the overall system performance. MILP switching control and multi-layer supercapacitor management ensure accurate execution of total power commands and that each cell operates within safe boundaries, ultimately achieving a synergistic improvement in the lifespan, economy, and operational reliability of the energy storage system.

[0092] This invention adaptively smooths the original power signal to generate a smoothed power that meets grid connection requirements and calculates the total power of the hybrid energy storage. Based on this, an optimized variational mode decomposition algorithm is used to accurately decompose the total power into high-frequency and low-frequency components according to frequency characteristics, thereby allocating power with different characteristics to the most suitable energy storage elements (supercapacitors and batteries) from the source. Then, using the low-frequency power components obtained from the decomposition, combined with the real-time status data of the battery cells, a dynamic initial allocation is performed to generate a power plan that has considered preliminary equalization. Then, starting from this plan, upper-level economic optimization and lower-level multi-objective coordination optimization are performed sequentially to transform a single power allocation command into an optimal solution that comprehensively considers operating costs, system losses, and battery health status. Finally, through execution-level control strategies such as mixed integer linear programming and multi-layer management of supercapacitors, the optimized power command is transformed into the specific switching state and power setpoint of each energy storage unit, forming a closed-loop precise control from macro power command to micro unit action, ultimately achieving multiple goals of improving economy, extending lifespan, and ensuring safety.

[0093] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. An energy storage management system that coordinates power distribution and battery optimization, characterized in that: It includes a coordination and scheduling layer, a power allocation layer, and a unit execution layer; The coordination and scheduling layer is used as the interface between the system and the external environment. It is responsible for receiving the original power signal, generating smooth power that meets the grid connection requirements through smoothing and frequency domain decomposition algorithms, and using frequency domain decomposition technology to decouple the total power of hybrid energy storage into high-frequency power components and low-frequency power components. Then, it generates a preliminary power allocation scheme for battery cells based on the low-frequency power components. The power allocation layer is used to receive the low-frequency power components and preliminary power allocation schemes of the battery cells issued by the coordination and scheduling layer. With the economic operation of the system and multi-objective coordination as the core, it outputs the final power allocation instruction of the battery system that comprehensively considers economic cost, system loss and SOC balance through sequence optimization solution. The unit execution layer is used to receive the final power allocation command and high-frequency power component of the battery system sent down from the upper layer. Through mixed integer programming and high-frequency multi-layer management strategy, it transforms the final power allocation command and high-frequency power component of the battery system into specific energy storage unit switching status and power command. The coordination and scheduling layer includes a battery power dynamic initial allocation module. Based on the low-frequency power components obtained by decomposition and combined with the real-time status data of each battery cell, the battery power dynamic initial allocation module performs dynamic power pre-allocation based on capacity and with SOC balance as the goal, and generates a preliminary power allocation scheme for the battery cells. The preliminary power allocation scheme for generating battery cells includes the following steps: Calculate the basic power allocation value: Calculate the capacity factor based on the rated capacity of each battery cell, and then allocate the low-frequency power component according to the capacity factor to obtain the basic power allocation value; Calculate the dynamic adjustment factor: Calculate the system average SOC, and based on the deviation of each unit's SOC from the average value, calculate the dynamic adjustment factor through a nonlinear function; Generate initial allocation scheme: Combine the basic allocation power value with the dynamic adjustment factor to calculate the initial battery allocation power, perform power limit check, and output the initial allocation scheme.

2. The energy storage management system for coordinating power distribution and battery optimization according to claim 1, characterized in that: The coordination and scheduling layer also includes a power smoothing and decomposition module; The power smoothing and decomposition module is responsible for processing the original power signal, generating smooth power that meets grid connection requirements through an adaptive smoothing algorithm, and decomposing the total power of the hybrid energy storage that needs to be carried by energy storage according to frequency characteristics using signal decomposition technology. The process of decomposing the total power of the hybrid energy storage system that needs to be stored according to frequency characteristics includes the following steps: Power smoothing: An adaptive moving average algorithm is used to dynamically adjust the window length L and calculate the smoothed power until the grid connection standard is met or the maximum window limit is reached. The formula for calculating the smoothing power is as follows: ； In the formula, P smooth [t] represents the smoothing power at time t; L represents the sliding window length; The original power sample value at historical time ti; L max This is the preset maximum window length limit; Hybrid power calculation: Calculate the difference between the original power signal and the smoothed power to obtain the total hybrid energy storage power that the energy storage system needs to bear; Power allocation and frequency domain allocation: Optimize VMD parameters to decompose the total power of hybrid energy storage with the goal of minimizing composite entropy; then analyze the instantaneous frequency of each modal component through Hilbert transform, merge the modal components according to the spectral characteristics, and finally output the high-frequency power component and the low-frequency power component.

3. The energy storage management system for coordinating power distribution and battery optimization according to claim 1, characterized in that: The formula for calculating the capacity factor is: ； In the formula, m i C is the capacity factor of the i-th battery cell; i The rated capacity (Ah) of the i-th battery cell; ∑ j C j The total capacity of all battery cells; The formula for calculating the basic power allocation value is as follows: P base,i =P low ×m i ; In the formula, P base,i Assign power value based on the base, P low Low-frequency power components; The formula for calculating the dynamic adjustment factor is as follows: ； In the formula, α i α is the dynamic adjustment factor for the i-th battery cell; base is the basic adjustment factor; k is the adjustment coefficient; SOC i The current state of charge (SOC) of the i-th battery cell. avg The average SOC of all battery cells; The formula for calculating the initial power allocation of the battery is as follows: ； In the formula, Initially allocate power to the battery.

4. The energy storage management system for coordinating power distribution and battery optimization according to claim 1, characterized in that: The power allocation layer includes an upper-layer economic optimization module and a lower-layer multi-objective coordinated optimization module, wherein: The upper-level economic optimization module is used to solve a more economical power allocation scheme by constructing an optimization problem with the objective function of minimizing the total weighted operating cost of the system, starting from the initial power allocation of the battery. The process of finding a more economical power allocation scheme includes the following steps: Calculate the resistance coefficient: comprehensively evaluate the degree of deviation of the SOC from the ideal value and the SOH state of each battery cell, calculate the original resistance coefficient, and perform normalization processing; Constructing the economic objective function: Establish an objective function with the goal of minimizing the total weighted cost, expressed as: ； In the formula, C loss,i For internal loss costs, C aging,i For aging cost, N is the total number of battery cells, R is the cost of aging. i This is the normalized resistance coefficient; Solving the optimization problem: Under the constraints of power balance, SOC safety and unit power limits, the objective function is solved using a sequential quadratic programming algorithm, and the economically optimized power is output.

5. The energy storage management system for coordinating power distribution and battery optimization according to claim 4, characterized in that: The original resistance coefficient The calculation formula is: ； In the formula, SOC ref β is the SOC reference value; β is the SOC deviation penalty coefficient; γ is the SOH penalty coefficient; The formulas for calculating the internal damage cost and the aging cost are as follows: ； ； In the formula, This is the unit internal loss cost coefficient; These are the coefficients of the quadratic and linear terms in the internal loss power model, respectively. For scheduling time steps; Investment i Let be the investment cost of the i-th battery cell; Let be the rated capacity of the i-th battery cell; Let be the cycle life of the i-th battery cell; In order to time step Internal energy throughput For the first The power of each battery cell.

6. The energy storage management system for coordinating power distribution and battery optimization according to claim 5, characterized in that: The lower-level multi-objective coordination and optimization module is used to coordinate the two conflicting objectives of total system loss and SOC balance on the basis of economically optimized power. It finds the Pareto optimal solution set through multi-objective optimization algorithm and selects the final compromise power allocation scheme according to the real-time operating conditions of the system. The selection of the final compromise power allocation scheme based on the real-time operating conditions of the system includes the following steps: Construct a multi-objective optimization problem: Define two objective functions: Minimize total charge / discharge loss. Minimize SOC variance And set constraints for power balance, SOC safety and cell power limits; Solving the Pareto front: A multi-objective genetic algorithm is used to generate a set of non-dominated solutions through population evolution, and the objective function value of each solution is normalized. Choosing a final compromise: Based on the current total power demand and the system's average health status, dynamically calculate the weights of the two objectives. and The weighted composite score of each solution in the Pareto front is calculated, and the solution with the highest score is selected as the final multi-objective optimized power of the battery.

7. The energy storage management system for coordinating power distribution and battery optimization according to claim 6, characterized in that: The expressions for the two objective functions are: ； ； In the formula, The target is the total charge and discharge loss. The goal is to achieve SOC balance. Let be the power loss of the i-th battery cell; Let be the overall efficiency of the i-th battery cell; The predicted SOC value for the next time step. ; The predicted average SOC value for the next time step; The formula for calculating the weighted composite score of each solution is as follows: ； ； ； In the formula, w1 and w2 are the loss targets, respectively. and balance goals The weights; The basic weight for the loss target; This is the power demand weighting coefficient; This refers to the rated total power of the battery pack. Total power requirement; For health status weighting coefficients; SOH i The health state of the i-th battery cell; the score is the weighted composite score for each Pareto solution; , The objective function value is the normalized value; N is the total number of battery cells; The battery's multi-objective optimized power P final,i for: 。 8. The energy storage management system for coordinating power distribution and battery optimization according to claim 1, characterized in that: The unit execution layer includes a low-frequency scale battery MILP control module and a high-frequency scale supercapacitor multi-layer management module, wherein: The low-frequency scale battery MILP control module, based on the battery's multi-objective power optimization, optimizes and determines the switching state of each battery cell by establishing a mixed integer linear programming model, and obtains the optimal combination of battery switching states and corresponding power. The optimal combination of battery switching states and the corresponding power includes the following steps: Problem modeling: Define decision variables and construct the objective function. ； In the formula, J is the objective function value; N is the total number of battery cells; This represents the total power command obtained from the lower-level multi-objective optimization; Represents the weighting coefficients of the SOC equilibrium term; SOC i Indicates the first The current state of charge of each battery cell; This represents the average SOC of all battery cells in the active state. This indicates that the SOC balance penalty is calculated only for the input units; Among them, the decision variables include the switching state. and power ; Set constraints; Switching logic and system operation constraints: At least one unit in any series branch must be switched on; Solution and Output: The solution is obtained using the MILP solver, and the optimal combination of battery switching states is output. And the corresponding power value after fine-tuning.

9. The energy storage management system for coordinating power distribution and battery optimization according to claim 8, characterized in that: The constraints include: Power balance guiding constraints: In the formula For the allowable power tracking error; SOC security constraints: In the formula, SOC i The current state of charge of the i-th battery cell, Δt bat E represents the time step of the battery control layer. rated,i This represents the rated capacity of the i-th battery cell; Power limit constraints: In the formula P max,i This represents the absolute value of the maximum allowable charging and discharging power of the i-th battery cell.

10. An energy storage management system for coordinating power distribution and battery optimization according to claim 9, characterized in that: The high-frequency scale supercapacitor multi-layer management module is used to dynamically adjust the power of each supercapacitor unit for high-frequency power components through a four-layer management strategy that includes constraints, selection, optimization and compensation. The dynamic adjustment of the power of each supercapacitor unit includes the following steps: Constraint layer: Real-time checks whether the SOC and power of each supercapacitor unit exceed safety limits, and forcibly prunes any power commands that exceed the limits; Selection layer: Based on the preset SOC range rules, select the cells that need to be optimized and adjusted from all supercapacitor cells; Optimization layer: For the selected cells, the power command is dynamically adjusted by applying a charge / discharge power correction factor based on the SOC level, so as to guide the SOC back to the normal range. Compensation layer: Calculates the difference in total system power before and after optimization and adjustment, and the unselected units in the healthy SOC range share the power difference according to the capacity or SOC margin ratio.

Images