An energy storage SOC equalization control method based on IHHO algorithm

Abstract

The application discloses a kind of energy storage SOC equalization control methods based on IHHO algorithm, comprising the following steps: S1, the state of charge and output voltage of each energy storage unit are collected, the state of charge and voltage value are exchanged by consistency communication, the average state of charge and average voltage of all energy storage units are calculated;S2, the power distribution factor and acceleration factor considering capacity difference are introduced, according to the deviation of the current state of charge and average state of charge of each energy storage unit, the expected output power of each energy storage unit is calculated;S3, the weighted sum of power deviation and bus voltage deviation is used to construct objective function.The beneficial effects of the present application: the power distribution factor and acceleration factor considering capacity difference are introduced in the present application, and the power deviation and voltage deviation are weighted in the objective function, the optimal droop coefficient and reference voltage are obtained by real-time optimization using improved Harris eagle optimization algorithm, which can accelerate the state of charge equalization speed between energy storage units with different capacities.

Description

Technical Field

[0001] This invention relates to the technical field of energy storage system control in DC microgrids, specifically to an energy storage SOC equalization control method based on the IHHO algorithm. Background Technology

[0002] In DC microgrids, distributed power sources such as photovoltaics and wind power exhibit power fluctuations due to environmental factors. To ensure stable system operation, multiple energy storage units are typically connected to balance the power between the load and the power source. In practical applications, different energy storage units differ in terms of state of charge (SOC), capacity, and line impedance. If traditional droop control is used for power distribution, the following problems arise: First, traditional droop control achieves power distribution by introducing virtual impedance into the line. An excessively large droop coefficient can lead to a severe drop in bus voltage, while an excessively small droop coefficient cannot effectively suppress the impact of line impedance on distribution accuracy. Second, in scenarios where multiple energy storage units of different capacities operate in parallel, the SOC equalization speed between units is slow, easily leading to overcharging or over-discharging, thereby shortening battery life and reducing system stability. Third, existing adaptive droop control strategies mostly rely on fixed calculation formulas or empirical parameter tuning, making it difficult to achieve a dynamic optimal trade-off between SOC equalization speed and bus voltage drop. When system operating conditions change, the control effect significantly decreases.

[0003] To address the aforementioned issues, existing research has attempted to introduce intelligent optimization algorithms for online tuning of the droop coefficient. However, the traditional HHO algorithm suffers from drawbacks when solving multi-peak, multi-dimensional, and nonlinear optimization problems, such as an imbalance between global exploration and local exploitation capabilities, a tendency to get trapped in local optima, and insufficient convergence accuracy. As a result, it is difficult to directly apply it to the real-time SOC equalization control of energy storage systems. Summary of the Invention

[0004] Purpose of the invention: To provide a storage SOC equalization control method based on the IHHO algorithm to solve the above-mentioned problems existing in the prior art.

[0005] Technical Solution: A SOC equalization control method for energy storage based on the IHHO algorithm, comprising the following steps: S1, collecting the state of charge and output voltage of each energy storage unit, exchanging the state of charge and voltage values ​​via consensus communication, and calculating the average state of charge and average voltage of all energy storage units; S2, introducing a power allocation factor and acceleration factor considering capacity differences, and calculating the expected output power of each energy storage unit based on the deviation between the current state of charge and the average state of charge; S3, constructing an objective function using a weighted sum of power deviation and bus voltage deviation; S4, using an improved Harris Eagle optimization algorithm to optimize the objective function in real time, and outputting the optimal droop coefficient and optimal reference voltage in real time; S5, substituting the real-time output optimal droop coefficient and optimal reference voltage into the droop control formula to generate control signals for each energy storage unit, and starting the next round of calculation; the method is applied to the scenario of parallel operation of energy storage units of different capacities in an islanded DC microgrid.

[0006] Preferably, the power deviation is calculated based on the difference between the actual output power and the expected output power of each energy storage unit; the voltage deviation is calculated based on the difference between the reference voltage and the actual output voltage of each energy storage unit.

[0007] Preferably, the improved Harris Eagle optimization algorithm introduces an improved escape energy formula and a hierarchical differential evolution mutation strategy on the basis of the traditional Harris Eagle optimization algorithm; the improved escape energy formula achieves a dynamic balance between global exploration and local exploitation by introducing random perturbation and adaptive exponential adjustment factor; the hierarchical differential evolution mutation strategy adopts different scaling factors and crossover probability parameters in the hard encirclement stage and the soft encirclement stage, respectively.

[0008] Preferably, the droop control formula is: ; in, U ref This is the reference voltage for the DC bus. U i and P i The first i The output voltage and output power of each energy storage unit; K i This is the droop coefficient.

[0009] Preferably, the expression for the output power of the energy storage unit is: ; Combining this with the droop control formula, we get: ; in, U bus This is the DC bus voltage. Idci For the first i The output current of each energy storage unit R i For the first i Line impedance.

[0010] Preferably, the expression for the power allocation factor is: ; in, m i Representing the i Power allocation factor of each energy storage unit CB i Representing the i The energy storage capacity of each energy storage unit The average state of charge of all energy storage units. SOC i For the first i The state of charge of each energy storage unit; The expression for the acceleration factor is: ; in, Q i For the first i The acceleration factor of each energy storage unit is used to replace the fixed coefficient and speed up the convergence. q , p To adjust the parameters, t This is the adjustment coefficient.

[0011] Preferably, the desired output power of the energy storage unit is: ; in, For the first i The expected output power of each energy storage unit; The global power deviation formula is defined as follows: ; in, F 1 For global power deviation, P ave The average output power of all energy storage units; The expression for average output power is: ; The global voltage deviation formula is: ; in, F 2 For global voltage deviation, U NiThis refers to the reference voltage values ​​of each energy storage unit output in real time using the IHHO algorithm.

[0012] Preferably, the global power deviation F 1 and global voltage deviation F 2 The comprehensive optimization objective function is defined as follows: ; in, F To comprehensively optimize the objective function, c The weighting coefficients of the power deviation function affect the SOC equalization speed of each energy storage unit.

[0013] The beneficial effects of this invention are as follows: By introducing a power allocation factor and an acceleration factor that consider capacity differences, and weighting power deviation and voltage deviation in the objective function, this invention utilizes an improved Harris Eagle optimization algorithm to obtain the optimal droop coefficient and reference voltage in real time, thereby accelerating the state-of-charge equalization speed among energy storage units of different capacities. Furthermore, by introducing an improved escape energy formula and a hierarchical differential evolution mutation strategy into the improved Harris Eagle optimization algorithm, this invention balances global exploration and local exploitation capabilities, effectively avoiding the defect of traditional Harris Eagle optimization algorithms being prone to getting trapped in local optima, and improving the convergence accuracy of the algorithm in multi-peak, multi-dimensional, and nonlinear optimization problems. Finally, this invention uses an improved Harris Eagle optimization algorithm to output the optimal droop coefficient and reference voltage in real time, eliminating the need for manual parameter tuning based on experience. Compared to existing adaptive droop control strategies that rely on fixed calculation formulas, this invention exhibits stronger adaptability and robustness under different operating conditions, enabling rapid state-of-charge equalization while controlling bus voltage deviation. Attached Figure Description

[0014] Figure 1 This is a typical DC microgrid structure diagram of the present invention; Figure 2 This is a block diagram of the optimized droop control of the present invention; Figure 3 This is a flowchart of the IHHO algorithm of the present invention; Figure 4 This is a diagram illustrating the SOC convergence and consistency process of the present invention. Figure 5 This is a power distribution diagram for each energy storage unit in this invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Specific Implementation

[0016] As attached Figure 1-5 As shown in this embodiment, an energy storage SOC equalization control method based on the IHHO algorithm includes the following steps: S1. Collect the state of charge and output voltage of each energy storage unit, exchange the state of charge and voltage values ​​through consistent communication, and calculate the average state of charge and average voltage of all energy storage units. S2. Introduce a power allocation factor and acceleration factor that take into account capacity differences, and calculate the expected output power of each energy storage unit based on the deviation between the current state of charge and the average state of charge of each energy storage unit. S3. Construct the objective function using the weighted sum of power deviation and bus voltage deviation; S4. The improved Harris Eagle optimization algorithm is used to optimize the objective function in real time, and the optimal droop coefficient and the optimal reference voltage are output in real time. S5. Substitute the real-time output of the optimal droop coefficient and the optimal reference voltage into the droop control formula to generate control signals for each energy storage unit and start the next round of calculation. The method is applied to scenarios where energy storage units of different capacities operate in parallel in an isolated DC microgrid.

[0017] like Figure 2 As shown, each energy storage unit obtains its own SOC and converter output voltage through sensor sampling, and sends the adjacent data based on consensus communication to the central controller to obtain the average state of charge and bus voltage Ubus.

[0018] The HHO algorithm is a swarm intelligence algorithm that simulates the group hunting behavior of Harris eagles to achieve global optimization. The algorithm consists of two phases: global exploration and local exploitation. It dynamically switches between exploration and exploitation strategies to update the Harris eagles' positions to approach prey and obtain the optimal solution. Its main principles are as follows: Global exploration phase To balance the search and development phases, the HHO algorithm defines the escape energy. E To simulate the process of prey fleeing during a Harris eagle chase, the expression is as follows: ; in, E This represents the energy used by the prey to escape. E0 represents the initial escape energy, expressed as a random number in the range (-1, 1); T represents the maximum number of iterations.

[0019] When |E|≥1, the algorithm is in the global exploration phase, searching for the approximate location of the prey, and its expression is as follows: ; Where X( t ), X( t+ 1) Represent the first in the Harris Eagle algorithm. t and t The eagle's position at +1 iteration; X rand ( t ) is the first t The eagle's random position in the next iteration; X rabbit ( t ) represents the first t Prey location at the next iteration; rand and q All are random numbers between (0,1); UB and LB represent the upper and lower limits of the search range.

[0020] Partial development phase When the escape energy |E| < 1, the algorithm enters the local development phase. Define random numbers. r This represents successful escape (r < 0.5) and unsuccessful escape (r < 0.5). r The chance of ≥0.5 is used to classify the local development stage into four cases based on the escape energy magnitude.

[0021] Case 1: Soft encirclement (|E|≥0.5, r≥0.5). In this case, the prey has sufficient escape energy but fails to escape. The Harris Hawk employs a soft encirclement strategy, expressed as: ; in, J The jump energy is a random number between (0,2).

[0022] Case 2: Hard enclosure (|E|<0.5, r ≥0.5). At this point, the prey has no energy left to escape the encirclement. The Harris Eagle adopts a hard-surround strategy, and the position expression is: .

[0023] Scenario 3: Gradual, rapid dive with soft encirclement (|E|≥0.5, r <0.5). At this point, the prey has sufficient energy to escape, and the Harris Eagle employs a more intelligent strategy to surround and capture it, adjusting its coordinates according to the prey's location, based on the formula... Update the current position and compare it with the fitness of the current position. If there is no improvement after comparison, the encirclement fails, and the formula is used. Perform a Levy-based flight walk. Here, S is a D-dimensional random vector; LF represents the Levy flight function.

[0024] Scenario 4: Gradual, rapid dive-and-surround (|E|<0.5, r<0.5). The prey has a chance to escape the encirclement, but lacks sufficient escape energy. At this point, the Harris Eagle initiates a hard surround, attempting to reduce the distance to the prey, with position updates as shown in the formula. As shown. If a rapid dive fails, a levy flight will be performed.

[0025] Traditional HHO algorithms suffer from drawbacks such as an imbalance between global exploration and local exploitation due to the linear decrease in escape energy E, and an overly simplistic exploration strategy. When dealing with complex, multi-modal, multi-dimensional, and nonlinear problems, they are prone to getting trapped in local optima and exhibit low convergence accuracy. To address these issues, an improved IHHO algorithm is proposed. Figure 3 Here is its algorithm flowchart.

[0026] Improvement 1: Improved escape energy formula To improve the algorithm's optimization speed and accuracy, the escape energy formula needs to be modified to balance the global exploration and local exploitation strategies. Considering that prey escape often has a certain degree of randomness in reality, this paper introduces random perturbation and an adaptive exponential adjustment factor into the formula, resulting in the improved formula shown below: ; in, t T and T represent the current iteration number and the maximum iteration number, respectively; rand A random number between (0,1); m and n These represent the initial and final values ​​of E, respectively. Through the synergistic effect of random perturbation and the exponential adjustment factor, the escape energy can be adaptively adjusted during the iteration process. This strategy ensures a balance between exploration and exploitation in the HHO algorithm throughout the iterative optimization process, thereby improving the overall performance and convergence accuracy of the algorithm.

[0027] Improvement 2: Hierarchical Differential Evolutionary Mutation Strategy Traditional HHO algorithms often suffer from decreased population diversity and dependence on the current optimal solution in the later stages of iteration. Once the optimal solution gets trapped in a local optimum, the population struggles to escape this localized region. Therefore, this paper introduces a hierarchical differential evolution mutation strategy, employing differentiated mutations at different siege stages to ensure population diversity while enhancing the algorithm's local exploitation capabilities. The mutation and crossover operation formulas for differential evolution are as follows: ; ; in, Representing the t The generation i The variation vector of each individual; F CR is the scaling factor; , , Three distinct individuals are randomly selected from the current population; This is the test vector generated after the crossover operation.

[0028] Based on the characteristics of the HHO algorithm at different siege stages, this paper designs a hierarchical differential evolution strategy as follows: Hard encirclement phase: At this point, the prey lacks the energy to escape, so a scaling factor is selected. F= 0.3 + 0.4 •(t / T) It exhibits a linearly increasing characteristic; the crossover probability is set to CR=0.8. This parameter enables the algorithm to explore potential optimal solution regions with progressively increasing search intensity, guided by the current optimal solution, thereby improving the algorithm's convergence accuracy and local exploitation capability.

[0029] Soft encirclement phase: At this stage, the prey possesses strong escape energy and needs to be searched over a large area; therefore, a scaling factor is selected. F= 0.5 •( 1 -t / T) It exhibits a linearly decreasing characteristic; the crossover probability is set to CR=0.6. This parameter selection can generate a large variable length in the early stages of iteration, increasing the global exploration capability. At the same time, as the number of iterations increases, the search range will gradually shrink, achieving the goal of balancing the search and exploration capabilities of the algorithm.

[0030] The optimization process of the improved IHHO algorithm is as follows: 1. Initialization By pre-setting the maximum number of algorithm iterations T, the population size N, and the value range of each decision variable, the positions of all Harris Eagle individuals are randomly initialized within the search space, and each individual position corresponds to a set of decision variable combinations.

[0031] 2. Calculate the fitness value and determine the global optimum. The fitness value of each Harris Eagle individual is calculated according to the objective function formula, where the objective function is the weighted sum of power deviation and voltage deviation; the position of the individual with the smallest fitness value in the current population is recorded as the global optimal solution.

[0032] 3. Update prey escape energy The escape energy E at the current iteration number is calculated using the improved escape energy formula. 4. Determine the absolute value of the escape energy and execute the corresponding strategy. When |E|≥1, the algorithm enters the global exploration phase, using a random position strategy to update the individual Harris Eagle's position; when |E|<1, the algorithm enters the local development phase, selecting a soft encirclement strategy, a hard encirclement strategy, a progressive rapid dive soft encirclement strategy, or a progressive rapid dive hard encirclement strategy to update the individual's position based on the comparison between the random number r and the absolute value of the escape energy |E|.

[0033] 5. Update the global optimal solution Calculate the updated fitness values ​​of all individuals. If there is an individual whose fitness value is less than the current global optimum, then update the position of the global optimum.

[0034] 6. Determine the termination condition of the iteration Determine whether the current iteration number t has reached the maximum iteration number T. If not, return to continue iterating; if it has, output the global optimal solution, that is, the optimal droop coefficient Ki and the optimal reference voltage UNi of each energy storage unit.

[0035] The power deviation is calculated based on the difference between the actual output power and the expected output power of each energy storage unit; the voltage deviation is calculated based on the difference between the reference voltage and the actual output voltage of each energy storage unit.

[0036] The improved Harris Eagle optimization algorithm introduces an improved escape energy formula and a hierarchical differential evolution mutation strategy on the basis of the traditional Harris Eagle optimization algorithm. The improved escape energy formula achieves a dynamic balance between global exploration and local exploitation by introducing random perturbation and adaptive exponential adjustment factor. The hierarchical differential evolution mutation strategy adopts different scaling factors and crossover probability parameters in the hard encirclement stage and the soft encirclement stage, respectively.

[0037] The droop control formula is: ; in, U ref This is the reference voltage for the DC bus. U i and P i The first i The output voltage and output power of each energy storage unit; K i This is the droop coefficient.

[0038] The expression for the output power of the energy storage unit is: ; Combining this with the droop control formula, we get: ; in,U bus This is the DC bus voltage. I dci For the first i The output current of each energy storage unit R i For the first i Line impedance.

[0039] To account for differences in energy storage capacity during power allocation, this paper proposes a power allocation factor, the expression of which is: ; in, m i Representing the i Power allocation factor of each energy storage unit CB i Representing the i The energy storage capacity of each energy storage unit The average state of charge of all energy storage units. SOC i For the first i The state of charge of each energy storage unit; The expression for the acceleration factor is: ; in, Q i For the first i The acceleration factor of each energy storage unit is used to replace the fixed coefficient and speed up the convergence. q , p To adjust the parameters, t This is an adjustment coefficient. When the SOC difference between the energy storage units is small, the convergence speed of SOC also slows down. Therefore, this paper introduces an acceleration factor. Q Instead of a fixed coefficient, this accelerates convergence. According to this power allocation parameter, when the State of Charge (SOC) is consistent, the right half of the value is the same, achieving the goal of allocating power proportionally to energy storage capacity. Simultaneously, based on the properties of the function, the allocation factor is limited to a certain range to avoid excessively large power allocation ratios when the SOC difference is large, which could lead to system instability. Each energy storage unit exchanges SOC and voltage values ​​via consistency communication, thereby obtaining the average state of charge and average voltage.

[0040] The desired output power of the energy storage unit is set according to the allocation factor as follows: ; in, For the first i The expected output power of each energy storage unit; Therefore, the global power deviation formula is defined as follows: ; in, F 1 For global power deviation, P ave The average output power of all energy storage units; The expression for average output power is: ; The global voltage deviation formula is: ; in, F 2 For global voltage deviation, U Ni This refers to the reference voltage values ​​of each energy storage unit output in real time using the IHHO algorithm.

[0041] global power deviation F 1 and global voltage deviation F 2 The comprehensive optimization objective function is defined as follows: ; Where F is the comprehensive optimization objective function, and c is the weighting coefficient of the power deviation function, which affects the SOC equalization speed of each energy storage unit. The preferred embodiments have been shown and described, but should not be construed as limiting the invention itself. Various changes in form and detail may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. A method for equalizing energy storage SOC based on the IHHO algorithm, characterized in that: Includes the following steps: S1. Collect the state of charge and output voltage of each energy storage unit, exchange the state of charge and voltage values ​​through consistent communication, and calculate the average state of charge and average voltage of all energy storage units. S2. Introduce a power allocation factor and acceleration factor that take into account capacity differences, and calculate the expected output power of each energy storage unit based on the deviation between the current state of charge and the average state of charge of each energy storage unit. S3. Construct the objective function using the weighted sum of power deviation and bus voltage deviation; S4. The improved Harris Eagle optimization algorithm is used to optimize the objective function in real time, and the optimal droop coefficient and the optimal reference voltage are output in real time. S5. Substitute the real-time output of the optimal droop coefficient and the optimal reference voltage into the droop control formula to generate control signals for each energy storage unit and start the next round of calculation. The method is applied to scenarios where energy storage units of different capacities operate in parallel in an isolated DC microgrid.

2. The energy storage SOC equalization control method based on the IHHO algorithm according to claim 1, characterized in that: The power deviation is calculated based on the difference between the actual output power and the expected output power of each energy storage unit; the voltage deviation is calculated based on the difference between the reference voltage and the actual output voltage of each energy storage unit.

3. The energy storage SOC equalization control method based on the IHHO algorithm according to claim 2, characterized in that: The improved Harris Eagle optimization algorithm introduces an improved escape energy formula and a hierarchical differential evolution mutation strategy on the basis of the traditional Harris Eagle optimization algorithm. The improved escape energy formula achieves a dynamic balance between global exploration and local exploitation by introducing random perturbation and adaptive exponential adjustment factor. The hierarchical differential evolution mutation strategy employs different scaling factors and crossover probability parameters in the hard-enclosed and soft-enclosed stages, respectively.

4. The energy storage SOC equalization control method based on the IHHO algorithm according to claim 3, characterized in that: The droop control formula is as follows: ； in, U ref This is the reference voltage for the DC bus. U i and P i The first i The output voltage and output power of each energy storage unit; K i This is the droop coefficient.

5. The energy storage SOC equalization control method based on the IHHO algorithm according to claim 4, characterized in that: The expression for the output power of the energy storage unit is: ； Combining this with the droop control formula, we get: ； in, U bus This is the DC bus voltage. I dci For the first i The output current of each energy storage unit R i For the first i Line impedance.

6. The energy storage SOC equalization control method based on the IHHO algorithm according to claim 5, characterized in that: The expression for the power allocation factor is: ； in, m i Representing the i Power allocation factor of each energy storage unit CB i Representing the i The energy storage capacity of each energy storage unit The average state of charge of all energy storage units. SOC i For the first i The state of charge of each energy storage unit; The expression for the acceleration factor is: ； in, Q i For the first i The acceleration factor of each energy storage unit is used to replace the fixed coefficient and speed up the convergence. q , p To adjust the parameters, t This is the adjustment coefficient.

7. The energy storage SOC equalization control method based on the IHHO algorithm according to claim 6, characterized in that: The expected output power of the energy storage unit is: ； in, For the first i The expected output power of each energy storage unit; The global power deviation formula is defined as follows: ； in, F 1 For global power deviation, P ave This represents the average output power of all energy storage units. The expression for average output power is: ； The global voltage deviation formula is: ； in, F 2 For global voltage deviation, U Ni This refers to the reference voltage values ​​of each energy storage unit output in real time using the IHHO algorithm.

8. The energy storage SOC equalization control method based on the IHHO algorithm according to claim 7, characterized in that: global power deviation F 1 and global voltage deviation F 2 The comprehensive optimization objective function is defined as follows: ； in, F To comprehensively optimize the objective function, c The weighting coefficients of the power deviation function affect the SOC equalization speed of each energy storage unit.

Images