Self-adaptive droop control method considering SOC balance and voltage deviation

Abstract

The invention discloses a self-adaptive droop control method giving consideration to SOC balance and voltage deviation, and the method comprises the steps: constructing an island DC micro-grid system, considering the power distribution principle of SOC balance and energy storage capacity, and setting a distribution factor associated with the SOC and the energy storage capacity for each energy storage device in the system; determining a power deviation function and a voltage deviation function of the island DC micro-grid system associated with the SOC, and obtaining a comprehensive optimization objective function after weighting; the SOC of each energy storage unit, the voltage reference value of the output end of the DC / DC converter and the line current are mutually communicated and exchanged through a consistency algorithm; and optimizing and solving the comprehensive optimization objective function by using an SHADE algorithm to obtain a droop coefficient and a voltage reference value of each energy storage unit converter. According to the self-adaptive droop control method, the droop control parameters can be dynamically adjusted, the problem that a traditional fixed parameter droop control strategy cannot adapt to actual operation conditions is solved, and meanwhile the system stability is improved and the power distribution is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of island DC microgrid control, and in particular to an adaptive droop control method that takes into account both SOC balance and voltage deviation. Background Art

[0002] DC microgrids are widely used in modern power systems, especially in smart grids and distributed energy systems. Power transmission within microgrids is efficient and flexible, enabling effective distributed energy dispatch. The development of DC microgrids is conducive to fully utilizing distributed energy resources.

[0003] SOC balancing is a technology that uses active or passive control to align the state of charge (SOC) of each battery cell in a battery pack, thereby improving the overall performance and lifespan of the battery pack. Its core goal is to address capacity and energy mismatches caused by manufacturing variations, usage environment, or aging.

[0004] Microgrids operate in two modes: grid-connected and islanded. In islanded mode, reducing bus voltage deviation and improving power distribution accuracy are key objectives of microgrid control strategies. Droop control is a commonly used control strategy in microgrids. Traditional droop control cannot achieve high-precision power distribution due to the influence of line impedance. Furthermore, the virtual impedance introduced by the droop coefficient results in a voltage drop, causing the terminal voltage to be lower than the output voltage. Research at home and abroad has improved on traditional droop control, proposing various adaptive droop control strategies, such as those using an adaptive PI controller, introducing the bus voltage change rate, and using fuzzy control. Summary of the Invention

[0005] 1. Technical problems to be solved:

[0006] In response to the above technical problems, the present invention provides an adaptive droop control method that takes into account both SOC balance and voltage deviation. This method takes into account the influence of line impedance and can optimize both power distribution and voltage deviation under the consistent target of the state of charge (SOC) of the energy storage device without adding a secondary voltage compensation link, thereby realizing a new SOC adaptive regulation control strategy.

[0007] 2. Technical solution:

[0008] An adaptive droop control method taking into account both SOC balance and voltage deviation is characterized by comprising:

[0009] Step 1: Construct an islanded DC microgrid system powered by energy storage and featuring voltage-power droop control and SOC balancing, including lines, loads, DC / DC converters, energy storage devices, and energy storage devices.

[0010] Step 2: Considering the power allocation principle of SOC balance and energy storage capacity, set the energy storage capacity allocation factor for each energy storage device in the system;

[0011] Step 3: Determine the power deviation function and voltage deviation function of the island DC microgrid system associated with the SOC, and obtain the comprehensive optimization objective function after weighting;

[0012] Step 4: Through the consistency algorithm, the SOC, DC / DC converter output voltage reference value, and line current of each energy storage unit are communicated and exchanged, and transmitted to the central controller of the system; the SHADE algorithm is used to optimize and solve the comprehensive optimization objective function to obtain the droop coefficient and voltage reference value of each energy storage unit converter.

[0013] Furthermore, in step 1, the expression for the voltage-power droop control of the SOC balance is:

[0014] U i =U ref -K i *P i (1)

[0015] In the above formula, U ref K is the rated reference value of the output voltage of the energy storage unit converter; i is the droop coefficient of the i-th energy storage unit; P i is the power delivered to the public load by the i-th energy storage unit;

[0016] Among them, the power P delivered to the public load by the i-th energy storage unit is i , as follows:

[0017]

[0018] In the above formula, U pcc Indicates the actual current of the circuit where the energy storage unit is located; I dci Indicates the line current of the line where the energy storage unit is located; U i is the voltage reference value of the DC / DC converter output corresponding to the energy storage unit; R i is the resistance of the circuit;

[0019] Substituting (1) into (2) we get:

[0020]

[0021] And because U ref The rated reference value of the output voltage of the energy storage unit converter is not the actual output reference voltage U Ni ,

[0022] Therefore, the actual output power of the energy storage unit is:

[0023]

[0024] Furthermore, step 2 specifically includes:

[0025] The relationship between the output power provided by any two energy storage devices in the system to the common load becomes:

[0026]

[0027] In the above formula, P i , P j are the powers delivered to the common load by the i-th and j-th energy storage units respectively; when the droop coefficient is large enough, the power ratio is the ratio of the reciprocal of the droop coefficient.

[0028] Associate the SOC of each energy storage device with the power allocation and define the allocation factor m of the i-th energy storage device i , as follows:

[0029]

[0030] In the above formula, CB i is the capacity of the i-th energy storage unit; x is the preset energy storage convergence coefficient, which is related to the SOC balancing speed of the energy storage device and affects the power and voltage variation range. It should be selected according to actual needs. Represents the average state of charge of the system, that is, the state of charge of all energy storage devices is added and the average is taken; when the SOC of the system is balanced, the right half of the allocation factor of each energy storage device is Some of the values are the same, thus achieving power distribution according to the capacity ratio of each energy storage device in the system.

[0031] Furthermore, in step 3, determining the comprehensive optimization objective function of the SOC balancing model of the islanded DC microgrid system specifically includes the following steps:

[0032] S31: Determine the target power of all energy storage devices in the system:

[0033] P ave =(m1P1+m2P2+m3P3+L+m n P n ) / n (7)

[0034] In the above formula, P ave is the target power of all energy storage devices, and the subscript n represents the nth energy storage device

[0035] S32: The line impedance R where the i-th energy storage device is located i It is calculated by the droop coefficient at the previous moment, as follows:

[0036]

[0037] In the above formula, k' is the droop coefficient of the line at a certain moment; U pcc ' is the bus voltage of the line at the previous moment; P i 'The power delivered by the i-th energy storage unit of the line at the last moment;

[0038] S33: Power deviation ΔP is performed on each energy storage device i Calculate, where the power deviation function is:

[0039]

[0040] S34: Calculate the total power deviation P of the entire system error Deviation from total voltage U error , where the total power deviation function and the total voltage deviation function are as follows:

[0041]

[0042] S35: Obtain comprehensive optimization objective function:

[0043] F(k n ,U Nn )=cP error +(1-c)U error (11)

[0044] In the above formula, F(k n ,U Nn ) represents the power deviation function P error And the voltage deviation function U error The comprehensive optimization objective function obtained after weighting can achieve the adjustment of power deviation and voltage deviation;

[0045] In order to achieve the SOC balance of the system, the weighting coefficient c is set as large as possible while ensuring that the voltage deviation is reasonable, and the value is 0.985.

[0046] Furthermore, in step 4, the SHADE algorithm is used to optimize and solve the comprehensive optimization objective function, which specifically includes the following steps:

[0047] S41: Initializing parameters and population to obtain a first generation population, where each individual in the population serves as a candidate solution of the robust nonlinear optimization model;

[0048] S42: setting a mutation operation strategy, which generates a mutation vector according to the difference between the current individual and the best individual;

[0049] S43: Mix the mutation vector and the target vector through the crossover operation formula;

[0050] S44: A classic greedy selection strategy is used to decide whether an individual enters the next generation. That is, if the fitness value of the trial vector is better than the fitness value of the target vector, the target vector is replaced by the trial vector, otherwise the target vector is retained.

[0051] S45: Use historical success memory strategy to enhance the efficiency of parameter selection;

[0052] S46: The maximum number of iterations is preset. After each generation completes the fitness evaluation, it will check whether the convergence conditions are met. If so, the algorithm terminates the calculation and outputs the optimal solution. The optimal solution is output as the droop coefficient and reference voltage value of each energy storage unit, which are applied to the circuit as new control parameters, and the next sampling and calculation are performed. If not, S41-S45 are repeated.

[0053] 3.Beneficial effects:

[0054] (1) The present invention provides an adaptive droop control method that takes into account both SOC balance and voltage deviation, which can improve system stability. Without adding a secondary voltage compensation link, the droop coefficient and voltage reference value are dynamically adjusted according to the SHADE algorithm to adapt to load fluctuations.

[0055] (2) The present invention provides an adaptive droop control method that takes into account both SOC balance and voltage deviation. Unlike the common method of directly defining the droop coefficient, it can adapt to and optimize the power distribution of each energy unit in the microgrid by designing a new distribution factor in the form of a power function. When considering the line impedance, it can achieve a uniform distribution according to the capacity ratio when the SOC is consistent, improve the distribution accuracy, extend the service life of the energy storage equipment, and optimize the system power distribution.

[0056] (3) The present invention provides an adaptive droop control method that takes into account both SOC balance and voltage deviation. There is no need to design an initial droop coefficient and voltage reference value. The system can automatically adjust the control strategy according to the SOC state and load changes. It operates well under constant power load and resistive load, adapts to microgrid environments of different sizes and types, and enhances system adaptability.

[0057] (4) The present invention provides an adaptive droop control method that takes into account both SOC balance and voltage deviation, thereby avoiding energy waste caused by voltage instability or uneven power distribution, improving the overall operating efficiency of the system, and enhancing the economy of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 This is the traditional voltage-power droop control involved in this method;

[0059] Figure 2 The framework diagram of the SOC balancing method capable of realizing the adaptive control strategy of this method;

[0060] Figure 3 This is the flow chart of the method using the SHADE algorithm to find the optimal solution;

[0061] Figure 4 This is the SOC convergence consistency process diagram obtained in the verification example;

[0062] Figure 5 The common bus power diagram for each energy storage unit obtained in the verification example;

[0063] Figure 6 This is the ratio of the droop coefficient k2 to k1 obtained in the verification example;

[0064] Figure 7 This is the ratio of the droop coefficient k3 to k1 obtained in the verification example;

[0065] Figure 8 To verify the example, the reference voltage fluctuation diagram of the output end of the energy storage unit converter 1 is shown;

[0066] Figure 9 This is the voltage fluctuation diagram at the bus terminal in the verification example. DETAILED DESCRIPTION

[0067] The present invention will be described in detail below with reference to the accompanying drawings.

[0068] As attached Figure 1 To the attached Figure 3 As shown, an adaptive droop control method taking into account both SOC balance and voltage deviation is characterized by comprising:

[0069] Step 1: Construct an islanded DC microgrid system powered by energy storage and featuring voltage-power droop control and SOC balancing, including lines, loads, DC / DC converters, energy storage devices, and energy storage devices.

[0070] Step 2: Considering the power allocation principle of SOC balance and energy storage capacity, set the energy storage capacity allocation factor for each energy storage device in the system;

[0071] Step 3: Determine the power deviation function and voltage deviation function of the island DC microgrid system associated with the SOC, and obtain the comprehensive optimization objective function after weighting;

[0072] Step 4: Through the consistency algorithm, the SOC, DC / DC converter output voltage reference value, and line current of each energy storage unit are communicated and exchanged, and transmitted to the central controller of the system; the SHADE algorithm is used to optimize and solve the comprehensive optimization objective function to obtain the droop coefficient and voltage reference value of each energy storage unit converter.

[0073] Furthermore, in step 1, the expression for the voltage-power droop control of the SOC balance is:

[0074] U i =U ref -K i *P i (1)

[0075] In the above formula, U ref K is the rated reference value of the output voltage of the energy storage unit converter; i is the droop coefficient of the i-th energy storage unit; P i is the power delivered to the public load by the i-th energy storage unit;

[0076] Among them, the power P delivered to the public load by the i-th energy storage unit is i , as follows:

[0077]

[0078] In the above formula, U pcc Indicates the actual current of the circuit where the energy storage unit is located; I dci Indicates the line current of the line where the energy storage unit is located; U i is the voltage reference value of the DC / DC converter output corresponding to the energy storage unit; R i is the resistance of the circuit;

[0079] Substituting (1) into (2) we get:

[0080]

[0081] And because U ref The rated reference value of the output voltage of the energy storage unit converter is not the actual output reference voltage U Ni ,

[0082] Therefore, the actual output power of the energy storage unit is:

[0083]

[0084] Furthermore, step 2 specifically includes:

[0085] The relationship between the output power provided by any two energy storage devices in the system to the common load becomes:

[0086]

[0087] In the above formula, P i , P jare the powers delivered to the common load by the i-th and j-th energy storage units respectively; when the droop coefficient is large enough, the power ratio is the ratio of the reciprocal of the droop coefficient.

[0088] Associate the SOC of each energy storage device with the power allocation and define the allocation factor m of the i-th energy storage device i , as follows:

[0089]

[0090] In the above formula, CB i is the capacity of the i-th energy storage unit; x is the preset energy storage convergence coefficient, which is related to the SOC balancing speed of the energy storage device and affects the power and voltage variation range. It should be selected according to actual needs. Represents the average state of charge of the system, that is, the state of charge of all energy storage devices is added and the average is taken; when the SOC of the system is balanced, the right half of the allocation factor of each energy storage device is Some of the values are the same, thus achieving power distribution according to the capacity ratio of each energy storage device in the system.

[0091] Furthermore, in step 3, determining the comprehensive optimization objective function of the SOC balancing model of the islanded DC microgrid system specifically includes the following steps:

[0092] S31: Determine the target power of all energy storage devices in the system:

[0093] P ave =(m1P1+m2P2+m3P3+L+m n P n ) / n (7)

[0094] In the above formula, P ave is the target power of all energy storage devices, and the subscript n represents the nth energy storage device

[0095] S32: The line impedance R where the i-th energy storage device is located i It is calculated by the droop coefficient at the previous moment, as follows:

[0096]

[0097] In the above formula, k' is the droop coefficient of the line at a certain moment; U pcc ' is the bus voltage of the line at the previous moment; P i 'The power delivered by the i-th energy storage unit of the line at the last moment;

[0098] S33: Power deviation ΔP is performed on each energy storage device i Calculate, where the power deviation function is:

[0099]

[0100] S34: Calculate the total power deviation P of the entire system error Deviation from total voltage U error , where the total power deviation function and the total voltage deviation function are as follows:

[0101]

[0102] S35: Obtain comprehensive optimization objective function:

[0103] F(k n ,U Nn )=cP error +(1-c)U err or (11)

[0104] In the above formula, F(k n ,U Nn ) represents the power deviation function P error And the voltage deviation function U error The comprehensive optimization objective function obtained after weighting can achieve the adjustment of power deviation and voltage deviation;

[0105] In order to achieve the SOC balance of the system, the weighting coefficient c is set as large as possible while ensuring that the voltage deviation is reasonable, and the value is 0.985.

[0106] Furthermore, in step 4, the SHADE algorithm is used to optimize and solve the comprehensive optimization objective function, which specifically includes the following steps:

[0107] S41: Initializing parameters and population to obtain a first generation population, where each individual in the population serves as a candidate solution of the robust nonlinear optimization model;

[0108] S42: setting a mutation operation strategy, which generates a mutation vector according to the difference between the current individual and the best individual;

[0109] S43: Mix the mutation vector and the target vector through the crossover operation formula;

[0110] S44: A classic greedy selection strategy is used to decide whether an individual enters the next generation. That is, if the fitness value of the trial vector is better than the fitness value of the target vector, the target vector is replaced by the trial vector, otherwise the target vector is retained.

[0111] S45: Use historical success memory strategy to enhance the efficiency of parameter selection;

[0112] S46: The maximum number of iterations is preset. After each generation completes the fitness evaluation, it will check whether the convergence conditions are met. If so, the algorithm terminates the calculation and outputs the optimal solution. The optimal solution is output as the droop coefficient and reference voltage value of each energy storage unit, which are applied to the circuit as new control parameters, and the next sampling and calculation are performed. If not, S41-S45 are repeated.

[0113] Verification example:

[0114] In this verification example, each energy storage unit obtains the state of charge SOC1, SOC2, SOC3...SOC through sensor sampling. n The voltages of the converter output terminals U_bus1, U_bus2, U_bus3L and U_busn are sent to the central controller based on the consistency communication of adjacent data, and the average value is taken to obtain the average state of charge. and bus voltage U pcc .

[0115] In this embodiment, the effectiveness of this method is verified by the MATLAB / SIMULINK platform. Considering the line impedance and energy storage capacity, three sets of parallel energy storage units are taken as an example to study the SOC balance and voltage deviation optimization of energy storage equipment in the implementation of an island DC microgrid. A simulation model is built. Figure 2 As shown in the figure, the line impedances of the three energy storage units are different, and the energy storage capacity ratio of each energy storage unit is 5:5:3; the initial SOC values are 75%, 73%, and 70% respectively; and the common load is a constant power load. Figure 5 and 6 As shown in the figure, when this control method is adopted, when SOC1 is greater than SOC2, the droop coefficient k2 gradually decreases from 7 times of k1, and the actual transmission power of energy storage unit 1 also decreases from 7 times of the transmission power of energy storage unit 1. In this process, the deviation between SOC1 and SOC2 gradually decreases. When SOC1 is equal to SOC2, k2 is close to k1. Similarly, as shown in the figure below, Figure 7 As shown, as the energy storage unit 3 with the smallest SOC1, in order to achieve SOC balance, the transmission power should be the smallest, that is, the droop coefficient should be the largest. The actual droop coefficient k3 begins to decay at seventy times of k1, which means that the transmission power of energy storage unit 1 is only one-seventieth of that of energy storage unit 3 at the beginning, and one-tenth of the transmission power of energy storage unit 2. Finally, the droop coefficient k3 decays to nearly three-fifths of the droop coefficient k1, achieving the actual transmission power of three-fifths of that of energy storage unit 1, achieving the purpose of distribution according to capacity ratio after balancing. Figure 4 As shown, SOC finally maintains equilibrium. Figure 8 As shown, taking the reference voltage of the output end of the energy storage unit converter 1 as an example, the standard reference voltage of 750V is gradually increased to 760V. Figure 9As shown in the figure, without adding a voltage secondary compensation link, the droop parameters selected by the algorithm can also keep the bus voltage above 740V, meeting the standard of voltage deviation less than 5%.

[0116] The power deviation function and voltage deviation function constructed in this implementation are shown in the following formula:

[0117]

[0118] SHADE (Self-adaptive Historical Differential Evolution) in this method is an adaptive differential evolution (DE) algorithm that enhances search efficiency by adaptively adjusting the algorithm's control parameters, such as the scaling factor F and the crossover rate Cr. SHADE uses a historical success memory strategy to record previously successful control parameters for use in subsequent iterations, thereby accelerating convergence and improving the global performance of the search. Its optimization process includes the following key steps:

[0119] 1. Initialization

[0120] The SHADE algorithm first generates an initial population P0, and each individual v in the population i It is initialized by a D-dimensional random number, that is, the position of each individual in the search space is determined by a random number.

[0121]

[0122] in, is the position of the i-th individual in generation 0 (initialization generation). upper and x lower are the upper and lower bounds of the search space respectively. rand(0,1) is a randomly generated uniformly distributed value used to generate the initial solution.

[0123] 2. Mutation Operation

[0124] The mutation operation is used to enhance the diversity of the population by generating mutation vectors. SHADE uses the current-to-pbest / 1 mutation strategy, which generates mutation vectors based on the difference between the current individual and the best individual:

[0125]

[0126] in, For the current individual, is the best individual in the current population, and are other individuals randomly selected from the population, and F is the scaling factor (which controls the step size of the mutation).

[0127] 3. Crossover Operation

[0128] The crossover operation generates a trial vector by mixing the mutation vector with the target vector. The crossover operation is performed using the following formula:

[0129]

[0130] Where Cr is the crossover rate, which determines how many elements in the trial vector come from the mutation vector.

[0131] 4. Select an action

[0132] The selection operation determines whether an individual enters the next generation based on its fitness value. SHADE uses a classic greedy selection strategy, that is, if the fitness value of the trial vector is better than the fitness value of the target vector, the trial vector replaces the target vector; otherwise, the target vector is retained. The selection operation formula is as follows:

[0133]

[0134] Among them, f(v' i ) is the fitness value of the individual, is the fitness value of the trial vector.

[0135] 5. Historically successful memory strategies

[0136] The SHADE algorithm uses a historical success memory strategy to enhance the efficiency of parameter selection. The historical memory records past successful parameter pairs and uses these historical parameters to guide the current search process. Specifically, the algorithm updates the historical success memory based on the fitness value, storing the control parameters that performed well:

[0137] M Cr ,M F =update(M Cr ,M F ,v' i )

[0138] Among them, M Cr and M F are the crossover rate and scaling factor stored in the history memory.

[0139] 6. Convergence Check

[0140] After completing fitness evaluation at each iteration, the algorithm checks whether convergence conditions have been met. Typically, convergence conditions are reached after reaching a specified maximum number of iterations or when fitness values no longer significantly improve. The algorithm terminates the calculation and outputs the optimal solution. The optimal solution is output as the droop coefficient and reference voltage value for each energy storage unit. These are then applied to the circuit as new control parameters for the next sampling and calculation.

[0141] Although the present invention has been disclosed above in terms of preferred embodiments, they are not intended to limit the present invention. Anyone skilled in the art can make various changes or modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection defined by the claims of this application.

Claims

1. An adaptive droop control method that takes into account both SOC balance and voltage deviation, characterized by: include: Step 1: Construct an islanded DC microgrid system powered by energy storage and featuring voltage-power droop control and SOC balancing, including lines, loads, DC / DC converters, energy storage devices, and energy storage devices. Step 2: Considering the power allocation principle of SOC balance and energy storage capacity, set the energy storage capacity allocation factor for each energy storage device in the system; Step 3: Determine the power deviation function and voltage deviation function of the island DC microgrid system associated with the SOC, and obtain the comprehensive optimization objective function after weighting; Step 4: Through the consistency algorithm, the SOC, DC / DC converter output voltage reference value, and line current of each energy storage unit are communicated and exchanged, and transmitted to the central controller of the system; the SHADE algorithm is used to optimize and solve the comprehensive optimization objective function to obtain the droop coefficient and voltage reference value of each energy storage unit converter.

2. The adaptive droop control method for balancing SOC balance and voltage deviation according to claim 1, characterized in that: In step 1, the expression of the voltage-power droop control of the SOC balance is: IN i =U ref -K i *P i (1) In the above formula, U ref K is the rated reference value of the output voltage of the energy storage unit converter; i is the droop coefficient of the i-th energy storage unit; P i is the power delivered to the public load by the i-th energy storage unit; Among them, the power P delivered to the public load by the i-th energy storage unit is i , as follows: In the above formula, U pcc Indicates the actual current of the circuit where the energy storage unit is located; I dci Indicates the line current of the line where the energy storage unit is located; U i is the voltage reference value of the DC / DC converter output corresponding to the energy storage unit; R i is the resistance of the circuit; Substituting (1) into (2) we get: And because U ref The rated reference value of the output voltage of the energy storage unit converter is not the actual output reference voltage U Ni , So the actual output power of the energy storage unit 3. The adaptive droop control method for balancing SOC balance and voltage deviation according to claim 2, characterized in that: Step 2 specifically includes: The relationship between the output power provided by any two energy storage devices in the system to the common load becomes: In the above formula, P i , P j are the powers delivered to the common load by the i-th and j-th energy storage units, respectively. When the droop coefficient is large enough, the power ratio is the ratio of the inverse of the droop coefficient. Associate the SOC of each energy storage device with the power allocation and define the allocation factor m of the i-th energy storage device i , as follows: In the above formula, CB i is the capacity of the i-th energy storage unit; x is the preset energy storage convergence coefficient, which is related to the SOC balancing speed of the energy storage device and affects the power and voltage variation range. It should be selected according to actual needs. Represents the average state of charge of the system, that is, the state of charge of all energy storage devices is added and the average is taken; when the SOC of the system is balanced, the right half of the allocation factor of each energy storage device is Some of the values are the same, thus achieving power distribution according to the capacity ratio of each energy storage device in the system.

4. The adaptive droop control method for balancing SOC balance and voltage deviation according to claim 3, characterized in that: In step 3, determining the comprehensive optimization objective function of the SOC balancing model of the islanded DC microgrid system specifically includes the following steps: S31: Determine the target power of all energy storage devices in the system: P ave =(m1P1+m2P2+m3P3+L+m n P n ) / n (7) In the above formula, P ave is the target power of all energy storage devices, and the subscript n represents the nth energy storage device S32: The line impedance R where the i-th energy storage device is located i It is calculated by the droop coefficient at the previous moment, as follows: In the above formula, k' is the droop coefficient of the line at a certain moment; U pcc ' is the bus voltage of the line at the previous moment; P' i The power delivered by the i-th energy storage unit of the line at the last moment; S33: Power deviation ΔP is performed on each energy storage device i Calculate, where the power deviation function is: S34: Calculate the total power deviation P of the entire system error Deviation from total voltage U error , where the total power deviation function and the total voltage deviation function are as follows: S35: Obtain comprehensive optimization objective function: F(k n ,U Nn )=cP error +(1-c)U err or (11) In the above formula, F(k n ,U Nn ) represents the power deviation function P error And the voltage deviation function U error The comprehensive optimization objective function obtained after weighting can achieve the adjustment of power deviation and voltage deviation; In order to achieve the SOC balance of the system, the weighting coefficient c is set as large as possible while ensuring that the voltage deviation is reasonable, and the value is 0.

985.

5. The adaptive droop control method taking into account both SOC balance and voltage deviation according to claim 1, characterized in that: In step 4, the SHADE algorithm is used to find the optimal solution for the comprehensive optimization objective function, which specifically includes the following steps: S41: Initializing parameters and population to obtain a first generation population, where each individual in the population serves as a candidate solution of the robust nonlinear optimization model; S42: setting a mutation operation strategy, which generates a mutation vector according to the difference between the current individual and the best individual; S43: Mix the mutation vector and the target vector through the crossover operation formula; S44: A classic greedy selection strategy is used to decide whether an individual enters the next generation. That is, if the fitness value of the trial vector is better than the fitness value of the target vector, the target vector is replaced by the trial vector, otherwise the target vector is retained. S45: Use historical success memory strategy to enhance the efficiency of parameter selection; S46: The maximum number of iterations is preset. After each generation completes the fitness evaluation, it will check whether the convergence conditions are met. If so, the algorithm terminates the calculation and outputs the optimal solution. The optimal solution is output as the droop coefficient and reference voltage value of each energy storage unit, which are applied to the circuit as new control parameters, and the next sampling and calculation are performed. If not, S41-S45 are repeated.

Images