An Optimal Scheduling Method and System for Electro-Hydrogen Coupled Systems Based on a Multi-Stage Hybrid Search Algorithm

Abstract

This invention relates to an optimization scheduling method and system for an electric-hydrogen coupling system based on a multi-stage hybrid search algorithm. The method employs a high-precision nonlinear model and specifically designs a two-stage scheme that divides the scheduling solution into a multi-step variable neighborhood search algorithm and an adaptive simulated annealing algorithm. This satisfies the computational efficiency requirements of the solution, while leveraging the ability of adaptive perturbation, global optimization, and escaping local optima. It generates a high-precision optimal solution while breaking through local optima.

Description

Technical Field

[0001] This invention belongs to the field of power system dispatching technology, specifically relating to an optimized dispatching method and system for an electric-hydrogen coupled system based on a multi-stage hybrid search algorithm. Background Technology

[0002] With the increasing proportion of renewable energy generation, effectively mitigating its volatility and intermittency has become a major challenge for the stable operation of the power grid. The electro-hydrogen coupling system, which converts excess electrical energy into hydrogen energy for storage via an electrolyzer and then converts the hydrogen energy back into electrical energy via a fuel cell when needed, provides an effective technological solution to this problem.

[0003] In the optimal scheduling of electro-hydrogen coupling systems, the core lies in establishing an accurate system operation model and employing efficient optimization algorithms to solve it, aiming to minimize the system's operating cost. However, the optimal scheduling of electro-hydrogen coupling systems is a complex, high-dimensional, nonlinear optimization problem. Traditional intelligent algorithms commonly used in existing research (such as particle swarm optimization and genetic algorithms) often suffer from premature convergence and susceptibility to local optima when solving such problems, leading to unstable and low-quality solutions. While some traditional search algorithms offer better stability, they suffer from slow convergence speeds and weak global exploration capabilities, making it difficult to find the global optimum. Therefore, current technology lacks an optimization algorithm that can guarantee both accuracy and efficiency to address this challenge.

[0004] In summary, how to establish a scheduling scheme generation method that can efficiently solve the problem and avoid getting trapped in local optima, so as to obtain a better scheduling scheme with higher immediacy, is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0005] One objective of this invention is to at least solve one or more of the aforementioned problems existing in the prior art. In other words, one objective of this invention is to provide an optimized scheduling method and system for an electric-hydrogen coupling system based on a multi-stage hybrid search algorithm that satisfies one or more of the aforementioned requirements, comprising:

[0006] Based on the nonlinear operating characteristics of key equipment in the electro-hydrogen coupling system, a refined nonlinear operating model of the electro-hydrogen coupling system is established.

[0007] The objective function for optimization is set to minimize the system operating cost of the nonlinear operating model, and constraints are set accordingly.

[0008] Using the operating power of each device in the system within the scheduling period as the optimization variable, the initial scheduling scheme is generated by solving the problem using a multi-step variable neighborhood search algorithm based on the model, objective function and constraints.

[0009] Based on the initial scheduling scheme, the operating power of each device in the initial scheduling scheme is further optimized using the adaptive simulated annealing algorithm to generate the optimal scheduling scheme.

[0010] The electro-hydrogen coupling system is scheduled according to the optimal scheduling scheme.

[0011] As a preferred implementation, an initial scheduling scheme is generated using a multi-step variable neighborhood search algorithm based on the objective function and constraints, including:

[0012] a. Generate a multi-stage non-uniform step size sequence, where each stage contains several non-uniform step sizes from large to small; construct a multi-dimensional solution space with multiple coordinate axes using the operating power of each device as the coordinate axis, and set the initial solution;

[0013] b. Using the first step size of the current stage as the initial step size, perform neighborhood probing starting from the initial solution to find the direction that makes the objective function value decrease;

[0014] c. Perform dynamic step-size expansion search in the direction that makes the objective function value decrease until the objective function value no longer decreases, then restore the initial step size and switch the search direction;

[0015] d. If no solution can be improved in all directions, use the next step length of the current stage as the initial step length and repeat step bc until all step lengths of the current stage are searched.

[0016] e. Switch to the next stage and repeat step bd iteratively using the non-uniform step size in the next stage until the iteration is completed and the initial scheduling scheme is generated.

[0017] As a preferred implementation, dynamic step-size expansion search includes:

[0018] The search continues in the direction that makes the objective function value decrease. If the objective function value decreases continuously, the step size is increased to improve the search speed until the objective function value stops decreasing. Then, the initial step size is restored and the search direction is switched.

[0019] As a preferred implementation, the adaptive simulated annealing algorithm is used to further optimize the operating power of each device in the initial scheduling scheme, including:

[0020] Initialize temperature parameters;

[0021] Starting from the initial scheduling scheme, a new solution is generated using neighborhood perturbation based on the current temperature. The Metropolis acceptance criterion is used to determine whether to accept the new solution in order to optimize the scheduling scheme and escape local optima.

[0022] The process iterates step by step, gradually reducing the temperature, until the iteration terminates, generating the optimal scheduling scheme.

[0023] In one preferred embodiment, the key equipment includes an electrolyzer and a fuel cell.

[0024] As a further preferred implementation method, the process of establishing the nonlinear operation model is as follows:

[0025] Based on the coupling relationship between the working efficiency of the electrolyzer and fuel cell and the input power in the system, the efficiency curves of the electrolyzer and fuel cell are fitted into polynomials using interpolation.

[0026] A refined nonlinear operating model of the electro-hydrogen coupling system is established by using polynomials and combining them with the operating models of each device.

[0027] In a preferred implementation, the system operating cost is the difference between the total cost of purchasing electricity, heat, and hydrogen and the total revenue from selling electricity, heat, and hydrogen.

[0028] As a preferred embodiment, the constraints include:

[0029] System equipment model constraints, electricity purchase and sale constraints, heat purchase and sale power constraints, hydrogen purchase and sale power constraints, wind turbine output constraints, photovoltaic output constraints, electrical power balance constraints, thermal power balance constraints, hydrogen power balance constraints, and key equipment efficiency constraints.

[0030] As a preferred implementation method, the optimization variables include: purchased and sold electricity power, purchased and sold heat power, purchased and sold hydrogen power, and the operating power of each device.

[0031] On the other hand, the present invention also provides an optimized scheduling system for an electro-hydrogen coupling system based on a multi-stage hybrid search algorithm, comprising:

[0032] The model generation module is used to establish a refined nonlinear operating model of the electro-hydrogen coupling system based on the nonlinear operating characteristics of key equipment in the system.

[0033] The initialization module is used to set the objective function for optimization with the goal of minimizing the system operating cost of the nonlinear operating model, and to set the constraints.

[0034] The multi-step variable neighborhood search module is used to solve the initial scheduling scheme by taking the operating power of each device in the system within the scheduling cycle as the optimization variable, based on the model, objective function and constraints, and using the multi-step variable neighborhood search algorithm.

[0035] The adaptive simulated annealing module is used to further optimize the operating power of each device in the initial scheduling scheme using the adaptive simulated annealing algorithm, based on the initial scheduling scheme, to generate the optimal scheduling scheme.

[0036] The scheduling execution module is used to schedule the electro-hydrogen coupling system according to the optimal scheduling scheme.

[0037] Compared with existing technologies, the electro-hydrogen coupling system optimization scheduling method and system based on a multi-stage hybrid search algorithm provided by this invention have the following beneficial effects:

[0038] The method and system of this invention employ high-precision nonlinear models at key equipment points in the system operation model, improving scheduling accuracy from the root of modeling and ensuring that the physical basis of the optimization solution is highly consistent with actual operating conditions. To address the massive computational load brought about by the high-precision nonlinear model, this invention provides a multi-stage hybrid search algorithm. It specifically designs a two-stage scheme dividing the scheduling solution into a multi-step variable neighborhood search algorithm and an adaptive simulated annealing algorithm. The multi-stage non-uniform step size mechanism of the multi-step variable neighborhood search algorithm enables large-scale, structured exploration in a vast solution space filled with local optima traps, effectively traversing numerous inferior local optima regions, avoiding the blindness of early searches, and quickly converging to a high-quality initial solution, perfectly meeting the computational efficiency requirements of the solution. Based on the initial scheduling scheme, the second stage employs an adaptive simulated annealing algorithm, leveraging its adaptive perturbation and global optimization capabilities to escape local optima, generating a high-precision optimal solution while breaking through local optima. Attached Figure Description

[0039] Figure 1 This is a flowchart of the optimized scheduling method for an electro-hydrogen coupling system based on a multi-stage hybrid search algorithm according to the present invention;

[0040] Figure 2 This is a schematic diagram of the system framework of the electro-hydrogen coupling system of the present invention;

[0041] Figure 3a This is a schematic diagram of the fitting curve of the electrolytic cell of the present invention;

[0042] Figure 3b This is a schematic diagram of the fitting curve of the fuel cell of the present invention;

[0043] Figure 4 This is a diagram comparing the convergence results of different algorithms when the scheduling is solved multiple times;

[0044] Figure 5 This is a schematic diagram of the power balance of the optimal scheduling scheme of the method of the present invention;

[0045] Figure 6 This is a schematic diagram of the thermal power balance of the optimal scheduling scheme of the method of the present invention;

[0046] Figure 7 This is a schematic diagram of the hydrogen power balance of the optimal scheduling scheme of the method of the present invention;

[0047] Figure 8This is a schematic diagram of the structure of the electro-hydrogen coupling system optimization scheduling system based on a multi-stage hybrid search algorithm of the present invention. Detailed Implementation

[0048] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0049] The following description provides examples and does not limit the scope, applicability, or examples set forth in the claims. Changes may be made to the function and arrangement of the described elements without departing from the scope of the invention. Various processes or components may be appropriately omitted, substituted, or added to the various examples. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Furthermore, features described with respect to some examples may be combined into other examples.

[0050] Embodiments of the present invention provide an optimized scheduling method for an electro-hydrogen coupling system based on a multi-stage hybrid search algorithm, the flowchart of which is shown below. Figure 1 As shown, it includes:

[0051] S1. Based on the nonlinear operating characteristics of key equipment in the electro-hydrogen coupling system, establish a refined nonlinear operating model of the electro-hydrogen coupling system.

[0052] The system framework diagram of the electro-hydrogen coupling system is shown below. Figure 2 As shown, it includes wind turbines, photovoltaics, loads, batteries, electric boilers, thermal storage tanks, electrolyzers, hydrogen storage tanks, and fuel cells.

[0053] Key components of an electro-hydrogen coupling system include the electrolyzer and the fuel cell. In actual operation, the relationship between the efficiency of the electrolyzer and the fuel cell and the input power is nonlinear, making it difficult to solve using a solver. Therefore, existing research often employs constant power or piecewise linearization methods for simplification. In this invention, to more accurately optimize the scheduling of the electro-hydrogen coupling system, an interpolation fitting method is used to fit the efficiency curves of the electrolyzer and the fuel cell into polynomials. Using these polynomials, combined with models of other components, a refined nonlinear operating model of the electro-hydrogen coupling system is established.

[0054] This embodiment provides an example of fitting the efficiency curves of an electrolyzer and a fuel cell to a polynomial. The fitting curve of the electrolyzer is as follows: Figure 3a As shown in Figure 3b, the fitting curve of the fuel cell is obtained after fitting, and the polynomial of the nonlinear efficiency model of the electrolyzer is:

[0055]

[0056] In the formula, ,in To input the power of the electrolytic cell, This refers to the capacity of the electrolytic cell.

[0057] The polynomial of the nonlinear efficiency model for fuel cells is:

[0058]

[0059] In the formula, ,in To input the power of the fuel cell, This refers to the capacity of the fuel cell.

[0060] S2. Set the objective function for optimization with the goal of minimizing the system operating cost of the nonlinear operating model, and set the constraints.

[0061] The objective function for optimization is to minimize the system operating cost so that the solution corresponds to the optimization objective of the system scheduling scheme.

[0062] Specifically, in this embodiment, system operating costs include the costs of purchasing electricity, heat, and hydrogen, and revenue includes revenue from selling electricity, heat, and hydrogen. Therefore, the objective function is set as follows: .

[0063] in:

[0064]

[0065]

[0066]

[0067]

[0068]

[0069]

[0070] For time intervals; , and These are the costs of purchasing electricity, heat, and hydrogen, respectively. , and These are revenue from electricity sales, revenue from heat sales, and revenue from hydrogen sales; and These are the electricity purchase price and the electricity sales price, respectively. and These are the purchase price and the selling price of heat, respectively. and These are the purchase price of hydrogen and the selling price of hydrogen, respectively. and These refer to power purchased and power sold, respectively. and These are respectively the heat purchase capacity and the heat sales capacity; and These are the hydrogen purchase capacity and hydrogen sales capacity, respectively.

[0071] The constraints include system equipment model constraints, electricity purchase and sale constraints, heat purchase and sale power constraints, hydrogen purchase and sale power constraints, wind turbine output constraints, photovoltaic output constraints, electric power balance constraints, thermal power balance constraints, hydrogen power balance constraints, and key equipment efficiency constraints. Among them, the key equipment efficiency constraints are polynomials of the nonlinear efficiency models of the electrolyzer and fuel cell set in step S1.

[0072] In this embodiment, the battery power constraint is:

[0073]

[0074] Among them, subscript Indexed by time period; and These represent the charging and discharging states of the battery. and These refer to the charging and discharging power of the battery; and These are the upper limits of battery charging and discharging power.

[0075] Battery energy constraints:

[0076]

[0077] in: , and These represent the battery's stored capacity at the beginning, end, and time period t, respectively. and These are the upper and lower limits of the battery's storage capacity. and These refer to the battery charging and discharging efficiency, respectively.

[0078] Electric boiler constraints:

[0079]

[0080] in: and These are the input and output power of the electric boiler, respectively. The heating efficiency of the electric boiler; and These are the upper and lower limits of the input power of the electric boiler; This indicates the working status of the electric boiler. 1 represents that the electric boiler is in working condition, and 0 represents that the electric boiler is in off condition. This represents the maximum number of times an electric boiler can be started and stopped per day.

[0081] Thermal storage tank power constraints:

[0082]

[0083] in: and These represent the charging and discharging states of the thermal storage tank, respectively. and These are the charging and discharging power of the thermal storage tank, respectively. and These are the upper limits of the heat storage tank's charging and discharging power, respectively.

[0084] Energy constraints of thermal storage tanks:

[0085]

[0086] in: , and These represent the heat storage capacity of the thermal storage tank during the initial period, the final period, and the time period t, respectively. The energy balance cycle of the thermal storage tank indicates that the amount of heat stored in the tank is equal from the initial state to the end of its balance cycle. and These are the upper and lower limits of the heat storage capacity of the thermal storage tank; and These refer to the heat release efficiency of the thermal storage tank.

[0087] Electrolytic cell start-up and shutdown constraints:

[0088]

[0089]

[0090]

[0091]

[0092]

[0093] in: This represents the total number of sampling periods; and These represent the start-up and shutdown states of the electrolytic cell, respectively. The variable is 0-1, where 1 indicates that the electrolytic cell is in the working state and 0 indicates that the electrolytic cell is in the off state. The unit time interval; and These are the daily maximum number of times an electrolytic cell can be started and stopped.

[0094] Electrolytic cell power constraints:

[0095] The upper and lower power limits for the electrolytic cell are:

[0096]

[0097] in: This refers to the operating power of the electrolytic cell; and These are the upper and lower limits of the operating power of the electrolytic cell under working conditions.

[0098] The power constraint for the electrolytic cell ramp-up is:

[0099]

[0100] in: This represents the maximum ramping power of the electrolytic cell per unit time period during operation.

[0101] Electrolyzer output constraints:

[0102]

[0103] in: The equivalent power of hydrogen production in the electrolyzer; This represents the hydrogen production efficiency of the electrolyzer.

[0104] Fuel cell start-stop constraints:

[0105]

[0106]

[0107]

[0108]

[0109]

[0110] in: and These represent the start-up and shutdown states of the fuel cell, respectively. These are 0-1 variables, representing the operating state of the fuel cell; The unit time interval; and These are the daily maximum number of times a fuel cell can be started and stopped.

[0111] Fuel cell power constraints:

[0112] The upper and lower power limits for fuel cell operation are:

[0113]

[0114] in: This represents the equivalent input power of the fuel cell; and These are the upper and lower limits of the operating power of the fuel cell when it is powered on.

[0115] The fuel cell ramping power constraint is:

[0116]

[0117] in: This represents the maximum ramp power of the hydrogen fuel cell per unit time period when it is powered on.

[0118] Fuel cell output constraints:

[0119]

[0120] in: The power output of the fuel cell; This refers to the power generation efficiency of fuel cells.

[0121] Hydrogen storage tank power constraints:

[0122]

[0123] in: and These represent the charging and discharging states of the hydrogen storage tank. and The hydrogen charging and discharging power of the hydrogen storage tank, respectively. and These are the upper limits of hydrogen charging and discharging power for the hydrogen storage tank.

[0124] Energy constraints of hydrogen storage tanks:

[0125]

[0126] in: , and These represent the hydrogen storage capacity of the hydrogen storage tank at the beginning, end, and time period t, respectively. and These represent the upper and lower limits of hydrogen storage capacity in the hydrogen storage tank. and These represent the hydrogen filling and discharging efficiency of the hydrogen storage tank.

[0127] Power purchase and sale constraints:

[0128]

[0129] in: and These refer to the power purchased and sold; and These are the state variables for electricity purchase and sale; and These are the upper limits for the power capacity to be purchased and sold.

[0130] Purchase and sale of heat power constraints:

[0131]

[0132] in: and These refer to the heat output for purchase and sale; and These are the state variables of the heat of purchase and sale; and These are the upper limits for the heat power to be purchased and sold.

[0133] Hydrogen purchase and sale power constraints:

[0134]

[0135] in: and The respective powers of purchasing and selling hydrogen; and These are the state variables for hydrogen purchase and sale; and These are the upper limits for the power to purchase and sell hydrogen.

[0136] Fan output constraints:

[0137]

[0138] in: This refers to the output power of the fan. This is the upper limit of the fan's output power.

[0139] Photovoltaic output constraints:

[0140]

[0141] in: Photovoltaic output power; This represents the upper limit of photovoltaic output power.

[0142] Electric power balance constraints:

[0143]

[0144] in: This refers to the output power of the fan. Photovoltaic output power; For electrical load; and These are the charging and discharging powers, respectively. and These refer to the power purchased and sold; The power input to the electrolytic cell; This refers to the electrical power output of the fuel cell.

[0145] Thermal power balance constraint:

[0146]

[0147] in: This is heat load data; and These are the charging and discharging heat power, respectively; This refers to the output power of the electric boiler. and These refer to the heat output for purchase and sale; and These are the waste heat recovered during the operation of the electrolyzer and the fuel cell, respectively.

[0148] Hydrogen power balance constraint:

[0149]

[0150] in: This is hydrogen load data; and These represent the hydrogen charging and discharging power, respectively. and The respective powers of purchasing and selling hydrogen; and These represent the power output from the electrolyzer and the power input to the fuel cell, respectively.

[0151] S3. Using the operating power of each device in the system within the scheduling period as the optimization variable, the initial scheduling scheme is generated by solving the problem using a multi-step variable neighborhood search algorithm based on the model, objective function and constraints.

[0152] Because the solution of the hydrogen-electric system has a complex solution space, and in order to improve the solution accuracy, a high-precision nonlinear model is introduced in step S1, which further increases the solution complexity. Under this premise, traditional linear solvers cannot solve the problem accurately, while conventional accurate solution methods, even if they can obtain an accurate solution, will be constrained by the complex solution space, resulting in slow computation.

[0153] Therefore, this invention uses a multi-step variable neighborhood search algorithm to solve the problem, which improves the solution speed under complex conditions and takes into account both global exploration and local optimization capabilities.

[0154] Specifically, this invention also provides a method for solving the problem using a multi-step variable neighborhood search algorithm based on the model, objective function, and constraints, including the following steps:

[0155] S31. Generate a multi-stage non-uniform step size sequence, where each stage of the non-uniform step size sequence contains several non-uniform step sizes from large to small.

[0156] The non-uniform step size sequence is generated using the following method: based on the size of the solution space, i.e. the upper and lower limits of the power of the device to be optimized and the required solution accuracy, a basic non-uniform step size set from coarse to fine is determined, and this basic set is repeated multiple times to form a multi-round search step size sequence. This multi-round sequence helps the algorithm to explore the solution space more fully.

[0157] Meanwhile, step S31 also constructs a multi-dimensional solution space with multiple coordinate axes using the operating power of each device as the coordinate axis, and generates a scheduling scheme that satisfies all constraints as an initial solution. This initial solution serves as the starting point for subsequent searches, and the objective function value of the initial solution is recorded.

[0158] S32. Using the first step size of the current stage as the initial step size, perform neighborhood probing starting from the initial solution to find the direction that makes the objective function value decrease.

[0159] Specifically, step S32 starts from the initial point and moves probingly along the positive and negative directions of one coordinate axis in the multidimensional solution space with a first step size to find the direction in which the objective function value decreases. If the objective function value decreases in a certain direction, the detection is successful and the direction is marked as a valid detection direction. The next detection continues along the valid detection direction.

[0160] S33. Perform a dynamic step-size expansion search in the direction that makes the objective function value decrease until the objective function value no longer decreases, then restore the initial step size and switch the search direction.

[0161] Specifically, the dynamic step-size expansion search method is as follows: After determining the effective detection direction in step S32, detection continues along that direction. If two consecutive detections along the same direction are successful, the step size is doubled and detection continues along that direction to accelerate the detection process. During this process, if the objective function value does not decrease after a certain detection, it means that the search in that direction has reached a local extremum. In this case, the initial step size of step S33 is restored, and the search is switched to the next direction that has not yet been searched, and the dynamic step-size expansion search is performed again.

[0162] S34. If no solution can be improved in all directions, switch to the next smaller step size in the current step size sequence as the initial step size, and repeat steps S31-S33 until all step size searches in the current stage are completed.

[0163] S35. Switch to the next stage and repeat steps S32-S34 using the non-uniform step size in the next stage until the iteration is completed and the initial scheduling scheme is generated.

[0164] Through the iterations of steps S31-S35 described above, the multi-step variable neighborhood search algorithm of this embodiment first uses the first step size of the first stage in the multi-stage non-uniform step size sequence to perform a dynamic step size expansion search. After completing the search in all directions, it switches to a smaller next step size in that stage to perform a dynamic step size expansion search again. When all step sizes in that stage have been used, it switches to the next stage to perform a dynamic step size expansion search. This search algorithm, which iteratively goes from coarse to fine and speeds up the process as much as possible, can achieve rapid solutions in complex solution spaces and has excellent ability to break through local optima.

[0165] S4. Based on the initial scheduling scheme, the operating power of each device in the initial scheduling scheme is further optimized using the adaptive simulated annealing algorithm to generate the optimal scheduling scheme.

[0166] After obtaining a relatively high-quality initial scheduling scheme through step S3, in order to further seek the global optimal solution, step S4 adopts the adaptive simulated annealing algorithm for further optimization.

[0167] This embodiment provides an implementation of the adaptive simulated annealing algorithm in step S4, including:

[0168] S41. Initialize temperature parameters, including initial temperature, temperature decay coefficient, and termination temperature.

[0169] S42. Starting from the initial scheduling scheme, generate a new solution using neighborhood perturbations based on the current temperature.

[0170] The initial scheduling scheme obtained in the first stage is used as the starting point for the second stage, and the current optimal solution and its objective function value are recorded. During the process of generating new solutions through neighborhood perturbation, the perturbation amplitude is related to the ratio of the current temperature to the initial temperature. In the early stages of the algorithm, the temperature is higher, the ratio is larger, and the perturbation amplitude is correspondingly larger, enabling the algorithm to explore a large area and find potentially better regions. As the temperature decreases, the perturbation amplitude adaptively shrinks, allowing the algorithm to perform a more refined local search within the already located high-quality regions, thus achieving a smooth transition from global exploration to local optimization.

[0171] S43. After generating a new solution, determine whether to accept the new solution according to the Metropolis acceptance criterion in order to optimize the scheduling scheme and escape local optima.

[0172] Specifically, Metropolis's acceptance criterion is to judge the objective function value of the new solution and compare it with the old solution. If the objective function value of the new solution is better, the new solution is accepted directly and the current state is updated. If the objective function value of the new solution is worse, the solution is accepted with an exponential probability that is negatively correlated with the current temperature, so as to balance the ability of global exploration and local optimization and avoid the algorithm from getting trapped in local optima too early.

[0173] S43. Iteration steps S41-S42: During the process, the temperature is gradually reduced until the iteration terminates, generating the optimal scheduling scheme.

[0174] Specifically, during the iteration process, an exponential cooling strategy is used to gradually reduce the system temperature in order to reduce the randomness of the search process. When the temperature drops to the preset minimum value or the number of times the objective function is called reaches the maximum number of iterations, the global optimal solution and its objective function value are finally output, which is the optimal scheduling scheme.

[0175] S5. Schedule the electro-hydrogen coupling system according to the optimal scheduling scheme.

[0176] The optimal scheduling scheme generated according to this embodiment includes the operating power of each device in the electro-hydrogen coupling system, such as the wind turbine, photovoltaic, battery, electrolyzer, fuel cell, and hydrogen storage tank, within a single scheduling time period. By sending the control commands for these operating powers to the system control unit, the economical, efficient, and precise scheduling and operation of the entire electro-hydrogen coupling system can be achieved.

[0177] To verify the superiority of the multi-stage hybrid search algorithm of this invention in solving the optimal scheduling problem of the electro-hydrogen coupling system, the best, worst, and average solutions obtained by the particle swarm optimization algorithm, simulated annealing algorithm, neighborhood search method, and the multi-stage hybrid search algorithm were compared. The comparison results are shown in Table 1 below. Convergence results of different algorithms are shown in the following figures. Figure 4 As shown, due to the randomness and divergence of the particle swarm optimization (PSO) algorithm, its solutions vary significantly and are generally poor, resulting in excessively high optimization scheduling costs and poor performance. Simulated annealing yields solutions significantly better than both PSO and neighborhood search algorithms, but the gap between its best and worst solutions is large, indicating high dispersion and unstable results. Neighborhood search, on the other hand, offers high search stability and very stable solutions, but it is prone to getting trapped in local optima, has weak global optima exploration capabilities, and converges slowly.

[0178] In contrast, the multi-stage hybrid search algorithm of this invention performs rapid optimization in the early stage and fine local optimization in the later stage, breaking through the local optimum. Its solution results and initial convergence speed are significantly better than other algorithms, and the solution results are more stable. While ensuring the reliability of the solution, it provides better support for the operation and scheduling of the electro-hydrogen coupling system.

[0179]

[0180] Table 1

[0181] In a specific example, the optimal scheduling scheme for power balance obtained by the multi-stage hybrid algorithm described in this invention is as follows: Figure 5 As shown, the thermal power balance is as follows Figure 6 As shown, the hydrogen power balance is as follows Figure 7 As shown.

[0182] Another embodiment of the present invention provides an optimized scheduling system for an electro-hydrogen coupling system based on a multi-stage hybrid search algorithm, used for a specific implementation of the method in the above embodiments, and its structural schematic diagram is shown below. Figure 8 As shown, it specifically includes:

[0183] The model generation module 100 is used to establish a refined nonlinear operating model of the electro-hydrogen coupling system based on the nonlinear operating characteristics of key equipment in the system.

[0184] Initialization module 200 is used to set the objective function for optimization with the goal of minimizing the system operating cost of the nonlinear operating model, and to set the constraints.

[0185] The multi-step variable neighborhood search module 300 is used to solve the initial scheduling scheme by taking the operating power of each device in the system within the scheduling cycle as the optimization variable, based on the model, objective function and constraints, using the multi-step variable neighborhood search algorithm.

[0186] The Adaptive Simulated Annealing Module 400 is used to further optimize the operating power of each device in the initial scheduling scheme using the Adaptive Simulated Annealing algorithm, based on the initial scheduling scheme, to generate the optimal scheduling scheme.

[0187] The scheduling execution module 500 is used to schedule the electro-hydrogen coupling system according to the optimal scheduling scheme.

[0188] In a further embodiment of the present invention, the process by which the model generation module establishes a nonlinear running model is specifically as follows:

[0189] Based on the coupling relationship between the working efficiency of the electrolyzer and fuel cell and the input power in the system, the efficiency curves of the electrolyzer and fuel cell are fitted into polynomials using interpolation.

[0190] A refined nonlinear operating model of the electro-hydrogen coupling system is established by using polynomials and combining them with the operating models of each device.

[0191] In a further embodiment of the invention, the multi-step variable neighborhood search module is configured to perform the following method:

[0192] a. Generate a multi-stage non-uniform step size sequence, where each stage contains several non-uniform step sizes from large to small; construct a multi-dimensional solution space with multiple coordinate axes using the operating power of each device as the coordinate axis, and set the initial solution;

[0193] b. Using the first step size of the current stage as the initial step size, perform neighborhood probing starting from the initial solution to find the direction that makes the objective function value decrease;

[0194] c. Perform dynamic step-size expansion search in the direction that makes the objective function value decrease until the objective function value no longer decreases, then restore the initial step size and switch the search direction;

[0195] d. If no solution can be improved in all directions, use the next step length of the current stage as the initial step length and repeat step bc until all step lengths of the current stage are searched.

[0196] e. Switch to the next stage and repeat step bd iteratively using the non-uniform step size in the next stage until the iteration is completed and the initial scheduling scheme is generated.

[0197] In a further embodiment of the invention, the adaptive simulated annealing module is configured to perform the following method:

[0198] Initialize temperature parameters;

[0199] Starting from the initial scheduling scheme, a new solution is generated using neighborhood perturbation based on the current temperature. The Metropolis acceptance criterion is used to determine whether to accept the new solution in order to optimize the scheduling scheme and escape local optima.

[0200] The process iterates step by step, gradually reducing the temperature, until the iteration terminates, generating the optimal scheduling scheme.

[0201] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Those skilled in the art will readily conceive of other embodiments of this disclosure upon considering the specification and practicing the disclosure herein. This invention is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described herein. The specification and embodiments are to be considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. A method for optimizing the scheduling of an electro-hydrogen coupling system based on a multi-stage hybrid search algorithm, characterized in that, include: Based on the nonlinear operating characteristics of key equipment in the electro-hydrogen coupling system, a refined nonlinear operating model of the electro-hydrogen coupling system is established. An objective function for optimization is set with the goal of minimizing the system operating cost of the nonlinear operating model, and constraints are set accordingly; Using the operating power of each device in the system within the scheduling period as the optimization variable, the initial scheduling scheme is generated by solving the problem using a multi-step variable neighborhood search algorithm based on the model, objective function and constraints. Based on the initial scheduling scheme, the operating power of each device in the initial scheduling scheme is further optimized using the adaptive simulated annealing algorithm to generate the optimal scheduling scheme. The electro-hydrogen coupling system is scheduled according to the optimal scheduling scheme. The initial scheduling scheme is generated by solving the model, objective function, and constraints using a multi-step variable neighborhood search algorithm, including: a. Generate a multi-stage non-uniform step size sequence, wherein each stage of the non-uniform step size sequence contains several non-uniform step sizes from large to small; construct a multi-dimensional solution space containing multiple coordinate axes with the operating power of each device as the coordinate axis, and set an initial solution; b. Using the first step size of the current stage as the initial step size, perform neighborhood probing starting from the initial solution to find the direction that makes the value of the objective function decrease; c. Perform a dynamic step-size expansion search in the direction that makes the value of the objective function decrease until the value of the objective function no longer decreases, then restore the initial step size and switch the search direction; d. If no solution can be improved in all directions, use the next step length of the current stage as the initial step length and repeat step bc until all step lengths of the current stage are searched. e. Switch to the next stage and repeat step bd iteratively using the non-uniform step size in the next stage until the iteration is completed and the initial scheduling scheme is generated.

2. The method for optimizing and scheduling an electro-hydrogen coupling system based on a multi-stage hybrid search algorithm as described in claim 1, characterized in that, The dynamic step-size expansion search includes: The search continues in the direction that makes the value of the objective function decrease. If the value of the objective function continues to decrease, the step size is increased to improve the search speed until the value of the objective function stops decreasing. Then, the initial step size is restored and the search direction is switched.

3. The method for optimizing and scheduling an electro-hydrogen coupling system based on a multi-stage hybrid search algorithm as described in claim 1, characterized in that, The operating power of each device in the initial scheduling scheme is further optimized using an adaptive simulated annealing algorithm, including: Initialize temperature parameters; Starting from the initial scheduling scheme, a new solution is generated using neighborhood perturbation based on the current temperature. The Metropolis acceptance criterion is used to determine whether to accept the new solution in order to optimize the scheduling scheme and escape local optima. The process iterates step by step, gradually reducing the temperature, until the iteration terminates, generating the optimal scheduling scheme.

4. The method for optimizing and scheduling an electro-hydrogen coupling system based on a multi-stage hybrid search algorithm as described in claim 1, characterized in that, The key equipment includes an electrolyzer and a fuel cell.

5. The method for optimizing and scheduling an electro-hydrogen coupling system based on a multi-stage hybrid search algorithm as described in claim 4, characterized in that, The specific process of establishing the nonlinear operation model is as follows: Based on the coupling relationship between the working efficiency of the electrolyzer and fuel cell and the input power in the system, the efficiency curves of the electrolyzer and fuel cell are fitted into polynomials using interpolation. Using the polynomial and combining it with the operating models of each device, a refined nonlinear operating model of the electro-hydrogen coupling system is established.

6. The method for optimizing and scheduling an electro-hydrogen coupling system based on a multi-stage hybrid search algorithm as described in claim 1, characterized in that, The system operating cost is the difference between the total cost of purchasing electricity, heat, and hydrogen and the total revenue from selling electricity, heat, and hydrogen.

7. The method for optimizing and scheduling an electro-hydrogen coupling system based on a multi-stage hybrid search algorithm as described in claim 1, characterized in that, The constraints include: System equipment model constraints, electricity purchase and sale constraints, heat purchase and sale power constraints, hydrogen purchase and sale power constraints, wind turbine output constraints, photovoltaic output constraints, electrical power balance constraints, thermal power balance constraints, hydrogen power balance constraints, and key equipment efficiency constraints.

8. The method for optimizing and scheduling an electro-hydrogen coupling system based on a multi-stage hybrid search algorithm as described in claim 1, characterized in that, The optimization variables include: electricity purchased and sold, heat purchased and sold, hydrogen purchased and sold, and the operating power of each piece of equipment.

9. An optimized scheduling system for an electro-hydrogen coupling system based on a multi-stage hybrid search algorithm, characterized in that, include: The model generation module is used to establish a refined nonlinear operating model of the electro-hydrogen coupling system based on the nonlinear operating characteristics of key equipment in the system. An initialization module is used to set the objective function for optimization with the goal of minimizing the system operating cost of the nonlinear operating model, and to set the constraints. The multi-step variable neighborhood search module is used to solve the initial scheduling scheme by taking the operating power of each device in the system within the scheduling cycle as the optimization variable, based on the model, objective function and constraints, and using the multi-step variable neighborhood search algorithm. The adaptive simulated annealing module is used to further optimize the operating power of each device in the initial scheduling scheme using the adaptive simulated annealing algorithm, based on the initial scheduling scheme, to generate the optimal scheduling scheme. The scheduling execution module is used to schedule the electro-hydrogen coupling system according to the optimal scheduling scheme; The multi-step variable neighborhood search module uses a multi-step variable neighborhood search algorithm to solve the model, objective function, and constraints to generate an initial scheduling scheme, including: a. Generate a multi-stage non-uniform step size sequence, wherein each stage of the non-uniform step size sequence contains several non-uniform step sizes from large to small; construct a multi-dimensional solution space containing multiple coordinate axes with the operating power of each device as the coordinate axis, and set an initial solution; b. Using the first step size of the current stage as the initial step size, perform neighborhood probing starting from the initial solution to find the direction that makes the value of the objective function decrease; c. Perform a dynamic step-size expansion search in the direction that makes the value of the objective function decrease until the value of the objective function no longer decreases, then restore the initial step size and switch the search direction; d. If no solution can be improved in all directions, use the next step length of the current stage as the initial step length and repeat step bc until all step lengths of the current stage are searched. e. Switch to the next stage and repeat step bd iteratively using the non-uniform step size in the next stage until the iteration is completed and the initial scheduling scheme is generated.

Images