A wind-solar-hydrogen-storage system economic dispatch method

Abstract

The application is a kind of wind-light-hydrogen-storage system economic scheduling method, which considers two kinds of energy storage modes of battery and hydrogen production, stores the surplus wind energy and light energy, and uses the battery to discharge to adjust the power balance when the load power is insufficient;Analyze the operation characteristics of the hydrogen production device, coordinate and control the two kinds of energy storage by dividing the operation interval;Optimize the solution of the coordinated control of the wind-light-hydrogen-storage system under the day-ahead wind, light and load data, and finally obtain the day-ahead power distribution plan of the battery and hydrogen production, so as to realize the economic scheduling of the system. The two kinds of energy storage modes of battery and hydrogen production absorb the surplus power, reduce the abandoned wind and light, sell hydrogen and obtain operation benefit;The discharge of the energy storage battery can supplement the load power, and the coordinated control of the two kinds of energy storage modes can improve the economy of the system and make the power generation as much as possible follow the load curve.

Description

Technical Field

[0001] This invention belongs to the field of optimized dispatching of new energy power systems, and relates to an economic dispatching method for wind-solar-hydrogen-storage systems. Background Technology

[0002] Replacing traditional fossil fuels with new energy sources such as wind and solar power is crucial for the current energy structure transformation. Ensuring the safe integration of new energy sources into the grid, coordinating the output of various energy sources, and optimizing the economics of power generation are prerequisites for the long-term development of the power industry. Therefore, research on the optimal scheduling of new energy systems is of great significance for the economical and stable operation of the power grid and improving the matching degree between power generation and load.

[0003] Among numerous power system optimization and scheduling algorithms, intelligent optimization algorithms are gradually demonstrating their advantages. The White Shark Optimizer (WSO), proposed by Malik Braik et al. in 2022, is highly effective in handling complex mathematical problems with multiple constraints. Belonging to the category of biomimetic metaheuristic algorithms, its core concept and basic ideas are inspired by the hunting behavior of white sharks, including their extraordinary hearing and sense of smell while navigating and foraging in the ocean. It establishes an equivalent mathematical model of foraging behavior to accommodate the white shark's balance between exploration and exploitation, helping the search agent explore and exploit every potential region of the search space to achieve comprehensive and accurate optimization of the proposed problem. This algorithm is novel in its approach and highly efficient in its strategy.

[0004] In existing research on renewable energy dispatch, most studies only utilize energy storage batteries for storing surplus electricity, neglecting hydrogen storage as an effective alternative. In renewable energy systems, incorporating hydrogen production units can convert abundant wind and solar energy into green, clean hydrogen for long-term storage. The hydrogen production process produces only oxygen as a byproduct, which can be directly released into the atmosphere, achieving zero pollution. The hydrogen produced can also be sold to various hydrogen refueling stations to generate operational revenue. Adding appropriate amounts of energy storage batteries to the system for power balance regulation and implementing differentiated control of the two types of energy storage can achieve greater operational efficiency, thus significantly improving the system's economic efficiency while ensuring supply load. Summary of the Invention

[0005] The purpose of this invention is to propose an economical scheduling method for a wind-solar-hydrogen-storage system to address the scheduling problem of new energy systems. This method considers both battery and hydrogen production as energy storage methods, storing surplus wind and solar energy and utilizing batteries to balance power during periods of insufficient load. It analyzes the operating characteristics of the hydrogen production unit, and through the division of its operating range, coordinates and optimizes the control of the two energy storage methods to achieve economical system scheduling.

[0006] The technical solution of this invention is as follows:

[0007] An economical dispatch method for a wind-solar-hydrogen-storage system includes the following steps:

[0008] Step 1: Establish a mathematical model of the new energy system, including wind turbines, photovoltaic cells, energy storage batteries, and hydrogen production devices. Select lithium iron phosphate batteries as energy storage batteries and alkaline electrolyzers as hydrogen production devices.

[0009] Step 2: Propose an objective function that includes system operating costs, operational revenue, and load shortage penalties, and add corresponding constraints; specifically, this includes the following steps:

[0010] Step 1: Obtain the objective function to be optimized by superimposing the system operating cost, operating revenue and load shortage penalty for T time periods;

[0011] Step 2: Set operating power and ramp rate limits for energy storage batteries and hydrogen production devices to ensure that each unit can operate normally during scheduling;

[0012] Step 3: Analyze the operating characteristics of the alkaline electrolyzer and coordinate the control of energy storage and the electrolyzer; specifically, this includes the following steps:

[0013] Step 1: Based on the mathematical model of the alkaline electrolyzer, determine the relationship between its power and operating efficiency, draw its power-efficiency curve using graphing software, and analyze the optimal operating range.

[0014] Step 2: Based on the optimal operating range of the alkaline electrolyzer, design a system operation strategy to achieve coordinated control of energy storage and the electrolyzer.

[0015] Step 4: Use Tent mapping to perform chaotic initialization of the WSO algorithm, introduce the Lévy flight strategy to enhance the algorithm's global optimization ability, and introduce the random walk strategy to enhance the algorithm's local optimization, thus obtaining the improved WSO algorithm, denoted as I-WSO, thereby improving the initialization and search capabilities of the WSO algorithm.

[0016] The day-ahead wind, solar, and load data are processed by dividing a 24-hour day into T (T=96) time periods, recording data every 15 minutes to obtain load maps for wind and solar power generation, as well as combined wind and solar power and local load curves. The combined wind and solar power and local load curves are obtained from the wind and solar power load maps. From the combined wind and solar power and local load curves, it can be determined whether the power in each time period is surplus or insufficient, and the relationship between wind and solar power generation and local load power in different time periods can be determined for the execution of coordinated control strategies. The I-WSO algorithm is used to optimize the objective function by executing the coordinated control strategy in step three under the condition of satisfying constraints, and the day-ahead scheduling plans for battery and hydrogen production energy storage methods are obtained.

[0017] To elaborate further, step one above specifically includes the following steps:

[0018] Step 1-1: Establish the equivalent mathematical model of the power generation unit:

[0019] Step 1-1-1: Establish a wind turbine model, with the wind turbine output power P. Wind (t) is shown in equation (1):

[0020]

[0021] In the formula, v(t) is the wind speed during time period t; V in Cut-in wind speed; V out To cut off the wind speed; V ra Rated wind speed; P ra This refers to the rated power of the fan;

[0022] Step 1-1-2: Establish a photovoltaic cell model, and determine the output power P of the photovoltaic cell. PV (t) is shown in equation (2):

[0023]

[0024] In the formula, P st Photovoltaic cell output power under standard test conditions; K D For photovoltaic cells, the derating factor; G act (t) represents the actual light intensity during time period t; G st Light intensity under standard test conditions; α is the power temperature coefficient; T w The operating temperature of the solar panel; T r For reference temperature;

[0025] Step 1-2: Establish the equivalent mathematical model of the energy storage unit:

[0026] Step 1-2-1: Establish a lithium iron phosphate battery model, and determine the remaining capacity C of the lithium iron phosphate battery. soc (t) is shown in equation (3):

[0027]

[0028] In the formula, ε is the battery self-discharge efficiency; η - For battery discharge efficiency; η + Battery charging efficiency; P SB (t) represents the battery interaction power during time period t, where P SB (t)<0 indicates discharge, P SB (t)>0 indicates charging;

[0029] Step 1-2-2: Establish a hydrogen energy storage unit model. The chemical reaction equation of the alkaline electrolyzer is shown in equation (4), and the input power of the hydrogen energy storage unit is expressed as equation (5):

[0030]

[0031] P cell =V cell I cell (5)

[0032]

[0033] I cell =i den S cell (7) In the formula, V cell This is the input voltage for the alkaline electrolytic cell; I cell The input current for the alkaline electrolytic cell; r1, r2, s, q1, q2, q3 are empirical coefficients; A is the electrode area; T is the operating temperature; i den S represents the current density. cell V represents the electrode area; rev The potential is reversible and is taken as 1.23V under standard conditions;

[0034] Step two above proposes an objective function and sets constraints, specifically including the following steps:

[0035] Step 2-1: Propose an objective function that includes system operating costs, operating revenue, and load shortage penalties, as shown in equation (8):

[0036] F = min(C) om -C ope +C loss (8)

[0037]

[0038] In the formula, F represents the total cost of the energy storage and hydrogen production equipment; C om For operating costs; C ope For operating revenue; C loss For load loss penalty; T is the number of time periods in a day, T = 96; D wind D PV D sb D cell The unit power operation and maintenance costs are respectively for wind turbines, photovoltaic cells, lithium iron phosphate batteries, and alkaline electrolyzers; C is the unit hydrogen price. q represents the volume of hydrogen produced during time period t; s P is the penalty factor. g (t) represents the power generation during time period t; P s (t) represents the load power during time period t;

[0039] Step 2-2: Set the constraints for each unit;

[0040] Step 2-2-1: Set the power balance limit as shown in equation (12):

[0041] P Wind (t)+P PV (t)=P SB (t)+P cell (t)+P s (t) (12) Step 2-2-2: Set the lithium iron phosphate battery interactive power limit as shown in equation (13), the state of charge limit as shown in equation (14), and the cycle operation limit as shown in equation (15):

[0042]

[0043] In the formula, The lower and upper limits of the interaction power of lithium iron phosphate batteries; SOC min SOC max These represent the lower and upper limits of the state of charge (SOC) of lithium iron phosphate batteries; C s This refers to the capacity of lithium iron phosphate batteries;

[0044] Step 2-2-3: Set the operating limits of the electrolytic cell as shown in equations (16) and (17):

[0045]

[0046] In the formula, These are the minimum and maximum power requirements for operating an alkaline electrolyzer. This represents the maximum ramp rate for an alkaline electrolytic cell.

[0047] Step three above analyzes the operating characteristics of the alkaline electrolyzer and designs the system operation control strategy, specifically including the following steps:

[0048] Step 3-1: Analyze the efficiency characteristics of the alkaline electrolyzer. The operating efficiency of the alkaline electrolyzer can be expressed as equation (17):

[0049]

[0050] In the formula, η is the operating efficiency of the alkaline electrolyzer; V th For thermal neutral potential, take V under standard conditions. th =1.48V; After obtaining the operating efficiency of the alkaline electrolyzer, the power-efficiency curve is plotted using graphing software. The curve is then analyzed for characteristics. The optimal operating range of the alkaline electrolyzer is determined by considering two factors: efficient operation of the alkaline electrolyzer and hydrogen production. The optimal operating range of the alkaline electrolyzer is obtained based on the power-efficiency curve to ensure efficient operation of the electrolyzer.

[0051] Step 3-2: Set the system operation control strategy with the goal of economic optimization:

[0052] Step 3-2-1: When the wind and solar power generation is greater than the local load power P load (t) and the upper limit power of alkaline electrolyzer operation The sum of When: The state of charge (SOC) of the lithium iron phosphate battery is greater than 0.5, i.e. Remaining power C soc (t), C s To optimize the capacity of lithium iron phosphate batteries, the alkaline electrolyzer is set to operate at its maximum power limit. Remaining power P SB Absorbed by lithium iron phosphate batteries; when the state of charge of the lithium iron phosphate battery is less than or equal to 0.5, i.e. At that time, the operating power range of the electrolytic cell was set to be... Remaining power P SB Absorbed by lithium iron phosphate batteries;

[0053] Step 3-2-2: When the wind and solar power generation is greater than the sum of the local load power and the lower limit of the alkaline electrolyzer's operating power, but less than or equal to the sum of the local load power and the upper limit of the alkaline electrolyzer's operating power, that is... At that time, the operating power range of the alkaline electrolyzer was set to... The excess or insufficient power is absorbed or supplemented by the lithium iron phosphate battery;

[0054] Step 3-2-3: When the wind and solar power generation is less than or equal to the sum of the local load power and the lower limit of the alkaline electrolyzer's operating power, i.e. To ensure power supply to the load, the alkaline electrolytic cell is set to operate at its lower power limit, i.e. Insufficient power is supplemented by battery discharge;

[0055] The improved WSO algorithm in step four above specifically includes the following steps:

[0056] Step 4-1: Initialize algorithm parameters, including maximum number of iterations I, population size N, variable space dimension D, and upper bound of variables x. max Lower bound x min ;

[0057] Step 4-2: Selecting the Tent mapping to perform chaotic processing on the initial solution of the WSO algorithm can improve the population quality and distribution uniformity while ensuring population diversity, which is beneficial to the subsequent search process of the algorithm.

[0058] Step 4-2-1: Randomly generate a value within (0,1) and denote it as z1, and denote d = 1;

[0059] Step 4-2-2: The Tent mapping expression is equation (18):

[0060]

[0061] In the formula, α is the number of particles in the chaotic sequence, and rand(0,1) is a random number in (0,1);

[0062] Step 4-2-3: Obtain the chaotic variable z from the Tent mapping expression through Bernoulli transformation. d+1 As shown in equation (19):

[0063]

[0064] Step 4-2-4: d = d + 1; Determine whether d reaches the variable space dimension D. If yes, save the generated D-dimensional chaotic sequence; otherwise, return to step 4-2-2 and recalculate the chaotic variable of the next dimension until a D-dimensional chaotic sequence is generated.

[0065] Step 4-2-5: Generate N D-dimensional chaotic sequences Z1,…,Z using steps 4-2-1 to 4-2-4. N z is a numerical value, Z is a sequence of z, and the initial position of the population is generated by equation (20), which is described in matrix form as shown in equation (21):

[0066] X i =x min +(x max -x min )Z i (20)

[0067]

[0068] Step 4-3: Calculate the global optimal solution X best Let k = 1;

[0069] Step 4-4: Calculate the individual's movement speed toward the prey, and update the formula as shown in (22):

[0070]

[0071] In the formula, i = 1, 2, ..., N, i represents the nth individual, and d represents the nth dimension of the individual; For the new velocity vector; X is the current velocity vector; best,k This represents the optimal position obtained in the kth generation. X is the current position vector;best The global optimal position; c1 and c2 are random numbers within (0,1); p1 and p2 represent the strength of the individual white shark; μ is the suggestion contraction factor; k is the current iteration number; p min p max Let p1 and p2 be the minimum and maximum values, respectively. min =0.5, p2=1.5; Y is the acceleration coefficient, Y=4.125;

[0072] Steps 4-5: Update to the location of the best prey; when a white shark hears the sound of waves caused by the movement of prey or smells its scent, its location will be updated as follows:

[0073]

[0074] In the formula, For the new individual position; A negation operator; a and b are binary vectors; X O f represents a logical vector; f represents the fluctuation frequency; mv represents the kinetic force; f min f max f represents the minimum and maximum oscillation frequencies. min =0.07, f max =0.75; a0 and a1 represent two positive numbers governing exploration and development activities;

[0075] Steps 4-6: Update the global optimal solution;

[0076] Steps 4-7: Introducing the Lévy Flight Strategy; To enhance the algorithm's global optimization capability, the Lévy flight strategy is introduced. To reduce the computational load and ensure the speed of optimization, the following strategy is executed when rand(0,1)>0.5, otherwise...

[0077]

[0078]

[0079] Steps 4-8: Update the optimal value for generation k+1. Since the Levy flight strategy causes individuals to move, but the quality of the new position is unknown, update the optimal value for generation k+1 according to the following formula:

[0080]

[0081] In the formula, F(X) best,k+1 ) is X best,k+1 The value obtained by substituting it into the objective function; for The value obtained by substituting it into the objective function;

[0082] Steps 4-9: Update to the position of the best individual; when r3 s When i=1, the white shark individual will move towards the optimal individual position and make the following update:

[0083]

[0084] In the formula, The position to be updated; sgn(r2-0.5) is used to change the search direction; r1, r2, and r3 are random numbers within (0,1); a1 represents the distance between the prey and the great white shark; a2 is a positive number controlling exploration and development, a2 = 0.0005;

[0085] Steps 4-10: Update position based on fish behavior: when r3 s When i>1, the white shark individual will not only update its position to the best individual according to equation (34), but also update its position according to the fish school behavior according to equation (38):

[0086]

[0087] Step 4-11: Introduce a random walk strategy to enhance the algorithm's local optimization ability, and calculate X. best,k+1 And perform random walk on it:

[0088] X′ best,k+1 =X best,k+1 +ε(X k+1,g -X k+1,w ) 40) In the formula, X′ best,k+1 X represents the optimal individual in generation k+1 after introducing the random walk strategy; ε is the scaling factor, ε ~ U(0,1); k+1,g X k+1,w There are two random solutions for the k+1 generation;

[0089] Step 4-12: Update the global optimal solution and the k+1 generation optimal solution:

[0090]

[0091] Step 4-13: k = k + 1, check if the maximum number of iterations I has been reached. If yes, exit the loop and execute step 4-14; otherwise, return to step 4-4.

[0092] Step 4-14: Output X best F(X) best );

[0093] Steps 4-15: Process the daytime wind, solar and load data, divide the 24 hours of a day into 96 time periods, record the data every 15 minutes, and obtain wind power generation, photovoltaic power generation and load curves;​​

[0094] Step 4-16: Utilize the I-WSO algorithm to optimize the objective function by performing coordinated control in Step 3 under the constraints; input wind, solar, and load data, substitute the objective function into the I-WSO algorithm, and iteratively derive the optimal cost and the day-ahead scheduling plan for lithium iron phosphate batteries and alkaline electrolyzers;

[0095] Compared with existing technologies, the beneficial effects of this paper are:

[0096] In response to the intermittent and fluctuating nature of renewable energy sources such as wind and solar power, which cannot meet load demand when connected to the grid, resulting in a large deviation between power generation and load curves and significant curtailment of wind and solar power, thus compromising economic efficiency and reliability, this invention proposes a solution. This solution utilizes both energy storage batteries and hydrogen production and storage to absorb surplus power, reducing wind and solar curtailment while simultaneously generating operational benefits through hydrogen sales. Furthermore, the invention supplements load power by discharging energy storage batteries and coordinates the control of both energy storage methods, thereby improving system economy while maximizing the alignment of power generation with the load curve.

[0097] This invention employs the I-WSO algorithm to solve the system model. The improved WSO algorithm boasts faster search capabilities and higher optimization accuracy. The I-WSO algorithm is used to optimize the system model under day-ahead wind, solar, and load data. Comparison with other intelligent algorithms shows that the I-WSO algorithm improves both convergence speed and accuracy, resulting in a 20.97% increase in system economy. This effectively enhances system economy and enables the generation of day-ahead power allocation plans for lithium iron phosphate batteries and alkaline electrolyzers, thereby achieving economical system scheduling. Attached Figure Description

[0098] Figure 1 This is a schematic diagram of the method flow for the economic dispatching method of the wind-solar-hydrogen-storage system of the present invention.

[0099] Figure 2 This is a schematic diagram of a new energy system structure according to an embodiment of the present invention.

[0100] Figure 3 It is the power-efficiency curve of an alkaline electrolyzer.

[0101] Figure 4 This is a flowchart of the control strategy for coordinated control of two types of energy storage in the system of this invention.

[0102] Figure 5 This is a flowchart of the scheduling algorithm of the present invention.

[0103] Figure 6 These are the power curves for wind turbines and photovoltaic cells.

[0104] Figure 7It is a power curve graph of combined wind and solar power generation and local load.

[0105] Figure 8 This is a comparison chart of optimization results obtained by different optimization algorithms.

[0106] Figure 9 This is the daytime scheduling plan for alkaline electrolyzers.

[0107] Figure 10 This refers to the day-ahead scheduling plan and status of charge of lithium iron phosphate batteries. Detailed Implementation

[0108] The embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0109] To describe the present invention in more detail, the technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0110] The structure diagram of the scheduling method in this embodiment is as follows: Figure 1 As shown, this includes modeling the new energy system, establishing a mathematical model of the new energy system, establishing an economic indicator objective function, setting operating constraints for each unit of the system, analyzing the operating characteristics of the alkaline electrolyzer and providing a control strategy for coordinated control of the system operation, processing day-ahead data, and using an improved WSO algorithm to solve the system model, thereby obtaining a day-ahead scheduling plan that takes into account economic efficiency.

[0111] The specific steps are as follows:

[0112] Step 1: Establish a mathematical model of the new energy system, including wind turbines, photovoltaic cells, energy storage batteries, and hydrogen production devices. The energy storage battery is selected as a lithium iron phosphate battery, and the hydrogen production device as an alkaline electrolyzer. The system model structure diagram is shown below. Figure 2 As shown;

[0113] Step 1-1: Establish the equivalent mathematical model of the power generation unit:

[0114] Step 1-1-1: Establish a wind turbine model with output power P. Wind (t) is shown in equation (1):

[0115]

[0116] In the formula, v(t) is the wind speed during time period t; V in Cut-in wind speed; V out To cut off the wind speed; V ra Rated wind speed; P ra This refers to the rated power of the fan;

[0117] Step 1-1-2: Establish a photovoltaic cell model with output power P. PV (t) is shown in equation (2):

[0118]

[0119] In the formula, P st Photovoltaic cell output power under standard test conditions; K D For photovoltaic cells, the derating factor; G act (t) represents the actual light intensity during time period t; G st Light intensity under standard test conditions; α is the power temperature coefficient; T w The operating temperature of the solar panel; T r For reference temperature;

[0120] Step 1-2: Establish the equivalent mathematical model of the energy storage unit:

[0121] Step 1-2-1: Establish a lithium iron phosphate battery model, with its remaining capacity C. soc (t) is shown in equation (3):

[0122]

[0123] In the formula, ε is the battery self-discharge efficiency; η - For battery discharge efficiency; η + Battery charging efficiency; P SB (t) represents the battery interaction power during time period t, which is the remaining power, where P SB (t)<0 indicates discharge, P SB (t)>0 indicates charging;

[0124] Step 1-2-2: Establish a hydrogen energy storage unit model. The chemical reaction equation of the alkaline electrolyzer is shown in equation (4), and the input power is expressed as equation (5):

[0125]

[0126] P cell =V cell I cell (5)

[0127]

[0128] I cell =i den S cell (7)

[0129] In the formula, V cell This is the input voltage for the alkaline electrolytic cell; I cell The input current for the alkaline electrolytic cell; r1, r2, s, q1, q2, q3 are empirical coefficients; A is the electrode area; T is the operating temperature; i den S represents the current density. cellV represents the electrode area; rev The potential is reversible and is taken as 1.23V under standard conditions;

[0130] Step 2: Propose an objective function that includes system operating costs, operational revenue, and load shortage penalties, and add corresponding constraints; specifically, this includes the following steps:

[0131] Step 2-1: Propose an objective function that includes system operating costs, operating revenue, and load shortage penalties, as shown in equation (8):

[0132] F = min(C) om -C ope +C loss (8)

[0133]

[0134]

[0135] In the formula, F represents the total cost of the energy storage and hydrogen production equipment; C om For maintenance costs; C ope For operating revenue; C loss For load loss penalty; T is the number of time periods in a day, T = 96; D wind D PV D sb D cell The unit power operation and maintenance costs are respectively for wind turbines, photovoltaic cells, lithium iron phosphate batteries, and alkaline electrolyzers; C is the unit hydrogen price. q represents the volume of hydrogen produced during time period t; s P is the penalty factor. g (t) represents the power generation during time period t; P s (t) represents the load power during time period t;

[0136] Step 2-2: Set the constraints for each unit;

[0137] Step 2-2-1: Set the power balance limit as shown in equation (12):

[0138] P Wind (t)+P PV (t)=P SB (t)+P cell (t)+P s (t) (12) Step 2-2-2: Set the lithium iron phosphate battery interactive power limit as shown in equation (13), the state of charge limit as shown in equation (14), and the cycle operation limit as shown in equation (15):

[0139]

[0140] Csoc (0)=C soc (T) (15)

[0141] In the formula, The lower and upper limits of the interaction power of lithium iron phosphate batteries; SOC min SOC max These represent the lower and upper limits of the state of charge (SOC) of lithium iron phosphate batteries; C s This refers to the capacity of lithium iron phosphate batteries;

[0142] Step 2-2-3: Set the operating limits of the electrolytic cell as shown in equations (16) and (17):

[0143]

[0144] In the formula, These are the minimum and maximum power requirements for operating an alkaline electrolyzer. This represents the maximum ramp rate for an alkaline electrolytic cell.

[0145] Step 3: Analyze the operating characteristics of the alkaline electrolyzer and design a system operation control strategy; specifically, this includes the following steps:

[0146] Step 3-1: Analyze the efficiency characteristics of the alkaline electrolyzer. The operating efficiency of the alkaline electrolyzer can be expressed as equation (17):

[0147]

[0148] In the formula, η is the operating efficiency of the alkaline electrolyzer; V th For thermal neutral potential, take V under standard conditions. th =1.48V; After obtaining the operating efficiency of the alkaline electrolyzer, the power-efficiency curve was plotted using graphing software, as shown below. Figure 3 As shown in the figure, the peak efficiency of the alkaline electrolyzer is around 0.2 pu. However, too low a power will result in very low hydrogen production. Therefore, it is necessary to consider both the efficient operation of the alkaline electrolyzer and the hydrogen production. The optimal operating range of the alkaline electrolyzer is set to 0.5 pu-0.9 pu.

[0149] Step 3-2: Based on the power-efficiency curve, the optimal operating range for the alkaline electrolyzer is obtained as 0.5 pu-0.9 pu. To achieve economic optimization, the following coordinated control system operation control strategy is implemented, and its flowchart is shown below. Figure 4 As shown:

[0150] Step 3-2-1: When the wind and solar power generation is greater than the sum of the local load power and the upper limit of the alkaline electrolyzer's operating power, i.e. When: The state of charge of the lithium iron phosphate battery is greater than 0.5%, and At that time, the alkaline electrolyzer is set to operate at its maximum power limit, i.e. The remaining power is absorbed by the lithium iron phosphate battery; when the state of charge of the lithium iron phosphate battery is less than 0.5%, that is... At that time, the operating power range of the electrolytic cell was set to be... The remaining power is absorbed by the lithium iron phosphate battery;

[0151] Step 3-2-2: When the sum of wind and solar power generation power and the local load power and the lower limit of alkaline electrolyzer operation power is less than the sum of the local load power and the upper limit of alkaline electrolyzer operation power, that is... At that time, the operating power range of the alkaline electrolyzer was set to... The excess or insufficient power is absorbed or supplemented by the lithium iron phosphate battery;

[0152] Step 3-2-3: When the wind and solar power generation is less than the sum of the local load power and the lower limit of the alkaline electrolyzer's operating power, i.e. To ensure power supply to the load, the alkaline electrolytic cell is set to operate at its lower power limit, i.e. Insufficient power is supplemented by battery discharge;

[0153] Step 4: Improve the initialization and search capabilities of the WSO algorithm to obtain the improved WSO algorithm, denoted as I-WSO, whose flowchart is shown below. Figure 5 As shown; the daytime wind, solar, and load data are processed, dividing the 24 hours into T (T=96) time periods, with data recorded every 15 minutes. The output power diagrams of wind turbines and photovoltaic cells are shown below. Figure 6 As shown, the combined wind and solar power generation and load power are as follows: Figure 7 As shown; the I-WSO algorithm is used to optimize the objective function under the constraints to obtain the day-ahead scheduling plan for each unit;

[0154] Step 4-1: Initialize algorithm parameters, including maximum number of iterations I, population size N, variable space dimension D, and upper bound of variables x. max Lower bound x min ;

[0155] Step 4-2: Improve the initialization strategy of the WSO algorithm; the random initialization of the initial population has too high randomness and uneven distribution, which is not conducive to the subsequent search process of the algorithm. Therefore, chaotic processing is performed on the initial solution, which can improve the population quality while ensuring population diversity. Since the traversal uniformity of the Tent mapping is better than that of the common Logistic mapping, the Tent mapping is selected to perform chaotic processing on the initial solution.

[0156] Step 4-2-1: Randomly generate a value within (0,1) and denote it as z1, and denote d = 1;

[0157] Step 4-2-2: The Tent mapping expression is equation (18):

[0158]

[0159] In the formula, α is the number of particles in the chaotic sequence, and rand(0,1) is a random number in (0,1);

[0160] Step 4-2-3: Obtain the chaotic variable z from the Tent mapping expression through Bernoulli transformation. d+1 As shown in equation (19):

[0161]

[0162] Step 4-2-4: d = d + 1; Determine whether d reaches the variable space dimension D. If yes, save the generated D-dimensional chaotic sequence; otherwise, return to step 4-2-2 and recalculate the chaotic variable of the next dimension until a D-dimensional chaotic sequence is generated.

[0163] Step 4-2-5: Generate N D-dimensional chaotic sequences Z1,…,Z using steps 4-2-1 to 4-2-4. N Z is a sequence composed of z, which is the chaotic variable. The initial position of the population is generated by Z using equation (20), and it is described in matrix form as shown in equation (21):

[0164] X i =x min +(x max -x min )Z i (20),

[0165]

[0166] Step 4-3: Calculate the global optimal solution X best Let k = 1;

[0167] Step 4-4: Calculate the individual's movement speed toward the prey, and update the formula as shown in (22):

[0168]

[0169] In the formula, i = 1, 2, ..., N; For the new velocity vector; X is the current velocity vector; best,k This represents the optimal position obtained in the kth generation. X is the current position vector; best The global optimal position; c1 and c2 are random numbers within (0,1); p1 and p2 represent the strength of the individual white shark; μ is the suggestion contraction factor; k is the current iteration number; pmin p max Let p1 and p2 be the minimum and maximum values, respectively. min =0.5, p2=1.5; Y is the acceleration coefficient, Y=4.125;

[0170] Steps 4-5: Update to the location of the best prey; when a white shark hears the sound of waves caused by the movement of prey or smells its scent, its location will be updated as follows:

[0171]

[0172] In the formula, For the new individual position; A negation operator; a and b are binary vectors; X O f represents a logical vector; f represents the fluctuation frequency; mv represents the kinetic force; f min f max f represents the minimum and maximum oscillation frequencies. min =0.07, f max =0.75; a0 and a1 represent two positive numbers governing exploration and development activities;

[0173] Steps 4-6: Update the global optimal solution;

[0174] Steps 4-7: Introducing the Lévy Flight Strategy; To enhance the algorithm's global optimization capability, the Lévy flight strategy is introduced. To reduce the computational load and ensure the speed of optimization, the following strategy is executed when rand(0,1)>0.5, otherwise...

[0175]

[0176] Steps 4-8: Update the optimal value for generation k+1. Since the Levy flight strategy causes individuals to move, but the quality of the new position is unknown, update the optimal value for generation k+1 according to the following formula:

[0177]

[0178] In the formula, F(X) best,k+1 ) is X best,k+1 The value obtained by substituting it into the objective function; for The value obtained by substituting it into the objective function;

[0179] Steps 4-9: Update to the position of the best individual; when r3 s When i=1, the white shark individual will move towards the optimal individual position and make the following update:

[0180] ​

[0181] In the formula, The position to be updated; sgn(r2-0.5) is used to change the search direction; r1, r2, and r3 are random numbers within (0,1); a1 represents the distance between the prey and the great white shark; a2 is a positive number controlling exploration and development, a2 = 0.0005;

[0182] Steps 4-10: Update position based on fish behavior: when r3 s When i>1, the white shark individual will not only update its position to the best individual according to equation (35), but also update its position according to the fish school behavior according to equation (38):

[0183]

[0184] Step 4-11: Introduce a random walk strategy to enhance the algorithm's local optimization ability. After obtaining the latest positions of N individuals in sequence, calculate X. best,k+1 And perform random walk on it:

[0185] X′ best,k+1 =X best,k+1 +ε(X k+1,i -X k+1,j (40)

[0186] In the formula, X′ best,k+1 X represents the optimal individual in generation k+1 after introducing the random walk strategy; ε is the scaling factor, ε ~ U(0,1); k+1,i X k+1,j There are two random solutions for the k+1 generation;

[0187] Step 4-12: Update the global optimal solution and the k+1 generation optimal solution:

[0188]

[0189] Step 4-13: k = k + 1, check if the maximum number of iterations I has been reached. If yes, exit the loop and execute step 4-14; otherwise, return to step 4-4.

[0190] Step 4-14: Output X best F(X) best ), which are the obtained objective function value and optimal solution vector;

[0191] Steps 4-15: Process the daytime wind, solar, and load data, dividing the 24 hours into 96 time periods, recording data every 15 minutes, to obtain wind power generation, solar power generation curves (with time h as the x-axis and power kW as the y-axis), and power curves of combined wind and solar power generation and local load power, as shown below. Figures 6-7 ; ​

[0192] Step 4-16: Use the I-WSO algorithm to perform coordinated control to optimize the objective function under the constraints; input wind, solar, and load data, substitute the objective function into the I-WSO algorithm, and iteratively obtain the optimal cost and the day-ahead scheduling plan for lithium iron phosphate batteries and alkaline electrolyzers;

[0193] In this embodiment, the day-ahead power curves of wind, solar, and load for a typical day are given. The data comes from EirGrid Group and global solar atlas. According to the method proposed in this invention, the day-ahead optimization scheduling of this set of data is performed to obtain the day-ahead scheduling plan for alkaline electrolyzers and lithium iron phosphate batteries for 96 time periods. Substituting the scheduling plan into the objective function, the daily operating cost can be obtained. In order to avoid randomness, 20 independent experimental calculations are performed, the average convergence curve of the 20 experiments is plotted, and the mean and variance are calculated. By comparing the results with other methods, the operating cost optimized by this invention is lower and the system economy is higher. For specific results, see the simulation example.

[0194] Step 4-17: Output the optimization results and their corresponding solutions, and analyze the obtained operating cost values, system operating benefits, power shortage rate, and day-ahead scheduling plans for alkaline electrolyzers and lithium iron phosphate batteries;

[0195] Simulation examples:

[0196] This example is applied to day-ahead optimization scheduling on a typical day. Day-ahead wind, solar, and load data for that day are processed, dividing the 24 hours into 96 time periods, with data recorded every 15 minutes. Using the wind-solar-hydrogen-storage system economic scheduling method based on the I-WSO algorithm of this invention, day-ahead optimization scheduling is performed on this typical day, and compared with other optimization algorithms to demonstrate that the proposed algorithm has a faster search speed and better convergence accuracy, can significantly reduce system operating costs, and ensures that power generation follows the load curve, thereby reducing power shortage losses and improving the overall system economy.

[0197] Based on the mathematical model of the new energy system of wind-solar-hydrogen-storage established in step one, this invention selects three alkaline electrolyzers with a rated power of 64kW and three lithium iron phosphate batteries with a rated capacity of 200kWh for day-ahead optimization scheduling, wherein the initial capacity of the lithium iron phosphate batteries is 120kWh. Based on step two, an economic objective function is constructed and constraints are set for each unit, including the parameters and values ​​shown in Table (1):

[0198] Table 1 Limiting parameters of each unit in the system

[0199]

[0200] To verify the performance of the I-WSO algorithm, typical single-peak and multi-peak test functions were selected for testing. To avoid randomness, 10 independent experiments were conducted, and the experimental results were compared with other intelligent optimization algorithms, as shown in Table 2. The parameters for all algorithms were set as follows: population size 20, number of iterations 1000, and variable space dimension 30.

[0201] Table 2 Optimization results of the test function

[0202]

[0203] By analyzing the average value and variance of the optimization results of each algorithm on the test function in Table 2, it can be seen that the I-WSO algorithm has the smallest average value and the lowest standard deviation, indicating that the scheduling method of the present invention has better optimization effect and better robustness.

[0204] Initialize the I-WSO algorithm parameters according to step four: maximum number of iterations I is 800, population size N is 40, variable space dimension D is 192, and variable upper bound x... max Lower bound x min Determined according to Table (1). Analyze the day-ahead wind, solar, and load data as follows: Figures 6-7 It is known that the combined wind and solar power output cannot meet the load demand during 0-2.5 hours, 6-8 hours, and 15-19 hours, while surplus power exists during other time periods. The proposed method is used to optimize the system's economic scheduling, and 20 independent experiments are conducted to verify the stability of the I-WSO algorithm. The average value of the 20 experimental results is used to plot the convergence curve, which is then compared with the original WSO algorithm and the ASO algorithm. Figure 8 As shown, the I-WSO algorithm converges to 6962.89 yuan in the 420th iteration, the WSO algorithm converges to 8810.02 yuan in the 500th iteration, and the ASO algorithm converges to 8116.84 yuan in the 480th iteration. The I-WSO algorithm proposed in this paper has the fastest convergence speed and the highest convergence accuracy, which is 20.97% higher than the WSO algorithm before the improvement.

[0205] The optimization results of the three algorithms are listed in Table 3. The analysis shows that the average value of the optimization result of the I-WSO algorithm is the best, while the ASO algorithm is slightly better than the WSO algorithm. Comparing the variances, the I-WSO algorithm has the lowest variance value, while the ASO algorithm has the highest, indicating that the I-WSO algorithm has the highest stability. In addition, the I-WSO algorithm also has the highest operating benefits, and its load power shortage rate of 0.0420 (less than 5%) is the lowest among the three, which fully proves that the method proposed in this invention can effectively improve the economy of the system.

[0206] Table 3. Results of 20 experiments for the three algorithms.

[0207]

[0208] The day-ahead scheduling plan for alkaline electrolyzers, optimized by the method proposed in this paper, is as follows: Figure 9 As shown, from 0:00 to 8:00, the electrolyzer operated at its lower power limit due to the low combined output of wind and solar power, with no significant fluctuations. From 8:00 to 11:00, the combined output of wind and solar power was high and could meet the load, so the electrolyzer power showed an upward trend during this period, with a large ramp rate, reaching the upper power limit at 10:00 and maintaining a high power operation from 10:00 to 13:00. From 12:00 to 14:00, the combined output of wind and solar power decreased, and the electrolyzer operating power also showed a downward trend, reaching the lower power limit at 14:00. From 14:00 to 19:00, the electrolyzer continued to operate at the lower limit without significant fluctuations. Starting from 19:00, due to the increase in wind power and the decrease in load power, there was a lot of surplus power, and the electrolyzer power showed a significant upward trend, reaching the upper limit again at 21:00 and maintaining it until 24:00.

[0209] Analysis of the battery interaction power and state of charge of lithium iron phosphate batteries, such as Figure 10 As shown, the battery is in a discharging state from 0-3, 6-8, and 15-20 hours, with the state of charge (SOC) continuously decreasing, reaching the lower limit at 3 and 20 hours, and operating close to the lower limit at 8. During charging from 3-6, 8-15, and 20-24 hours, the SOC continuously increases, reaching the upper limit at 14. Observation of the curves shows that the interactive power during battery charging and discharging meets the power limits, and the SOC does not exceed the maximum or minimum limits, indicating high battery utilization. The day-ahead scheduling plan for the electrolyzer and battery obtained through this method is relatively reasonable, with the system load under-supply rate below 5%, meeting the power demand and significantly improving overall economic efficiency, fully demonstrating the applicability of this invention in economical scheduling scenarios.

[0210] The above description is only one application scenario of the present invention. Any equivalent changes made within the scope of the patent application of the present invention, or applications to the economic dispatch of other new energy power systems, should fall within the scope of the present invention.

[0211] Matters not covered in this invention are common knowledge.

Claims

1. An economical dispatch method for a wind-solar-hydrogen-storage system, characterized in that, This scheduling method considers both battery and hydrogen production energy storage methods, stores surplus wind and solar energy, and uses batteries to discharge and regulate power balance when the load power is insufficient; it analyzes the operating characteristics of the hydrogen production unit and coordinates the two energy storage methods by dividing its operating range. The coordinated control of the wind-solar-hydrogen-storage system under day-ahead wind, solar and load data is optimized and solved. Finally, the day-ahead power allocation plan of battery and hydrogen production is obtained, thereby realizing the economic scheduling of the system. The specific process of coordinating and controlling the two types of energy storage by analyzing the operating characteristics of the hydrogen production unit and dividing its operating range is as follows: Step 3-1: The hydrogen production device is an alkaline electrolyzer, and the battery is a lithium iron phosphate battery. The efficiency characteristics of the alkaline electrolyzer are analyzed, and the operating efficiency of the alkaline electrolyzer is expressed by equation (17): In the formula, η is the operating efficiency of the alkaline electrolyzer; V th For thermal neutral potential, take V under standard conditions. th =1.48V; V cell This is the input voltage for the alkaline electrolytic cell; After obtaining the operating efficiency of the alkaline electrolyzer, a power-efficiency curve is plotted using graphing software. Considering the two factors of efficient operation and hydrogen production, the optimal operating range of the alkaline electrolyzer is determined based on the power-efficiency curve. Step 3-2: Set the system operation control strategy with the goal of economic optimization: Step 3-2-1: When the wind and solar power generation is greater than the local load power P load (t) and the upper limit power of alkaline electrolyzer operation The sum of When: The state of charge (SOC) of the lithium iron phosphate battery is greater than 0.5, i.e. Remaining power C soc (t), C s To optimize the capacity of lithium iron phosphate batteries, the alkaline electrolyzer is set to operate at its maximum power limit. Remaining power P SB Absorbed by lithium iron phosphate batteries; when the state of charge of the lithium iron phosphate battery is less than or equal to 0.5, i.e. At 0.5, the operating power range of the electrolytic cell is set as follows: The remaining power is absorbed by the lithium iron phosphate battery; Step 3-2-2: When the wind and solar power generation is greater than the sum of the local load power and the lower limit of the alkaline electrolyzer's operating power, but less than or equal to the sum of the local load power and the upper limit of the alkaline electrolyzer's operating power, that is... At that time, the operating power range of the alkaline electrolyzer was set to... The excess or insufficient power is absorbed or supplemented by the lithium iron phosphate battery; Step 3-2-3: When the wind and solar power generation is less than or equal to the sum of the local load power and the lower limit of the alkaline electrolyzer's operating power, i.e. To ensure power supply to the load, the alkaline electrolytic cell is set to operate at its lower power limit, i.e. Insufficient power is supplemented by battery discharge; Among them, P Wind (t) represents the output power of the wind turbine model; P PV (t) represents the output power of the photovoltaic cell model.

2. The economic dispatch method for a wind-solar-hydrogen-storage system according to claim 1, characterized in that, The scheduling method includes the following steps: Step 1: Establish a mathematical model of the new energy system, including wind turbines, photovoltaic cells, energy storage batteries, and hydrogen production devices. Select lithium iron phosphate batteries as energy storage batteries and alkaline electrolyzers as hydrogen production devices. Step 2: Set the objective function, including system operating costs, operating revenue, and load shortage penalties, and add corresponding constraints; specifically, this includes the following steps: Step 1: Obtain the objective function to be optimized by superimposing the system operating cost, operating revenue and load shortage penalty for T time periods; Step 2: Set limits on operating power and ramp rate for the energy storage battery and hydrogen production device to ensure that the energy storage battery and hydrogen production device can operate normally during scheduling; Step 3: Analyze the operating characteristics of the alkaline electrolyzer to achieve coordinated control of energy storage and the electrolyzer; Step 4: Use the Tent mapping to perform chaotic initialization on the WSO algorithm, introduce the Lévy flight strategy to enhance the algorithm's global optimization ability, and introduce the random walk strategy to enhance the algorithm's local optimization, thus obtaining the improved WSO algorithm, denoted as I-WSO. The wind, solar and load data are processed by recording data at intervals throughout the 24 hours of the day. The 24 hours of the day are divided into T consecutive time periods to obtain the combined wind and solar power and local load curves, and to determine the relationship between wind and solar power generation and local load power in different time periods. Then, the I-WSO algorithm is used to optimize the objective function under the constraints, and the day-ahead scheduling plans for the two energy storage methods, namely battery and hydrogen production, are obtained.

3. The economic dispatch method for a wind-solar-hydrogen-storage system according to claim 2, characterized in that, Data is recorded every 10-20 minutes.

4. The economic dispatch method for a wind-solar-hydrogen-storage system according to claim 2, characterized in that, The mathematical model of the new energy system includes the following: Step 1-1: Establish the equivalent mathematical model of the power generation unit: Step 1-1-1: Establish a wind turbine model with output power P. Wind (t) is the expression in equation (1): In the formula, v(t) is the wind speed during time period t; V in Cut-in wind speed; V out To cut off the wind speed; V ra Rated wind speed; P ra This refers to the rated power of the fan; Step 1-1-2: Establish a photovoltaic cell model with output power P. PV (t) is Equation (2): In the formula, P st Photovoltaic cell output power under standard test conditions; K D For photovoltaic cells, the derating factor; G act (t) represents the actual light intensity during time period t; G st Light intensity under standard test conditions; α is the power temperature coefficient; T w The operating temperature of the solar panel; T r For reference temperature; Step 1-2: Establish the equivalent mathematical model of the energy storage unit: Step 1-2-1: Establish a lithium iron phosphate battery model, with its remaining capacity C. soc (t) is the expression in equation (3): In the formula, ε is the battery self-discharge efficiency; η - For battery discharge efficiency; η + Battery charging efficiency; P SB (t) represents the battery interaction power during time period t, where P SB (t)<0 indicates discharge, P SB (t)>0 indicates charging; Step 1-2-2: Establish the hydrogen energy storage unit model, and the input power is expressed as equation (5): P cell =V cell I cell (5) I cell =i den S cell (7) In the formula, V cell This is the input voltage for the alkaline electrolytic cell; I cell The input current for the alkaline electrolytic cell; r1, r2, s, q1, q2, q3 are empirical coefficients; A is the electrode area; T is the operating temperature; i den S represents the current density. cell V represents the electrode area; rev The potential is reversible and is taken as 1.23V under standard conditions.

5. The economic dispatch method for a wind-solar-hydrogen-storage system according to claim 2, characterized in that, The objective function is given by equation (8): F=min(C om -C ope +C loss ) (8) In the formula, F represents the total cost of the energy storage and hydrogen production equipment; C om For operating costs; C ope For operating revenue; C loss For load power shortage penalty; T is the number of time periods in a day; D wind D PV D sb D cell The unit power operation and maintenance costs are respectively for wind turbines, photovoltaic cells, lithium iron phosphate batteries, and alkaline electrolyzers; C is the unit hydrogen price. q represents the volume of hydrogen produced during time period t; s P is the penalty factor. g (t) represents the power generation during time period t; P s (t) represents the load power during time period t; P SB (t) represents the battery interaction power during time period t; P PV (t) represents the output power of the photovoltaic cell; P Wind (t) Output power of wind turbine generator; P cell Provide power input to the hydrogen energy storage unit; The limiting conditions are as follows: The power balance limit is set as in equation (12): P Wind (t)+P PV (t)=P SB (t)+P cell (t)+P s (t) (12) The interaction power limit of the lithium iron phosphate battery is set as Equation (13), the state of charge limit is set as Equation (14), and the cycle operation limit is set as Equation (15): C soc (0)=C soc (T) (15) In the formula, These represent the lower and upper limits of the interaction power of lithium iron phosphate batteries; SOC min SOC max These represent the lower and upper limits of the state of charge (SOC) of lithium iron phosphate batteries; C s For lithium iron phosphate battery capacity; C soc (t) represents the remaining charge of the lithium iron phosphate battery; the operating limits of the electrolyzer are set as equations (16) and (17): In the formula, These are the minimum and maximum power requirements for operating an alkaline electrolyzer. This represents the maximum ramp rate for an alkaline electrolytic cell.

6. The economic dispatch method for a wind-solar-hydrogen-storage system according to claim 2, characterized in that, The improved WSO algorithm specifically includes the following steps: Step 4-1: Initialize algorithm parameters, including maximum number of iterations I, population size N, variable space dimension D, and upper bound of variables x. max Lower bound x min ; Step 4-2: Improve the initialization strategy of the WSO algorithm; Step 4-2-1: Randomly generate a value within (0,1) and denote it as z1, and denote d = 1; Step 4-2-2: The Tent mapping expression is equation (18): In the formula, α is the number of particles in the chaotic sequence, and rand(0,1) is a random number in (0,1); Step 4-2-3: Obtain the chaotic variable z from the Tent mapping expression through Bernoulli transformation. d+1 As shown in equation (19): Step 4-2-4: d = d + 1; Determine whether d reaches the variable space dimension D. If yes, save the generated D-dimensional chaotic sequence; otherwise, return to step 4-2-2 and recalculate the chaotic variable of the next dimension until a D-dimensional chaotic sequence is generated. Step 4-2-5: Generate N D-dimensional chaotic sequences Z1,…,Z using steps 4-2-1 to 4-2-4. N Z is a sequence composed of z, which is the chaotic variable. The initial position of the population is generated by Z using equation (20), and it is described in matrix form as shown in equation (21): X i =x min +(x max -x min )Z i (20), Step 4-3: Calculate the global optimal solution X best Let k = 1; Step 4-4: Calculate the individual's movement speed toward the prey, and update the formula as shown in (22): In the formula, i = 1, 2, ..., N; For the new velocity vector; X is the current velocity vector; best,k This represents the optimal position obtained in the kth generation. X is the current position vector; best The global optimal position; c1 and c2 are random numbers within (0,1); p1 and p2 represent the strength of the individual white shark; μ is the suggestion contraction factor; k is the current iteration number, k = 1, 2, ..., I; p min p max Let p1 and p2 be the minimum and maximum values, respectively. min =0.5, p2=1.5; Y is the acceleration coefficient, Y=4.125; Steps 4-5: Update to the location of the best prey; when a white shark hears the sound of waves caused by the movement of prey or smells its scent, its location will be updated as follows: In the formula, For the new individual position; A negation operator; a and b are binary vectors; X O f represents a logical vector; f represents the fluctuation frequency; mv represents the kinetic force; f min f max f represents the minimum and maximum oscillation frequencies. min =0.07, f max =0.75; a0 and a1 represent two positive numbers governing exploration and development activities; Steps 4-6: Update the global optimal solution; Steps 4-7: Introducing the Lévy Flight Strategy; To enhance the algorithm's global optimization capability, the Lévy flight strategy is introduced. To reduce the computational load and ensure the speed of optimization, the following strategy is executed when rand(0,1)>0.5, otherwise... Steps 4-8: Update the optimal value for generation k+1. Since the Levy flight strategy causes individuals to move, but the quality of the new position is unknown, update the optimal value for generation k+1 according to the following formula: In the formula, F(X) best,k+1 ) is X best,k+1 The value obtained by substituting it into the objective function; for The value obtained by substituting it into the objective function; Steps 4-9: Update to the position of the best individual; when r3 s When i=1, the white shark individual will move towards the optimal individual position and make the following update:​ In the formula, The position to be updated; sgn(r2-0.5) is used to change the search direction; r1, r2, and r3 are random numbers within (0,1); a1 represents the distance between the prey and the great white shark; a2 is a positive number controlling exploration and development, a2 = 0.0005; Steps 4-10: Update position based on fish behavior: when r3 s When i>1, the white shark individual will not only update its position to the best individual according to equation (35), but also update its position according to the fish school behavior according to equation (38):​ Step 4-11: Introduce a random walk strategy to enhance the algorithm's local optimization ability, and calculate X. best,k+1 And perform a random walk on it: X′ best,k+1 =X best,k+1 +ε(X k+1,g -X k+1,w (40) In the formula, X′ best,k+1 X represents the optimal individual in generation k+1 after introducing the random walk strategy; ε is the scaling factor, ε ~ U(0,1); k+1,g X k+1,w There are two random solutions for the k+1 generation; Step 4-12: Update the global optimal solution and the k+1 generation optimal solution: Step 4-13: k = k + 1, check if the maximum number of iterations I has been reached. If yes, exit the loop and execute step 4-14; otherwise, return to step 4-4 and update the position of the best prey. Step 4-14: Output X best 、F(X best ); Steps 4-15: Process the daytime wind, solar and load data, divide the 24 hours of a day into 96 time periods, record the data every 15 minutes, and obtain wind power generation, photovoltaic power generation and load curves; Step 4-16: Optimize the objective function using the I-WSO algorithm under the constraints; input wind, solar, and load data, substitute the objective function into the I-WSO algorithm, and iteratively obtain the optimal cost and the day-ahead scheduling plan for lithium iron phosphate batteries and alkaline electrolyzers.

Images