Optimal control method for hydrogen-electricity coupling system

Abstract

The invention belongs to the technical field of optimal control of a pressure delay osmotic energy-photovoltaic-solid oxide fuel cell and wind energy hybrid energy system, and particularly relates to an optimal control method for a hydrogen-electricity coupling system based on improved giant armadillo optimization. Based on a giant armadio optimization algorithm and a triangular topology aggregation optimizer, a key parameter, namely a random solution, is introduced in an exploration stage, an optimal solution is removed, the influence of a current solution on position updating is reduced, the convergence speed and the global search capability of the algorithm are improved, the solving quality is greatly improved, and the algorithm is more effective than a classical GAO algorithm, and has a wide application prospect. The real-time tracking and accuracy based on the IGAO algorithm are improved, and when the MPPT control problem is solved, the method has higher calculation accuracy and shorter operation time, can be used for a pressure delay osmotic energy-photovoltaic-solid oxide fuel cell and wind energy hybrid energy system to obtain higher power generation, and is beneficial to the development of the energy field.

Description

Technical Field

[0001] The invention belongs to the technical field of optimization control of a pressure-delayed osmotic energy-photovoltaic-solid oxide fuel cell and wind energy hybrid energy system, and in particular relates to an optimal control method for a hydrogen-electricity coupling system based on improved giant armadillo optimization. Background Art

[0002] Maximum power point tracking (MPPT) is a technology used to improve the efficiency of energy systems, especially in photovoltaic, pressure delayed penetration and other power generation systems. Its goal is to maximize the output power of the energy system by adjusting its operating point, thereby obtaining the maximum energy conversion efficiency. Traditional MPPT control algorithms are the perturbation and observation method and the incremental modified pulsation method, but these two algorithms will produce large oscillations near the maximum power point, resulting in large power loss and instability. At present, researchers are focusing on the development and improvement of MPPT control algorithms, and the newly developed MPPT algorithms have been applied to some energy systems, and have made great progress in power improvement and response speed characteristics. Metaheuristic algorithms are widely regarded as one of the most effective and famous random methods and are widely used to solve many optimization problems. When solving the maximum power point tracking problem of energy systems, metaheuristic optimization algorithms usually have strong global search capabilities, can search for optimal solutions in a large range, and help to jump out of local optimal solutions. Engineering practice has proved that metaheuristic optimization algorithms can usually adapt to the time-varying environmental conditions of different energy systems and have certain robustness and generalization capabilities. Meta-heuristic algorithms commonly used in engineering include genetic optimization algorithm (GA), grasshopper optimization algorithm (GOA) and giant armadillo optimization algorithm (GAO). Among them, the GA algorithm has a strong global search capability, can jump out of the local optimal solution, and has good parallelism. However, the calculation process of the GA algorithm is complex, the calculation time is long, and it involves the setting of many parameters, and the parameter adjustment is difficult. The classic triangular topology aggregation optimizer (TTAO) has accurate local search capabilities and fast local search speed, but its overall calculation complexity is large, the calculation time is lengthy, it is difficult to meet real-time computing capabilities, and it is difficult to apply in actual engineering. The GAO algorithm has a shorter calculation time and better local development capabilities, and it has advantages in both calculation time and calculation accuracy. However, the classic GAO algorithm over-considers the current solution and the optimal solution when performing global optimization, which makes it difficult to conduct a global search, making it easy to fall into the local optimal problem. In addition, each update requires the calculation of the optimal solution position, which has a certain complexity, a long calculation time, and low calculation efficiency, making it difficult to meet the real-time requirements of the system.

[0003] The giant armadillo algorithm is designed with two stages, approaching the anthill and attacking the anthill, corresponding to the exploration stage and the development stage respectively. The next position of the individual in its global search algorithm is updated according to the optimal solution and the current solution. The optimal solution and the current solution have a great influence on the update of the position, which makes the algorithm easy to fall into the local optimal problem, and it will prolong the calculation time when searching for the optimal solution. Although the triangle topology aggregation optimizer is lengthy in overall calculation time and difficult to apply in engineering practice, its global search capability is strong and has strong randomness. It does not consider the optimal solution as the basic parameter, but uses random individuals to participate in the position update. Based on this, the present invention transforms the giant armadillo algorithm. By combining the local search capability of the triangle topology aggregation optimizer, the performance of the algorithm is improved, so that the MPPT control method can obtain better dynamic performance. Summary of the invention

[0004] The purpose of the present invention is to avoid the shortcomings of the prior art and provide an optimal control method for a hydrogen-electric coupling system. By modifying the giant armadillo algorithm and combining the advantages of other strategies, an optimal control algorithm for a pressure-delayed osmotic energy-photovoltaic-solid oxide fuel cell and wind energy hybrid system is obtained to solve the problems existing in the prior art.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is an optimal control method for a hydrogen-electric coupling system, which is characterized by being used for maximum power tracking optimization of a photovoltaic subsystem, a wind energy subsystem, a pressure-delayed osmotic energy subsystem, and a solid oxide fuel cell subsystem, and comprises the following steps:

[0006] Step 1: Initialization, input the real-time ambient temperature and solar irradiance of the photovoltaic subsystem, the wind speed of the wind energy subsystem, the solution temperature, solution concentration (salt concentration) and inlet flow of the pressure-delayed osmotic energy subsystem, the operating temperature and hydrogen content of the solid oxide fuel cell subsystem; the current I-voltage V characteristic curves of the photovoltaic subsystem, the pressure-delayed osmotic energy subsystem and the solid oxide fuel cell subsystem, and the wind energy capture power Pwt and wind wheel speed ωn characteristic curve of the wind energy subsystem; set the population size N and the maximum number of iterations T of the improved algorithm;

[0007] Step 2: Create an initial population and generate i random variable parameters by calculation:

[0008] x i =lb i +rand×(ub i -lb i ); rand∈[0,1] (1)

[0009] Among them, x irefers to the i-th parameter value, i ranges from 1 to the population size N, lb i is the lower bound of the parameter, ub i is the upper bound of the parameter, rand represents a random number ranging from 0 to 1; x i,j is the jth dimension of the ith parameter value. In this hybrid energy system, the dimension of the parameter value is 1, so by default j is only 1. i,j That is x i ; For the photovoltaic subsystem, pressure-delayed osmotic energy subsystem and solid oxide fuel cell subsystem, the parameter x i is the current I; for the wind energy subsystem, the parameter x i is the wind wheel speed ωn;

[0010] Evaluate the initialized population, calculate the fitness of all individuals in the population, and calculate the fitness F of different individuals according to the current I-voltage V characteristic curves of the photovoltaic subsystem, the pressure delayed osmotic energy subsystem and the solid oxide fuel cell subsystem. The formula is:

[0011] F=I×V (2)

[0012] The individual fitness of the wind energy subsystem is the wind energy capture power Pwt corresponding to the current wind rotor speed ωn;

[0013] Step 3: Update position in the exploration phase, which means that in the exploration phase, the position update formula of the triangle topology aggregation optimizer is introduced to replace the original exploration phase formula, which improves the global exploration ability and algorithm accuracy of the algorithm and reduces the calculation time;

[0014] Furthermore, the updating of the position in the exploration phase refers to replacing the classical formula (3) and the classical formula (4) with the introduced formula (5). The classical formula (3) and the classical formula (4) refer to the design of the giant armadillo algorithm, in which for each population member representing the giant armadillo, the positions of other population members with better objective function values ​​are considered as termite mounds; the candidate termite mound set of each member in the population is specified using equation (3);

[0015] TM i ={X k :F k <F i andk≠i}, where i=1, 2,...,N and k∈{1, 2,...,N} (3)

[0016] Among them, TM i is the set of candidate termite mound locations for the i-th giant armadillo, X k is the population member whose objective function value is better than the i-th giant armadillo, F k is the value of its objective function;

[0017] The giant armadillo randomly selects a candidate termite mound and attacks it; based on modeling the movement of the giant armadillo toward the termite mound, the new position of each population member is calculated using formula (4),

[0018]

[0019] here, For the i-th individual in the exploration phase, STM i is the termite mound selected for the i-th giant armadillo, STM i,j is its jth dimension, x P1 i is the new position of the ith giant armadillo calculated according to the attack phase of the proposed giant armadillo algorithm, and x P1 i,j is its jth dimension, r i,j is a random number in the interval [0,1], I i,j is a randomly selected number 1 or 2;

[0020] The position update formula of the introduced triangle topology aggregation optimizer is:

[0021]

[0022] in, is the updated position of the jth dimension of the ith individual in the exploration phase; r is a random number between 0 and 1; x i，j is the position of the jth dimension of the i-th individual before the update; m is a random integer from 1 to the number of populations; x m，j is the position of the jth dimension of a random individual in the population;

[0023] In the exploration phase, the introduced formula (5) is used to replace the classical formula (3) and formula (4), and the variable parameters x_i of the photovoltaic subsystem, wind energy subsystem, pressure delayed osmotic energy subsystem, and solid oxide fuel cell subsystem are updated through formula (5).

[0024] Step 4: Selecting a better location means updating the pre-variable parameter x in the photovoltaic subsystem, wind energy subsystem, pressure-delayed osmotic energy subsystem, and solid oxide fuel cell subsystem. i The fitness and updated variable parameter x i The fitness of the variables is compared and the variable parameter x with the larger fitness value is selected. i , and retain it;

[0025] Step 5: Development phase, using the original update strategy, means using the giant armadillo’s original development phase algorithm to update the position;

[0026] Step 6: Selecting a better position again means updating the pre-variable parameter x in the photovoltaic subsystem, wind energy subsystem, pressure-delayed osmotic energy subsystem, and solid oxide fuel cell subsystem. i The fitness and updated variable parameter x i The fitness of the variables is compared, and the variable parameters with larger corresponding fitness values ​​are selected and retained;

[0027] Step 7: If the algorithm has reached the maximum number of iterations T set for the improved algorithm, the final result is output and the iteration ends; otherwise, go to step 3 and continue searching; finally, in the photovoltaic subsystem, wind energy subsystem, pressure delayed osmotic energy subsystem, and solid oxide fuel cell subsystem, the four variable parameters x with the largest corresponding fitness values ​​are obtained. i They are recorded as four optimal solutions respectively. The optimal solution is the power point obtained by maximum power tracking of different systems, and the corresponding fitness value is the maximum power.

[0028] The exploration phase in the third step of updating the position over-utilizes the optimal solution and the current solution, making it difficult for the algorithm to search globally, and the need to calculate the optimal solution prolongs the running time; the improved exploration phase no longer uses the optimal solution, reduces the impact of the current solution on the calculation, and introduces random individuals; it has better global search capabilities, reduces the convergence time of the algorithm, and ultimately improves the computational efficiency of the algorithm and thus improves the performance of the maximum power point tracking system.

[0029] The beneficial effects of the present invention are as follows: the present invention provides an optimal control method for a hydrogen-electric coupling system, which is based on a giant armadillo algorithm and a triangular topology aggregation optimizer. In the exploration stage, a key parameter of a random solution is introduced, while the optimal solution is removed and the influence of the current solution on the position update is reduced, thereby improving the convergence speed and global search capability of the algorithm, greatly improving the solution quality, and being more effective than the classic GAO algorithm. The real-time tracking and accuracy based on the Improved Giant Armadillo Optimization (IGAO) algorithm are improved, and when solving the MPPT control problem, the method has higher calculation accuracy and faster running time, and can be used for pressure-delayed osmotic energy-photovoltaic-solid oxide fuel cell and wind energy hybrid energy systems to obtain greater power generation, which is beneficial to the development of the energy field. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 is a flow chart of the present invention;

[0031] Figure 2 is a schematic diagram of the change in the environment of the photovoltaic module when the present invention is implemented;

[0032] Figure 3is a schematic diagram of the change of the pressure-delayed permeation environment when the present invention is implemented;

[0033] Figure 4 is a schematic diagram of the environmental changes of a solid oxide fuel cell when the present invention is implemented;

[0034] Figure 5 It is a schematic diagram of the change of the wind turbine environment when the present invention is implemented. DETAILED DESCRIPTION

[0035] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0036] In order to achieve the above-mentioned purpose, an optimal control method for a hydrogen-electric coupling system is provided. Figure 1 As shown, the invention is characterized by including a photovoltaic subsystem, a wind energy subsystem, a pressure-delayed osmotic energy subsystem, and a maximum power tracking optimization of a solid oxide fuel cell subsystem, and includes the following steps:

[0037] Step 1: Initialization, input the real-time ambient temperature and solar irradiance of the photovoltaic subsystem, the wind speed of the wind energy subsystem, the solution temperature, solution concentration (salt concentration) and inlet flow of the pressure-delayed osmotic energy subsystem, the operating temperature and hydrogen content of the solid oxide fuel cell subsystem; the current I-voltage V characteristic curves of the wind energy subsystem, the pressure-delayed osmotic energy subsystem and the solid oxide fuel cell subsystem, and the wind energy capture power Pwt and wind wheel speed ωn characteristic curve of the wind energy subsystem; set the population size N and the maximum number of iterations T of the improved algorithm;

[0038] Step 2: Create an initial population and generate i random variable parameters by calculation:

[0039] x i =lb i +rand×(ub i -lb i ); rand∈[0,1] (1)

[0040] Among them, x i refers to the i-th parameter value, i ranges from 1 to the population size N, lb i is the lower bound of the parameter, ub i is the upper bound of the parameter, rand represents a random number ranging from 0 to 1; x i,j is the jth dimension of the ith parameter value. In this hybrid energy system, the dimension of the parameter value is 1, so by default j is only 1. i,j That is x i ; For the photovoltaic subsystem, pressure-delayed osmotic energy subsystem and solid oxide fuel cell subsystem, the parameter xi is the current I; for the wind energy subsystem, the parameter x i is the wind wheel speed ωn;

[0041] Evaluate the initialized population, calculate the fitness of all individuals in the population, and calculate the fitness F of different individuals according to the current I-voltage V characteristic curves of the photovoltaic subsystem, the pressure delayed osmotic energy subsystem and the solid oxide fuel cell subsystem. The formula is:

[0042] F=I×V (2)

[0043] The individual fitness of the wind energy subsystem is the wind energy capture power Pwt corresponding to the current wind rotor speed ωn;

[0044] Step 3: Update position in the exploration phase, which means that in the exploration phase, the position update formula of the triangle topology aggregation optimizer is introduced to replace the original exploration phase formula, which improves the global exploration ability and algorithm accuracy of the algorithm and reduces the calculation time;

[0045] Furthermore, the updating of the position in the exploration phase refers to replacing the classical formula (3) and the classical formula (4) with the introduced formula (5). The classical formula (3) and the classical formula (4) refer to the design of the giant armadillo algorithm, in which for each population member representing the giant armadillo, the positions of other population members with better objective function values ​​are considered as termite mounds; the candidate termite mound set of each member in the population is specified using equation (3);

[0046] TM i ={X k :F k <F i andk≠i}, where i=1, 2,...,N an dk∈{1, 2,...,N} (3)

[0047] Among them, TM i is the set of candidate termite mound locations for the i-th giant armadillo, X k is the population member whose objective function value is better than the i-th giant armadillo, F k is the value of its objective function;

[0048] The giant armadillo randomly selects a candidate termite mound and attacks it; based on modeling the movement of the giant armadillo toward the termite mound, the new position of each population member is calculated using equation (4),

[0049]

[0050] here, For the i-th individual in the exploration phase, STM i is the termite mound selected for the i-th giant armadillo, STM i,jis its jth dimension, x P1 i is the new position of the ith giant armadillo calculated according to the attack phase of the proposed giant armadillo algorithm, and x P1 i,j is its jth dimension, r i,j is a random number in the interval [0,1], I i,j is a randomly selected number 1 or 2;

[0051] The position update formula of the introduced triangle topology aggregation optimizer is:

[0052]

[0053] in, is the updated position of the jth dimension of the ith individual in the exploration phase; r is a random number between 0 and 1; x i，j is the position of the jth dimension of the i-th individual before the update; m is a random integer from 1 to the number of populations; x m，j is the position of the jth dimension of a random individual in the population;

[0054] In the exploration phase, the introduced formula (5) is used to replace the classical formula (3) and formula (4), and the variable parameters x_i of the photovoltaic subsystem, wind energy subsystem, pressure delayed osmotic energy subsystem, and solid oxide fuel cell subsystem are updated through formula (5).

[0055] Step 4: Selecting a better location means updating the pre-variable parameter x in the photovoltaic subsystem, wind energy subsystem, pressure-delayed osmotic energy subsystem, and solid oxide fuel cell subsystem. i The fitness and updated variable parameter x i The fitness of the variables is compared and the variable parameter x with the larger fitness value is selected. i , and retain it;

[0056] Step 5: Development phase, using the original update strategy, means using the giant armadillo’s original development phase algorithm to update the position;

[0057] Step 6: Selecting a better position again means updating the pre-variable parameter x in the photovoltaic subsystem, wind energy subsystem, pressure-delayed osmotic energy subsystem, and solid oxide fuel cell subsystem. i The fitness and updated variable parameter x i The fitness of the variables is compared, and the variable parameters with larger corresponding fitness values ​​are selected and retained;

[0058] Step 7: If the algorithm has reached the maximum number of iterations T set for the improved algorithm, the final result is output and the iteration ends; otherwise, go to step 3 and continue searching; finally, in the photovoltaic subsystem, wind energy subsystem, pressure delayed osmotic energy subsystem, and solid oxide fuel cell subsystem, the four variable parameters x with the largest corresponding fitness values ​​are obtained. i They are recorded as four optimal solutions respectively. The optimal solution is the power point obtained by maximum power tracking of different systems, and the corresponding fitness value is the maximum power.

[0059] The exploration phase in the third step of updating the position over-utilizes the optimal solution and the current solution, making it difficult for the algorithm to search globally, and the need to calculate the optimal solution prolongs the running time; the improved exploration phase no longer uses the optimal solution, reduces the impact of the current solution on the calculation, and introduces random individuals; it has better global search capabilities, reduces the convergence time of the algorithm, and ultimately improves the computational efficiency of the algorithm and thus improves the performance of the maximum power point tracking system.

[0060] The present invention provides the following specific implementation methods to verify the effect of the present invention: The optimal control method for the hydrogen-electric coupling system described above firstly performs the initial setting of the algorithm parameters. To ensure the fairness of the experiment, it is first necessary to determine the optimal parameter settings of each algorithm for the problem. The population size of the genetic optimization algorithm (GA), the locust optimization algorithm (GOA), the giant armadillo algorithm (GAO), and the improved giant armadillo algorithm (IGAO) are all 5, the maximum number of iterations is 30, and the remaining algorithm internal parameters are kept at the default values;

[0061] like Figure 2 The optimization control method of the present invention is applied to the operation of the photovoltaic subsystem in a scene where the irradiance and temperature change. Figure 2 In (a), the horizontal axis represents three different scenes, and the vertical axis represents the solar irradiance value under different scenes; Figure 2 (b) The horizontal axis represents three different scenes, and the vertical axis represents the ambient temperature under different scenes. Under different scene conditions, the solar irradiance is 900W / m 2 , 800W / m 2 and 700W / m 2 At the same time, the ambient temperature is 20℃, 30℃ and 40℃ respectively. Under this scenario condition, various algorithms are compared and tested.

[0062] like Figure 3As shown, the optimization control method of the present invention is applied to the operation of the pressure-delayed osmotic energy subsystem in a scenario where the temperature and salinity change. Figure 3 In (a), the horizontal axis represents three different scenarios, and the vertical axis represents the solution temperature under different scenarios. Figure 3 (b) The horizontal axis represents three different scenarios, and the vertical axis represents the solution flow rate under different scenarios. Figure 3 In (c), the horizontal axis represents three different scenarios, and the vertical axis represents the solution concentration in different scenarios. The solution concentration is the salt concentration. In different scenarios, the solution temperature is 20℃, 30℃ and 40℃, the solution flow rate is 0.3kg / h, 0.4kg / h and 0.5kg / h, and the solution concentration is 30g / kg, 35g / kg and 40g / kg. The algorithms are compared under this scenario.

[0063] like Figure 4 As shown, the optimization control method of the present invention is applied to the operation of the solid oxide fuel cell subsystem in a scenario where the temperature and hydrogen concentration change. Figure 4 In (a), the horizontal axis represents three different scenarios, and the vertical axis represents the operating temperature under different scenarios, such as Figure 4 (b) The horizontal axis represents three different scenarios, and the vertical axis represents the hydrogen concentration in different scenarios. Under different scenario conditions, the operating temperatures are 1073K, 1123K, and 1173K, respectively. The hydrogen concentrations are 60%, 70%, and 80%, respectively. A comparative test is conducted on each algorithm under this scenario condition.

[0064] like Figure 5 As shown, the optimization control method of the present invention is applied to the operation of the wind energy subsystem in a scene with varying wind speed. Figure 5 The horizontal axis represents three different scenarios, and the vertical axis represents the wind speed in different scenarios. Under the scenario conditions, the wind speeds are 8m / s, 10m / s and 12m / s respectively. Under the scenario conditions, various algorithms are compared and tested.

[0065] The present invention uses the following control performance indicators when evaluating the MPPT control performance experiment based on the IGAO algorithm:

[0066] The average power output (APO) is used as a significant factor to evaluate the energy extraction of the proposed method. To reduce the random error of the operation process, all methods were executed ten times in the test.

[0067]

[0068] Where ε is the number of executions (set to 10), is the fitness function of the designed problem, represents the best fitness obtained by each technology in the i-th run.

[0069] Average response time (ART): It is a significant factor in evaluating the response time of the proposed method. Its mathematical expression is:

[0070]

[0071] Where T ε Indicates the εth running time.

[0072] Average performance boost (APB): The ratio of APO to ART, which evaluates the amount of power extracted per unit time. Its mathematical expression is as follows:

[0073]

[0074] This paper proposes an MPPT control method based on the IGAO algorithm and conducts comparative tests with MPPT control methods based on other metaheuristic algorithms under various working conditions to verify its reliability. In this process, simulation tests were carried out under four systems, including photovoltaic subsystem, pressure-delayed osmotic energy subsystem, solid oxide fuel cell subsystem and wind energy subsystem. Table 1 lists the comparison results of several methods of the photovoltaic subsystem under temperature and irradiance change scenarios, Table 2 lists the comparison results of several methods of the pressure-delayed osmotic energy subsystem under temperature, solution concentration and solution flow rate change scenarios, Table 3 lists the comparison results of several methods of the solid oxide fuel cell subsystem under temperature and hydrogen concentration change scenarios, Table 4 lists the comparison results of several methods of the wind energy subsystem under wind speed change scenarios, and the APO, ART and APB values ​​show that the MPPT control method based on the IGAO algorithm has good tracking performance.

[0075] Table 1. Comparison results of MPPT control methods based on different algorithms for photovoltaic subsystems under different environments

[0076]

[0077] Table 2. Comparison results of MPPT control methods based on different algorithms for pressure-delayed osmotic energy subsystem under different environments

[0078]

[0079]

[0080] Table 3. Comparison results of MPPT control methods based on different algorithms for solid oxide fuel cell subsystems under different environments

[0081]

[0082] Table 4. Comparison results of MPPT control methods based on different algorithms for wind energy subsystems in different environments

[0083]

[0084]

[0085] When solving the MPPT control problem, the IGAO algorithm is compared and analyzed with the GAO, GA, and GOA metaheuristic algorithms, which can meet the requirements of power calculation accuracy and running time at the same time. The proposed IGAO algorithm is evaluated using the performance indicators of the optimization design problem, including average power output (APO), average response time (ART), and average performance boost (APB). The impact of different operating environments on the energy system is fully considered, and four different scenarios are proposed to test the algorithm, which verifies that the algorithm has a good application effect on the MPPT control method.

[0086] The maximum power point tracking control method proposed in this paper was tested under different working conditions using four energy subsystems, and the MPPT control methods based on IGAO and other metaheuristic algorithms were compared and analyzed using the three performance indicators of APO, ART and APB to verify its reliability. The MPPT control method based on the IGAO algorithm performed best under the conditions of temperature and irradiance changes in the photovoltaic subsystem, and obtained the maximum APO, APB and the minimum ART under all conditions. Compared with the MPPT control method based on the classic GAO algorithm, APO and APB were increased by 11.9% and 17.3% respectively, and ART was reduced by 33.9%. In the scenario test of temperature and solution changes in the pressure-delayed osmotic energy subsystem, the IGAO algorithm had the best tracking effect, and obtained the maximum APO, APB and the minimum ART under all conditions. Compared with the MPPT control method based on the classic GAO algorithm, APO and APB were increased by 0.064% and 0.334% respectively, and ART was reduced by 0.33%. In the solid oxide fuel cell subsystem, the MPPT control method based on the IGAO algorithm has the best performance under the conditions of temperature and hydrogen concentration changes, and obtains the maximum APO, APB and minimum ART under all conditions. Compared with the MPPT control method based on the classic GAO algorithm, APO and APB are increased by 0.019% and 0.155% respectively, and ART is reduced by 0.16%. In the wind energy subsystem, the MPPT control method based on the IGAO algorithm has the best performance under the conditions of wind speed changes, and obtains the maximum APO, APB and minimum ART under all conditions. Compared with the MPPT control method based on the classic GAO algorithm, APO and APB are increased by 0.73% and 61.7% respectively, and ART is reduced by 36.4%.

[0087] The MPPT based on the IGAO algorithm is compared with the MPPT based on the other two meta-heuristic algorithms. The experimental results show that the MPPT based on the IGAO algorithm can achieve optimal performance under all environmental conditions and converge to the optimal parameters, which is better than other algorithms, making the MPPT more real-time and accurate in tracking the maximum power of the renewable energy system. The system can be adjusted in real time under changing environmental conditions. The results show that IGAO has global optimization capabilities, high optimization accuracy and convergence speed, which greatly improves the real-time and precision requirements of MPPT control, and has obvious improvements compared with the MPPT control method based on the GAO algorithm.

[0088] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An optimal control method for a hydrogen-electricity coupling system, characterized in that: The maximum power point tracking optimization for photovoltaic subsystem, wind energy subsystem, pressure delayed osmotic energy subsystem, and solid oxide fuel cell subsystem includes the following steps: The first step is initialization. Input the real-time ambient temperature and solar irradiance of the photovoltaic subsystem, the wind speed of the wind energy subsystem, the solution temperature, solution concentration (salt concentration) and inlet flow of the pressure-delayed osmotic energy subsystem, and the operating temperature and hydrogen content of the solid oxide fuel cell subsystem; calculate the current I-voltage V characteristic curves of the photovoltaic subsystem, the pressure-delayed osmotic energy subsystem and the solid oxide fuel cell subsystem, and the wind energy capture power Pwt and wind wheel speed ωn characteristic curve of the wind energy subsystem; set the population size N and the maximum number of iterations T of the improved algorithm; The second step is to create an initial population and generate i random variable parameters by calculation: x i =lb i +rand×(ub i -lb i );rand∈[0,1] (1) Among them, x i refers to the i-th parameter value, i ranges from 1 to the population size N, lb i is the lower bound of the parameter, ub i is the upper bound of the parameter, rand represents a random number ranging from 0 to 1; x i,j is the jth dimension of the ith parameter value. In this hybrid energy system, the dimension of the parameter value is 1, so by default j is only 1. i,j That is x i ; For the photovoltaic subsystem, pressure-delayed osmotic energy subsystem and solid oxide fuel cell subsystem, the parameter x i is the current I; for the wind energy subsystem, the parameter x i is the wind wheel speed ωn; Evaluate the initialized population, calculate the fitness of all individuals in the population, and calculate the fitness F of different individuals according to the current I-voltage V characteristic curves of the photovoltaic subsystem, the pressure-delayed osmotic energy subsystem, and the solid oxide fuel cell subsystem. The formula is: F=I×V (2) The individual fitness of the wind energy subsystem is the wind energy capture power Pwt corresponding to the current wind rotor speed ωn; The third step is to update the position in the exploration phase, which means that in the exploration phase, the position update formula of the triangle topology aggregation optimizer is introduced to replace the original exploration phase formula, which improves the global exploration ability and algorithm accuracy of the algorithm and reduces the calculation time; The fourth step is to select a better position, which means that in the photovoltaic subsystem, wind energy subsystem, pressure delayed osmotic energy subsystem, and solid oxide fuel cell subsystem, the pre-update variable parameter x i The fitness and updated variable parameter x i The fitness of the variables is compared and the variable parameter x with the larger fitness value is selected. i , and retain it; Step 5: Development phase, using the original update strategy, means using the giant armadillo’s original development phase algorithm to update the position; Step 6: Selecting a better position again means updating the pre-variable parameter x in the photovoltaic subsystem, wind energy subsystem, pressure-delayed osmotic energy subsystem, and solid oxide fuel cell subsystem. i The fitness and updated variable parameter x i The fitness of the variables is compared, and the variable parameters with larger corresponding fitness values ​​are selected and retained; Step 7: If the algorithm has reached the maximum number of iterations T set for the improved algorithm, the final result is output and the iteration ends; otherwise, go to step 3 and continue searching; finally, in the photovoltaic subsystem, wind energy subsystem, pressure delayed osmotic energy subsystem, and solid oxide fuel cell subsystem, the four variable parameters x with the largest corresponding fitness values ​​are obtained. i They are recorded as four optimal solutions respectively. The optimal solution is the power point obtained by maximum power tracking of different systems, and the corresponding fitness value is the maximum power.

2. The optimal control method for a hydrogen-electricity coupling system according to claim 1, characterized in that: The updating of the position in the exploration phase in the third step refers to that in the giant armadillo algorithm design, for each population member representing the giant armadillo, the position of other population members with better objective function values ​​is considered to be a termite mound; Use formula (3) to specify the set of candidate termite mounds for each member of the population; TM i ={X k :F k <F i andk≠i},where i=1,2,...,Nandk∈{1,2,...,N} (3) Among them, TM i is the set of candidate termite mound locations for the i-th giant armadillo, X k is the population member whose objective function value is better than the i-th giant armadillo, F k is the value of its objective function; The giant armadillo randomly selects a candidate termite mound and attacks it; based on modeling the movement of the giant armadillo toward the termite mound, the new position of each population member is calculated using formula (4), here, For the i-th individual in the exploration phase, STM i is the termite mound selected for the i-th giant armadillo, STM i,j is its jth dimension, x P1 i is the new position of the ith giant armadillo calculated according to the attack phase of the proposed giant armadillo algorithm, and x P1 i,j is its jth dimension, r i,j is a random number in the interval [0,1], I i,j is a randomly selected number 1 or 2; The position update formula of the introduced triangle topology aggregation optimizer is: in, is the updated position of the jth dimension of the ith individual in the exploration phase; r is a random number between 0 and 1; x i，j is the position of the jth dimension of the i-th individual before the update; m is a random integer from 1 to the number of populations; x m，j is the position of the jth dimension of a random individual in the population; In the exploration phase, the introduced formula (5) is used to replace the classical formula (3) and formula (4), and the variable parameters x of the photovoltaic subsystem, wind energy subsystem, pressure delayed osmotic energy subsystem, and solid oxide fuel cell subsystem are i Update using formula (5).

3. The optimal control method for a hydrogen-electricity coupling system according to claim 2, characterized in that: The exploration phase updates the position by over-utilizing the optimal solution and the current solution, which makes it difficult for the algorithm to search globally, and the running time is prolonged due to the need to calculate the optimal solution. The improved exploration phase no longer uses the optimal solution, reduces the impact of the current solution on the calculation, and introduces random individuals. It has better global search capabilities, reduces the convergence time of the algorithm, and ultimately improves the calculation efficiency of the algorithm and thus improves the performance of the maximum power point tracking system.

Images