Control strategy and energy storage capacity configuration method for multi-source combined frequency modulation

Abstract

The application relates to a multi-source combined frequency modulation control strategy and energy storage capacity configuration method, and belongs to the technical field of power systems.The application aims to solve the energy storage capacity optimization mathematical model established under the combined system by using a multi-objective particle swarm optimization algorithm, and finally compares and analyzes the multi-source combined frequency modulation control strategy and energy storage capacity configuration method under two experimental conditions.The application introduces a multi-source combined frequency modulation system, establishes the control strategy of the multi-source combined frequency modulation system in a three-machine nine-node system of PSCAD / EMTDC according to the primary frequency modulation principles and control methods of different types of energy, proposes an energy storage capacity optimization model, introduces an MOPSO algorithm, analyzes the energy storage capacity optimization result through an example, and analyzes the energy storage capacity configuration.The application can reasonably configure the energy storage capacity of the multi-source combined system, considers the system frequency stability, and improves the overall economy of the system.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power systems. BACKGROUND

[0002] In order to improve the utilization rate of new energy, new energy represented by wind power and photovoltaic has been widely used, but it has also brought great challenges to the frequency regulation of power systems. Over the past few decades, with the rapid development of society and the significant improvement of people's living quality, problems such as overconsumption of fossil energy and environmental degradation have arisen. Renewable energy represented by wind power and photovoltaic has received high attention from all walks of life and has been widely used in many fields due to its many advantages such as cleanliness, high efficiency and abundant resources.

[0003] From the existing domestic and foreign research results, it can be seen that the research on new energy participating in power grid frequency regulation mainly focuses on single machine frequency regulation, frequency regulation based on virtual synchronous generator and participation in microgrid island frequency regulation. The advantages of thermal power units and wind turbine units are combined to greatly improve the overall frequency fluctuation of the system. The rotor kinetic energy of the wind turbine and the super capacitor are coordinated to ensure that the doubly-fed wind turbine can provide fast and continuous active power support when the system frequency fluctuates. A frequency regulation control method for a microgrid containing light, diesel and storage is proposed to support the dynamic balance of active power in the entire microgrid. A three-level coordinated frequency control method based on discrete consensus effectively improves the lowest point of frequency when the frequency drops. For the differences between wind and light, a segmented method is used to cooperate with the conventional unit in the time domain and the frequency domain respectively. A method using a low-pass filter is proposed to fully utilize the frequency regulation advantage of energy storage. By comparing the frequency regulation capacity of conventional units, wind turbines and energy storage, a wind-storage-fire combined frequency regulation method based on frequency division regulation principle is proposed. According to the analysis of the above literature, the size of the energy storage capacity not only affects the overall frequency regulation capability of the system, but also relates to the overall economy and practicality. SUMMARY

[0004] The purpose of the application is to use a multi-objective particle swarm algorithm to solve the energy storage capacity optimization mathematical model established under the joint system, and finally to compare and analyze the control strategy of multi-source joint frequency regulation and the energy storage capacity configuration method under two experimental conditions.

[0005] The steps of the application are:

[0006] S1, wind turbine frequency regulation strategy

[0007] Active power change ΔP eg Calculated from the rate of change of system frequency deviation

[0008]

[0009] wherein K j represents the virtual inertia coefficient; Δf2 represents the frequency deviation amount borne by the wind turbine generator;

[0010] active power variation amount ΔP ex is calculated from the size of the system frequency deviation value

[0011] ΔP ex = K d Δf2 (2)

[0012] wherein K d represents the droop coefficient;

[0013] When the power system frequency fluctuates, an additional frequency response unit is formed by the virtual inertia control and the droop control;

[0014] S2, photovoltaic generator frequency modulation strategy

[0015] The load shedding level σ% of the photovoltaic array is defined as

[0016]

[0017] S3, energy storage system frequency modulation strategy

[0018] In the multi-source joint system, the frequency modulation effect of the joint frequency modulation system is evaluated by formula (4) as follows:

[0019]

[0020] wherein Δf jlh represents the difference between the jth actual sampling frequency value f jlh of the joint frequency modulation system and the rated frequency f0; Δf jct represents the difference between the jth actual sampling frequency value f jct of the conventional unit and the rated frequency f0; n represents the sampling point number; f1 represents the frequency modulation effect of the joint frequency modulation system, f1 ∈ [0, 1], when f1 = 0, it indicates that the active power balance of the joint frequency modulation system, the frequency does not fluctuate, and the energy storage capacity can meet the frequency modulation demand of the system; when f1 = 1, it indicates that the energy storage capacity cannot meet the frequency modulation demand;

[0021] The initial investment cost C bess of the energy storage is represented by formula (5) as follows:

[0022] C bess = c b E bess (5)

[0023] wherein c b represents the cost of configuring a unit capacity of energy storage; E bessC represents the capacity of the energy storage device;

[0024] The operation cost C of the energy storage device is represented by formula (6) yx :

[0025]

[0026] In the formula, c y represents the operation cost of the energy storage device depreciation; E b,t represents the amount of electricity provided by the energy storage device for system frequency regulation in the time period t; represents the amount of electricity output by the energy storage device for system frequency regulation in the time period t; represents the amount of electricity absorbed by the energy storage device for system frequency regulation in the time period t;

[0027] The abandoned light cost of photovoltaic is represented by formula (7) as follows:

[0028] C sun = c s P MPPT,t * t) (7)

[0029] In the formula, C sun represents the price of electricity sold by the photovoltaic field to the power grid; P MPPT,t represents the average value of the maximum power output by the photovoltaic in the maximum power tracking mode in the time period t;

[0030] The auxiliary frequency modulation service fee E f is represented by formula (8) as follows:

[0031] E f = c f E ∑ = c f (E W,t +E s,t +E b,t ) (8)

[0032] In the formula, c f represents the frequency modulation price of wind, light, and energy storage participating in primary frequency modulation service; E ∑ represents the total frequency modulation capacity; E W,t represents the frequency modulation electricity provided by the wind turbine in the time period t; E s,t represents the frequency modulation electricity provided by the photovoltaic in the time period t; E b,t represents the frequency modulation electricity provided by the energy storage in the time period t;

[0033] The annual net present value NPV of the combined system is represented by formula (9) as follows:

[0034]

[0035] In the formula, t1~t n represents the n data sampling time period; D represents the sampling time; Y represents the service life of the energy storage; and h represents the depreciation rate of the energy storage. represents the recovery coefficient of the energy storage.

[0036] The objective function for improving the frequency stability and annual net income of the multi-source joint system is shown in formula (10):

[0037]

[0038] The equality constraint condition of the model is shown in formula (11):

[0039] E dw,t =E G,t +E W,t +E s,t +E b,t (11)

[0040] In the formula, E G，t represents the power of the thermal power unit participating in frequency modulation in the t time period; E dw,t represents the power demand of the system frequency modulation in the t time period, that is, the power required by the entire network should be provided by wind, light, storage and fire.

[0041] SOC min = 20%, SOC max = 80%, that is, shown in formula (12):

[0042] SOC min ≤ SOC ≤ SOC max (12)

[0043] The basic flow of solving the energy storage capacity optimization configuration of the multi-source joint system is as follows: firstly, the frequency improvement effect and the overall economy of the multi-source joint system are taken as two optimization targets of the algorithm, and each energy storage capacity optimization scheme and the corresponding frequency deviation amount are taken as particles in the population, then initialization processing is performed, each particle is given a random initial position and speed, then the fitness is calculated, the particle fitness of the current position is compared with the fitness of the historical and global best position, and the best position of the particle is updated, then the position and speed of the particle are updated according to formula (11) and (12), whether the iteration number meets the maximum iteration number is judged, if not, the optimization is re-performed, if yes, the final result is output, and the running of the algorithm is ended.

[0044] The energy storage capacity of the multi-source joint system can be reasonably configured, the system frequency stability is considered, and the overall economy of the system is improved. BRIEF DESCRIPTION OF DRAWINGS

[0045] Figure 1 is a control strategy diagram of multi-source joint frequency modulation;

[0046] Figure 2 is a power frequency characteristic diagram of energy storage participating in primary frequency modulation;

[0047] Figure 3 is an MOPSO algorithm flowchart;

[0048] Figure 4 is a solution distribution condition diagram;

[0049] Figure 5 is a system frequency change curve diagram under four experimental conditions;

[0050] Figure 6 is a system frequency change curve diagram under four conditions;

[0051] Figure 7 is a virtual inertia control diagram;

[0052] Figure 8 is a virtual droop control diagram;

[0053] Figure 9 is a photovoltaic load shedding control operation principle diagram. DETAILED DESCRIPTION

[0054] The energy storage capacity of the multi-source combined system can be reasonably configured, the frequency stability is considered, and the economy of the whole system is maximized. The control strategy and storage capacity optimization model of the multi-source combined frequency control system are analyzed, the MOPSO algorithm is used to solve the energy storage capacity optimization configuration problem of the multi-source combined system. Finally, comparative analysis is carried out under two experimental conditions, and it is verified that the energy storage result meets the frequency modulation demand and also considers the economy of the whole system.

[0055] The energy storage capacity of the multi-source combined system is reasonably configured and researched, the net income of the whole system is maximized as much as possible under the condition of meeting the overall frequency modulation demand of the system, the frequency stability is considered, and the economy of the whole system is improved as much as possible. The energy storage capacity optimization target of the multi-source combined system is established.

[0056] The energy storage capacity optimization target of the multi-source combined system is established, the reasonable energy storage capacity is obtained by using the MOPSO algorithm, and finally experiments are carried out under two disturbance conditions, and the traditional thermal power, wind-storage-thermal system and 5MW energy storage capacity combined system are compared respectively, and the frequency change conditions of four scenes are verified. The setting of 3.62MW energy storage can not only improve the frequency stability of the system, but also improve the economy of the whole system.

[0057] The present application comprises the following steps:

[0058] (1) Introduce a multi-source joint frequency regulation system. Based on the primary frequency regulation principle and control method of different types of energy, build a control strategy for the multi-source joint frequency regulation system in the three-machine nine-node system of PSCAD / EMTDC.

[0059] (2) An energy storage capacity optimization model is proposed to evaluate the frequency regulation effect of the joint frequency regulation system in a multi-source joint system;

[0060] (3) Introduce the MOPSO algorithm and use the MOPSO algorithm to solve the problem of energy storage capacity optimization configuration of multi-source joint system;

[0061] (4) Analyze the results of energy storage capacity optimization and the configuration of energy storage capacity through numerical examples.

[0062] This invention addresses the energy storage capacity configuration problem in existing multi-source combined systems, aiming to maximize the overall net benefit of the system while meeting the overall frequency regulation requirements.

[0063] This invention establishes an optimization objective for the energy storage capacity of a multi-source combined system. The multi-objective particle swarm optimization algorithm is used to calculate the correctness and superiority of the optimal energy storage capacity configuration result.

[0064] This invention uses the MOPSO algorithm to calculate and analyze a reasonable energy storage capacity.

[0065] The present invention will now be described in detail:

[0066] Based on the primary frequency regulation principles and control methods of different types of energy sources, a system such as the PSCAD / EMTDC three-machine nine-node system was built. Figure 1 The control strategy of the multi-source joint frequency modulation system is shown.

[0067] Δf t This represents the real-time frequency deviation of the system; Δf1, Δf2, Δf3, and Δf4 represent the frequency regulation amounts undertaken by thermal power, wind power, photovoltaic power, and energy storage, respectively; P t P w P s P b These represent the active power outputs of thermal power, wind power, photovoltaic power, and energy storage after primary frequency regulation; ΔP wind Indicates the incremental frequency regulation control of wind power; P wref P represents the active power output during steady-state operation of wind power; w P represents the active power output of wind power after primary frequency regulation; bref P represents the active power output during steady-state operation of the energy storage system. b This indicates the active power output of the energy storage after one frequency regulation.

[0068] (1) Wind turbine frequency regulation strategy

[0069] As Figure 7 shown is a virtual inertia control principle diagram, a control mode designed by deriving from the basic principle of inertia response, active power variation ΔP eg is calculated by the rate of change of system frequency deviation, which is expressed by formula (1) as follows:

[0070]

[0071] In the formula, K j represents the virtual inertia coefficient; Δf2 represents the frequency deviation amount borne by the wind turbine.

[0072] Droop control is a control mode simulating the power-frequency droop characteristic of conventional units, and its principle is shown in Figure 8 The active power variation ΔP ex is calculated by the size of the system frequency deviation value, which is expressed by formula (2) as follows:

[0073] ΔP ex = K d Δf2 (2)

[0074] In the formula, K d represents the droop coefficient.

[0075] When the frequency of the power system fluctuates, an additional frequency response unit is formed by the virtual inertia control and the droop control, and the frequency of the entire power system can be adjusted by adjusting the active power output of the unit.

[0076] (2) Photovoltaic unit frequency modulation strategy

[0077] The present application adopts the disturbance observation method, under which the active power output of the photovoltaic is maximum, but the upward adjustment is limited when the frequency changes. If the photovoltaic is allowed to participate in primary frequency modulation, the actual operating voltage of the photovoltaic needs to be controlled to be slightly higher than the voltage at the maximum power point, and a certain reserve is reserved, and the photovoltaic load shedding operation principle is shown in Figure 9 .

[0078] In the figure, K p represents the unit adjustment power of the photovoltaic; ΔP sun represents the primary frequency modulation amount of the photovoltaic; P j represents the active power output of the photovoltaic in load shedding operation; P s represents the active power output of the photovoltaic after primary frequency modulation; and Δf3 represents the adjustment amount borne by the photovoltaic.

[0079] The load shedding level σ% of the photovoltaic array is defined as formula (3) as follows:

[0080]

[0081] (3) Frequency regulation strategy of energy storage system

[0082] The principle of energy storage participating in primary frequency regulation of power system is shown in Fig. 1. Figure 2 If the load suddenly increases, the load curve L1 will move to L2, and the BESS will quickly respond to increase the output to suppress the decrease of frequency. The system frequency will move from point a to point b, and the frequency deviation of the system is Δf e , the energy storage is in the linear output stage, and the battery outputs extra power ΔP b to keep the frequency at point c by setting the unit regulation power K bess . When the frequency exceeds the upper limit or lower limit of the frequency regulation dead zone, the energy storage battery is in the charging or discharging state.

[0083] wherein f u represents the upper limit of the frequency regulation dead zone; f d represents the lower limit of the frequency regulation dead zone; P BESS represents the upper limit of the battery output participating in frequency regulation; and -P BESS represents the lower limit of the battery output participating in frequency regulation.

[0084] In a multi-source joint system, the frequency regulation effect of the joint frequency regulation system is evaluated by formula (4):

[0085]

[0086] In the formula, Δf jlh represents the difference between the jth actual sampling frequency value f jlh of the joint frequency regulation system and the rated frequency f0; Δf jct represents the difference between the jth actual sampling frequency value f jct of the traditional unit and the rated frequency f0; n represents the number of sampling points; f1 represents the frequency regulation effect of the joint frequency regulation system, f1 ∈ [0, 1], when f1 = 0, the active power balance of the joint frequency regulation system is achieved, and the frequency does not fluctuate, and the energy storage capacity can meet the frequency regulation demand of the system; when f1 = 1, the energy storage capacity cannot meet the frequency regulation demand, so the smaller the value of f1, the better the effect of the energy storage capacity on improving the frequency of the system.

[0087] The initial investment cost C bess of the energy storage can be represented by formula (5):

[0088] C bess = c b E bess (5)

[0089] In the formula, c b represents the cost of configuring a unit capacity of energy storage; and E bess represents the capacity of the configured energy storage.

[0090] The energy storage will be charged or discharged frequently, which will affect the life of the device and speed up the aging process. The operation cost C generated by energy storage is represented by formula (6) yx :

[0091]

[0092] In the formula, c y represents the operation cost of energy storage device depreciation; E b,t represents the amount of electricity generated by the energy storage device participating in system frequency modulation in the time period t; represents the amount of electricity generated by the energy storage device participating in system frequency modulation in the time period t; represents the amount of electricity generated by the energy storage device participating in system frequency modulation in the time period t.

[0093] The photovoltaic power generation system operates according to the 20% load reduction rate to respond to the system frequency change in time, which will reduce the economic benefit of photovoltaic, so the abandoned light cost of photovoltaic is represented by formula (7):

[0094] C sun = c s (20% P MPPT,t *t) (7)

[0095] In the formula, C sun represents the price of photovoltaic field selling electricity to the grid; P MPPT,t represents the average value of the maximum power output by the photovoltaic in the maximum power tracking mode in the time period t.

[0096] In addition, the power system operator will pay auxiliary frequency modulation service charge E f for wind, light, and storage participating in frequency modulation service, which is represented by formula (8):

[0097] E f = c f E ∑ = c f (E W,t +E s,t +E b,t ) (8)

[0098] In the formula, c f represents the frequency modulation price of wind, light, and storage participating in primary frequency modulation service; E ∑ represents the total frequency modulation capacity; E W,t represents the frequency modulation capacity of the wind turbine in the time period t; E s,t represents the frequency modulation capacity of the photovoltaic in the time period t; E b,t represents the frequency modulation capacity of the energy storage in the time period t.

[0099] In summary, the annual net present value NPV of the combined system can be expressed by equation (9) as follows:

[0100]

[0101] In the formula, t1~t n n data sampling time period; D represents the sampling time; Y represents the service life of energy storage; h represents the depreciation rate of energy storage; represents the recovery coefficient of energy storage.

[0102] On the basis of meeting the best frequency stability of the multi-source combined frequency modulation system, the objective function of the multi-source combined system to improve the frequency stability and annual net income of the system is shown in equation (10) as follows:

[0103]

[0104] The equality constraint condition of the model is expressed by equation (11) as follows:

[0105] E dw,t = E G,t + E W,t + E s,t + E b,t (11)

[0106] In the formula, E G，t represents the amount of electricity of the thermal power unit participating in frequency modulation in t time period; E dw,t represents the system frequency modulation demand in t time period, i.e. the amount of electricity required by the entire network should be provided by wind, light, storage, and fire.

[0107] When the energy storage battery participates in frequency modulation, in order to prevent the occurrence of overcharging and overdischarging, SOC min = 20%, SOC max = 80%, i.e. expressed by equation (12) as follows:

[0108] SOC min ≤ SOC ≤ SOC max (12)

[0109] The MOPSO algorithm is used to solve the energy storage capacity optimization configuration problem of the multi-source combined system, and the basic process is as follows Figure 3The frequency improvement effect and the overall economy of the multi-source joint system are first taken as two targets of algorithm optimization, and each energy storage capacity optimization scheme and the corresponding frequency deviation amount are taken as particles in the population. Then, initialization processing is performed, a random initial position and speed are assigned to each particle, the fitness is calculated, the particle fitness of the current position is compared with the fitness of the historical and global best position, the best position of the particle is updated, the position and speed of the particle are updated according to formulas (11) and (12), it is judged whether the number of iterations meets the maximum number of iterations, if not, the optimization is performed again, if yes, the final result is output and the algorithm is ended.

[0110] Simulation analysis

[0111] After running the algorithm for many times, the approximate distribution of the solution space is obtained, as shown in FIG. 2. Figure 4 The abscissa f1 represents the frequency improvement effect of the system, and the ordinate f2 represents the annual net income of the system.

[0112] As can be seen from FIG. 2, Figure 4 the optimization scheme corresponding to the point No. 10 can make the objective function f2 reach the maximum, at this time, the frequency regulation effect is the worst, which is 0.821, but the net income of the system can reach the maximum, which is 1.608×10 3 6 (RMB), at this time, the corresponding optimization scheme is to configure the energy storage capacity with a size of 2.37 MW. With the gradual increase of the energy storage capacity, the net income of the system will gradually decrease, and the frequency regulation effect of the system will be obviously improved. When the energy storage capacity continues to increase, the optimization scheme corresponding to the point No. 1 can make the objective function f1 reach the minimum, at this time, the frequency regulation effect of the system can reach the best, which is 0.231, but the net income of the system is the smallest, which is 0.094×10 3 6 (RMB), at this time, the corresponding optimization scheme is to configure the energy storage capacity with a size of 2.37 MW. With the gradual increase of the energy storage capacity, the net income of the system will gradually decrease, and the frequency regulation effect of the system will be obviously improved. When the energy storage capacity continues to increase, the optimization scheme corresponding to the point No. 1 can make the objective function f1 reach the minimum, at this time, the frequency regulation effect of the system can reach the best, which is 0.231, but the net income of the system is the smallest, which is 0.094×10 3 6 (RMB), at this time, the corresponding optimization scheme is to configure the energy storage capacity with a size of 2.37 MW. With the gradual increase of the energy storage capacity, the net income of the system will gradually decrease, and the frequency regulation effect of the system will be obviously improved. When the energy storage capacity continues to increase, the optimization scheme corresponding to the point No. 1 can make the objective function f1 reach the minimum, at this time, the frequency regulation effect of the system can reach the best, which is 0.231, but the net income of the system is the smallest, which is 0.094×10

[0113] In order to verify the correctness and superiority of the best energy storage capacity configuration result obtained by the multi-objective particle swarm algorithm, the energy storage capacity of the joint system is set to 3.62 MW and compared with the system with the energy storage capacity set to 5 MW.

[0114] (1) Step load disturbance

[0115] Under Figure 5The frequency variation curve of the combined system with 3.62 MW energy storage capacity is shown by the red dotted line, and the frequency variation curve of the combined system with 5 MW energy storage capacity is shown by the purple solid line.

[0116] From Figure 5 It can be seen that when the energy storage capacity is set to 3.62 MW, the overall frequency modulation effect of the system is obviously better than that of the traditional unit and the wind-storage-fire system. Compared with the system with 5 MW energy storage capacity, the maximum frequency difference and the steady-state frequency deviation are increased by 0.0121 Hz and 0.0114 Hz respectively, and the primary frequency modulation time is increased by 0.51 s. However, the increase of energy storage capacity is accompanied by the increase of investment cost, which greatly reduces the net income of the system. Therefore, considering the dual factors of frequency modulation effect and system net income, the energy storage capacity of the multi-source combined system is configured to 3.62 MW, which can effectively improve the frequency fluctuation of the system and obtain better income return.

[0117] (2) Random square wave disturbance

[0118] Similarly, the same random square wave disturbance as above is set, and the Figure 6 is the frequency variation curve of the system under four experimental conditions.

[0119] From Figure 6 the same conclusion as Figure 5 can be drawn, which will not be repeated here.

[0120] Therefore, by comprehensively analyzing the above two load fluctuation conditions, it can be concluded that when the energy storage capacity is configured to 3.62 MW, the system frequency stability and economic efficiency can be improved, and at the same time, it fully proves the correctness and superiority of the results obtained by the multi-objective particle swarm optimization algorithm.

Claims

1. A control strategy for multi-source joint frequency regulation and a method for configuring energy storage capacity, characterized in that: The steps are as follows: S1, Wind Turbine Frequency Regulation Strategy Change in active power ΔP eg Calculated from the rate of change of system frequency deviation （1） In the formula, K j Δf2 represents the virtual inertia coefficient; Δf2 represents the frequency deviation borne by the wind turbine. Change in active power ΔP ex It is calculated from the magnitude of the system frequency deviation. （2） In the formula, K d Indicates the droop coefficient; When the frequency of the power system fluctuates, an additional frequency response unit is formed by virtual inertial control and droop control; S2, Photovoltaic Unit Frequency Regulation Strategy The load reduction level σ% of a photovoltaic array is defined as follows: （3） S3, Energy Storage System Frequency Regulation Strategy In a multi-source joint system, the frequency modulation effect of the joint frequency modulation system is evaluated by equation (4): （4） In the formula, Δf jlh f represents the j-th actual sampling frequency value using a combined frequency modulation system. jlh The difference between the rated frequency f0 and the frequency f0; Δf jct This represents the j-th actual sampling frequency value f of a conventional unit. jct The difference between the frequency and the rated frequency f0; n represents the number of sampling points; f1 represents the frequency modulation effect of the joint frequency modulation system, f1 ∈ [0, 1]. When f1 = 0, it means that the active power of the joint frequency modulation system is balanced, the frequency does not fluctuate, and the energy storage capacity can meet the frequency modulation requirements of the system; when f1 = 1, it means that the energy storage capacity cannot meet the frequency modulation requirements. Initial investment cost C for energy storage bess This can be expressed as: Equation (5) （5） In the formula, c b E represents the cost per unit capacity of energy storage; bess Indicates the capacity of the configured energy storage; Equation (6) is used to represent the operating cost C generated by energy storage. yx : （6） In the formula, c y E represents the operating cost of an energy storage device due to aging and depreciation; b,t This represents the amount of electricity that the energy storage device participates in system frequency regulation during the time period t. This represents the amount of electricity generated by the energy storage device during the time period t, which is used for system frequency regulation. This represents the amount of electricity absorbed by the energy storage device during the time period t, which is used for system frequency regulation. The cost of curtailment of photovoltaic power is expressed by equation (7): （7） In the formula, C sun This indicates the price at which a photovoltaic power plant sells electricity to the grid; P MPPT,t This represents the average maximum power output of the photovoltaic system operating in maximum power point tracking mode during the time period t. Payment for auxiliary frequency modulation service fee E f This can be expressed as: Equation (8) （8） In the formula, c f This indicates the frequency regulation price for wind, solar, and energy storage participating in primary frequency regulation services; E Ʃ Indicates total frequency modulation capacity; E W,t E represents the frequency-modulated power supplied by the fan during the time period t; s,t E represents the frequency-modulated power provided by photovoltaic power during the time period t; b,t This represents the frequency-regulating power provided by energy storage within the time period t; The annual net present value (NPV) of the combined system is expressed by equation (9): （9） In the formula, t1~t n This represents n data sampling time periods; D represents the sampling time; Y represents the lifespan of the energy storage; h represents the depreciation rate of the energy storage. Indicates the energy recovery coefficient of energy storage; The objective function for improving the system frequency stability and annual net income of a multi-source joint system is shown in equation (10): （10） The equality constraints of this model are expressed by equation (11): （11） In the formula, E G，t E represents the amount of electricity generated by thermal power units participating in frequency regulation during the time period t; dw,t This indicates the system's frequency regulation power demand during time period t, meaning that the power required by the entire network should be provided by wind, solar, energy storage, and thermal power. Configure SOC min =20%, SOC max =80%, which can be expressed as equation (12): （12） The basic process for solving the optimal configuration of energy storage capacity in a multi-source combined system is as follows: First, the frequency improvement effect and overall economy of the multi-source combined system are taken as the two objectives of the algorithm optimization. Each energy storage capacity optimization scheme and its corresponding frequency deviation are taken as particles in the population. Then, initialization is performed, and each particle is assigned a random initial position and velocity. Then, the fitness is calculated. The fitness value of the particle at the current position is compared with the fitness values ​​of the historical and global best positions, and the optimal position of the particle is updated. It is then determined whether the number of iterations at this time meets the maximum number of iterations. If it does not meet the maximum number of iterations, the optimization is repeated. If it does meet the maximum number of iterations, the final result is output and the algorithm ends.

Images