CEO-EEFO optimization algorithm-based hydrogen-electric coupling system power distribution method, electronic equipment and medium

Abstract

The invention discloses a CEO-EEFO optimization algorithm-based hydrogen-electric coupling system power distribution method, electronic equipment and a medium. The method comprises the steps of determining a population scale, an iteration condition and a variable dimension of a CEO-EEFO algorithm according to a system variable to be solved; an initialized CEO-EEFO algorithm is obtained based on the input parameters; a double-layer dynamic model is constructed based on an initialized CEO-EEFO algorithm, the double-layer dynamic model comprises an upper-layer power grid operator model and a lower-layer user model, the upper-layer power grid operator model comprises a system comprehensive income optimal target function, and the lower-layer user model comprises a power utilization cost and user perception comfort degree target function; the upper-layer power grid operator model is solved based on the variable dimension to obtain an electricity price strategy and a power distribution strategy, and the lower-layer user model receives the electricity price strategy to adjust the load demand and feed back the load demand to the upper-layer power grid operator model; and performing iterative calculation based on the population scale and the iteration condition, and outputting a final power distribution strategy. According to the invention, the power distribution precision and the overall economic index of the electro-hydrogen coupling system can be improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of multi-energy grid collaborative optimization and intelligent control, and particularly relates to a hydrogen-electric coupling system power distribution method based on a CEO-EEFO optimization algorithm, an electronic device and a medium. BACKGROUND

[0002] Under the background of continuous development in the energy field, energy demand is increasing day by day, and traditional energy supply modes face many challenges, such as low energy utilization efficiency and great environmental pressure, and the hydrogen-electric coupling system gradually becomes a research hotspot due to its advantages in energy storage and conversion.

[0003] In related technologies, the power distribution of the hydrogen-electric coupling system is a complex multi-objective optimization problem. On the one hand, the system operator needs to consider the comprehensive benefits of the system and reasonably formulate electricity price strategies and power distribution strategies; on the other hand, users not only pay attention to the cost of electricity, but also have requirements for the perceived comfort of electricity.

[0004] Traditional optimization algorithms, such as genetic algorithms and particle swarm algorithms, have slow convergence speed and are prone to local optimization when dealing with such complex multi-objective, multi-constrained double-layer dynamic optimization problems, and it is difficult to efficiently and accurately achieve optimal power distribution of the hydrogen-electric coupling system.

[0005] Therefore, a more efficient optimization algorithm is needed to solve the power distribution problem of the hydrogen-electric coupling system to improve the accuracy of the optimal power distribution of the hydrogen-electric coupling system. SUMMARY

[0006] The purpose of the present application is to improve the accuracy of the hydrogen-electric coupling system power distribution method based on the CEO-EEFO optimization algorithm by collecting multiple key physical parameters and constructing a closed continuous-time neural network model based on a sparse loop.

[0007] To achieve the above objectives, this invention proposes a power allocation method for hydrogen-electric coupling systems based on the CEO-EEFO optimization algorithm, comprising: determining the population size, iteration conditions, and variable dimensions of the CEO-EEFO algorithm according to the system variables to be solved, wherein the system variables include the momentary output power of the gas turbine, the momentary operating power of the electrolyzer, the momentary output power of the fuel cell, and the grid electricity price curve; initializing the CEO-EEFO algorithm based on input parameters to obtain an initialized CEO-EEFO algorithm, wherein the input parameters include illumination condition data, user load data, and initial electricity price data; and constructing a two-layer dynamic model based on the initialized CEO-EEFO algorithm. The two-layer dynamic model includes an upper-layer grid operator model and a lower-layer user model. The upper-layer grid operator model includes upper-layer constraints and an objective function for optimizing overall system revenue. The lower-layer user model includes lower-layer constraints and objective functions for electricity cost and user perceived comfort. Based on the variable dimensions, the upper-layer grid operator model is solved to obtain an electricity pricing strategy and a power allocation strategy. The lower-layer user model receives the electricity pricing strategy to adjust load demand and feeds it back to the upper-layer grid operator model. Iterative calculations are performed based on the population size and the iteration conditions to output the final power allocation strategy. The termination conditions include the maximum number of iterations or a Stackelberg equilibrium state.

[0008] In one optional implementation, the CEO-EEFO algorithm is initialized based on the input parameters to obtain an initialized CEO-EEFO algorithm. Specifically, this includes: generating an initial population based on a two-dimensional discrete memristor hyperchaotic mapping to obtain a homogenized population; and obtaining the initialized CEO-EEFO algorithm based on the homogenized population and the input parameters.

[0009] In one optional implementation, the CEO-EEFO algorithm generates an initial population based on a two-dimensional discrete memristor hyperchaotic mapping to obtain a homogenized population. Specifically, this includes selecting two different initial individuals from the current eel population. and The original individuals are then linearly mapped to specific attraction basin regions [-0.5, 0.5] and [-0.25, 0.25] of the chaotic map: In the formula, and It is an individual eel. , The initial position of the chaos after mapping; , These are the upper and lower bounds of the search space, respectively; the mapped values ​​are... and Using this as the initial value, iterate N times to generate N chaotic candidate individuals: Perform an inverse mapping on the N chaotic candidate individuals to obtain the actual search space of the original problem: Based on the actual search space, the original individuals are... , Generate N evolutionary directions: Using the Lyapunov index The linear energy factor of the decision-making EEFO algorithm for rest, migration, and hunting measurements. Replaced with super-chaotic energy factor ; In the formula, when hour, ;when hour, ;when At the same time, optimize the interaction strategy: ;when Enhanced interaction strategies: In the formula, This represents the step size for the algorithm's iterative search. , This is the best solution in the population at the current iteration number.

[0010] In one optional implementation, the upper-level constraints include system energy flow balance constraints, gas turbine ramp-up constraints, electrolyzer operation constraints, fuel cell operation constraints, and hydrogen storage device SOC constraints. The expression for the upper-level constraints is as follows: In the formula, , , They are respectively The output power of photovoltaic equipment, gas turbines, and fuel cells at all times; , They are respectively Real-time power grid load and power consumption of electrolytic cell equipment; This represents the difference in output power between the gas turbines. Set an upper limit for the power ramp-up of gas turbine equipment; , These are the minimum and maximum operating power limits for the electrolytic cell, respectively. The rated core power of the electrolytic cell; Rated operating power of fuel cell power generation equipment; Electrolysis efficiency of the electrolytic cell; The specific energy consumption for hydrogen production is the power required to produce one unit mass of hydrogen. For fuel cell power generation efficiency; The energy density of hydrogen is the energy released when a unit mass of hydrogen is completely converted. , For hydrogen storage equipment The hydrogen storage capacity status at the current moment and the previous moment; The rated hydrogen storage capacity of the hydrogen storage device; , The upper and lower limits of the hydrogen storage equipment capacity are defined as follows; the expression for the optimal objective function of the system's overall benefit is: In the formula, , , These are revenues from electricity sales, hydrogen sales, and carbon emission reduction, respectively. , , The costs are respectively the operating and maintenance costs of the photovoltaic system, the gas turbine, and the electrolyzer. This refers to the cost of the energy storage system.

[0011] In one optional implementation, the expression for the lower-level constraint is: ; In the formula, , These represent the upper and lower limits of the user's flexible load adjustment range, respectively; the expression for the objective function of the electricity cost and user perceived comfort objective function is as follows: In the formula, for Real-time electricity price on the power grid; The actual electricity consumption of the user at time t; Let t be the user's perceived loss cost.

[0012] In one optional implementation, iterative calculations are performed based on the population size and the iteration conditions, and a comprehensive performance evaluation is also output. The expression for the comprehensive performance evaluation is: In the formula, , , These are revenues from electricity sales, hydrogen sales, and carbon emission reduction, respectively. , , , These are the operating and maintenance costs of the photovoltaic system, gas turbine, electrolyzer, and fuel cell, respectively. This is due to the loss of hydrogen production efficiency; To assess the perceived cost of loss for users.

[0013] In one optional implementation, the expression for the user-perceived loss cost is: In the formula, For users to perceive the cost of loss; This represents the user's actual electricity consumption. Electricity consumption to meet user needs; The user perception loss coefficient during the low deviation phase; The user perception loss coefficient during the high deviation phase; Set a maximum value for the cost of perceived loss to the user.

[0014] In one optional implementation, the expression for the hydrogen production efficiency loss is: In the formula, Let t be the power consumed by the electrolytic cell during operation at time t; This is the minimum operating power of the electrolytic cell; The cost of starting and stopping the electrolytic cell.

[0015] The present invention also proposes an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute any of the power allocation methods for hydrogen-electric coupling systems based on the CEO-EEFO optimization algorithm described in the present invention.

[0016] The present invention also proposes a medium storing a computer program, which, when executed by a processor, implements the power allocation method for a hydrogen-electric coupling system based on the CEO-EEFO optimization algorithm described in any one of the claims.

[0017] The beneficial effects of this invention are as follows: by constructing a dynamic game framework with grid operators as leaders and user groups as responders, the upper layer optimizes equipment operation and maintenance costs and power allocation strategies to form new electricity pricing strategies. The lower layer receives the new electricity pricing strategies transmitted from the upper layer and optimizes load demand based on electricity costs and user perceived comfort. Through interactive iteration, the global collaboration of the model is achieved, thereby improving the overall economic indicators and power allocation accuracy of the electric-hydrogen coupling system. Attached Figure Description

[0018] Figure 1 A flowchart of a power allocation method for a hydrogen-electric coupling system based on the CEO-EEFO optimization algorithm provided in an embodiment of the present invention; Figure 2The overall flowchart of the CEO-EEFO algorithm, which is an improved version of the CEO-EEFO optimization algorithm for power allocation in a hydrogen-electric coupling system, provided for an embodiment of the present invention, is shown below. Figure 3 The diagram shows the interaction relationship of a two-layer dynamic game model for a power allocation method for a hydrogen-electric coupling system based on the CEO-EEFO optimization algorithm, as provided in an embodiment of the present invention. Detailed Implementation

[0019] The CEO-EEFO optimization algorithm, due to its potential advantages in complex optimization problems, has been introduced into the field of electric-hydrogen coupling system optimization to achieve a balance between overall system benefits and user electricity experience. Existing electric-hydrogen coupling system optimization methods suffer from four main shortcomings when dealing with uncertainties on both the source and load sides: First, traditional dynamic programming algorithms have significant limitations in multi-agent game scenarios. When the system dimension exceeds the processing capacity of traditional dynamic programming algorithms, the algorithm faces the "curse of dimensionality," leading to an exponential decrease in solution efficiency. This is especially true when the system has high-dimensional nonlinear constraints (such as multi-period coupled energy storage charging and discharging constraints, and multi-node voltage stability constraints), where traditional methods struggle to effectively solve for Nash equilibrium solutions and often fall into local optimum traps. Second, existing research generally adopts economically driven models under the assumption of complete rationality, using linear utility functions to characterize user response behavior, neglecting the emotional costs prevalent in actual demand-side management. When there are periodic power shortages or drastic price fluctuations, non-economic factors such as user psychological resistance due to disrupted electricity plans, trust costs arising from doubts about power supply reliability, and complaints caused by a sudden drop in comfort all contribute to these problems. This will significantly impact the demand response effect; third, conventional metaheuristic algorithms (such as genetic algorithms and particle swarm optimization algorithms) rely on random search and fixed parameter settings, which are poorly adaptable to high-dimensional nonlinear constraints (such as energy storage SOC limits and equipment ramp-up rates). They also have shortcomings in population initialization strategies and convergence control, resulting in limited solution accuracy and premature convergence. Furthermore, conventional constraint handling methods (such as penalty function methods) have large relaxation errors for non-convex constraints, which can easily generate conservative or infeasible solutions, affecting the safety of system operation; fourth, the existing power allocation strategies of electro-hydrogen coupling systems are insufficient in their synergistic optimization between the fluctuation of photovoltaic output and the dynamic response characteristics of hydrogen production equipment. The intermittency of photovoltaic power generation can easily lead to frequent start-stop or inefficient operation of electrolyzers, resulting in a significant decrease in hydrogen production efficiency. At the same time, the lack of a dynamic matching mechanism between energy storage devices (such as hydrogen storage tanks and fuel cells) and flexible loads in the hydrogen energy consumption process makes it difficult to achieve full-chain optimization of hydrogen energy production-storage-consumption, resulting in energy waste and economic losses.

[0020] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0021] like Figure 1 andFigure 3 As shown, according to an embodiment of the present invention, in one aspect, a power allocation method for a hydrogen-electric coupling system based on the CEO-EEFO optimization algorithm is provided, comprising the following steps: Step S101: Determine the population size, iteration conditions, and variable dimensions of the CEO-EEFO algorithm based on the system variables to be solved. The system variables include the gas turbine's momentary output power, the electrolyzer's momentary operating power, the fuel cell's momentary output power, and the grid electricity price curve. Step S103: Initialize the CEO-EEFO algorithm based on the input parameters to obtain the initialized CEO-EEFO algorithm. The input parameters include illumination condition data, user load data and initial electricity price data. Step S105: Construct a two-layer dynamic model based on the initialization CEO-EEFO algorithm. The two-layer dynamic model includes an upper-layer grid operator model and a lower-layer user model. The upper-layer grid operator model includes upper-layer constraints and the objective function for optimal overall system revenue. The lower-layer user model includes lower-layer constraints and the objective function for electricity cost and user perceived comfort. Step S107: Solve the upper-level grid operator model based on the variable dimension to obtain the electricity pricing strategy and power allocation strategy. The lower-level user model receives the electricity pricing strategy to adjust load demand and feeds it back to the upper-level grid operator model. Step S109: Perform iterative calculations based on population size and iteration conditions, and output the final power allocation strategy. Termination conditions include the maximum number of iterations or the Stackelberg equilibrium state.

[0022] In this embodiment, based on the variables to be solved (such as the gas turbine's instantaneous output power, the electrolyzer's instantaneous operating power, the fuel cell's instantaneous output power, and the grid electricity price curve), the boundary conditions of the CEO-EEFO algorithm parameters (such as gas turbine power ramp-up constraints, electrolyzer start-up and shutdown power constraints, and fuel cell power output upper and lower limit constraints) are set. Based on the variables to be solved and the boundary conditions, the population size, iteration conditions, and variable dimensions of the CEO-EEFO algorithm can be determined.

[0023] Among them, the core elements of the hydrogen-electric coupling system are analyzed in depth. Based on the system variables such as the output power of the gas turbine at any time (its output power is affected by factors such as fuel supply and equipment efficiency, and it serves as the basic power source for stable energy supply in the system), the operating power of the electrolyzer at any time (which converts electrical energy into hydrogen energy, and the power level directly affects hydrogen production and energy storage efficiency), the output power of the fuel cell at any time (which realizes the efficient conversion of hydrogen energy into electrical energy to supply power to the user side), and the grid electricity price curve (which reflects the supply and demand relationship and cost fluctuations in the electricity market), the core parameters of the CEO-EEFO algorithm are scientifically determined.

[0024] Among them, population size: based on the complexity of the system and the limitations of computing resources, an appropriate number of initial solution individuals are set to ensure that the algorithm has sufficient diversity exploration capability in the search space.

[0025] Iteration conditions: Taking into account the real-time requirements of system operation and computational accuracy, set iteration termination conditions, such as the maximum number of iterations and the convergence threshold of the objective function, to balance computational efficiency and optimization effect.

[0026] Variable Dimension: Based on the number of system variables, determine the dimension of the algorithm's search space to provide a basic framework for subsequent algorithm solutions.

[0027] Based on collected data on illumination conditions (which affect the power generation of renewable energy systems such as photovoltaic systems, and consequently the energy input structure of hydrogen-electric coupling systems), user load data (containing electricity demand characteristics of different time periods and different types of users, serving as a key basis for system power allocation), and initial electricity price data (reflecting the current electricity market price benchmark), the CEO-EEFO algorithm is initialized. By incorporating this actual operational data into the algorithm, a population containing initial feasible solutions is generated, providing a starting point for subsequent optimization processes and allowing the algorithm to begin searching from a solution space that closely approximates the actual system operating state.

[0028] Compared to traditional optimization algorithms, the CEO-EEFO optimization algorithm has stronger global search capabilities and faster convergence speed. It can quickly find near-global optimal power allocation schemes in complex hydrogen-electric coupling system power allocation problems, significantly improving optimization efficiency.

[0029] Based on the initialized CEO-EEFO algorithm, a two-layer dynamic model is constructed to realize the game and collaborative optimization between power grid operators and users.

[0030] Upper-level grid operator model: This model revolves around the objective function of maximizing overall system revenue, comprehensively considering factors such as generation costs, electricity sales revenue, investment and operation and maintenance costs of energy storage equipment, and carbon trading revenue, aiming to maximize the overall economic benefits of the system. Simultaneously, it incorporates upper-level constraints such as equipment operating parameter limitations (e.g., upper and lower limits of gas turbine power, maximum operating current of electrolyzers), the law of conservation of energy, and the requirements for safe and stable grid operation to ensure the feasibility and safety of the optimization strategy.

[0031] Lower-level user model: Guided by the objective functions of electricity cost and user perceived comfort, this model aims to reduce electricity costs by adjusting electricity consumption behavior (such as shifting some adjustable load to off-peak hours) while meeting users' basic electricity needs, and also taking into account users' requirements for power stability and continuity. The model includes lower-level constraint functions such as user load characteristic constraints (e.g., minimum guaranteed load, maximum adjustable load range) and equipment operation constraints (e.g., start-stop restrictions for equipment such as air conditioners) to ensure the rationality of the user-side optimization strategy.

[0032] By constructing a two-layer dynamic model, the different interests of grid operators and users are fully considered, so as to achieve the optimal overall system benefits, reduce users' electricity costs, improve users' electricity comfort, balance the interests of energy supply and consumption, and promote the sustainable development of the energy system.

[0033] Based on a defined number of variables, the CEO-EEFO algorithm is used to solve the upper-level grid operator model. Through optimization of the objective function, the optimal electricity pricing strategy (such as time-of-use pricing or tiered pricing schemes) and power allocation strategy (clearly defining the power output / input arrangements of each generation and energy storage device at different times) are obtained. After receiving the electricity pricing strategy from the upper level, the lower-level user model adjusts its load demand according to its own objective function and constraints (such as delaying the operation time of some non-emergency electrical equipment) and feeds the adjusted load demand back to the upper-level grid operator model. Based on the new load information, the upper-level model re-optimizes the electricity pricing and power allocation strategies, forming a dynamic interaction and collaborative optimization mechanism between the upper and lower-level models.

[0034] Based on the set population size, in each iteration, the CEO-EEFO algorithm performs selection, crossover, and mutation operations on the solutions in the current population by simulating optimization mechanisms such as biological evolution or physical phenomena, generating a new solution population. The algorithm determines whether the termination requirement is met according to the iteration conditions. It stops iterative calculation when the maximum number of iterations is reached or a Stackelberg equilibrium state is achieved (i.e., both the strategies of the upper-level grid operator and the lower-level user are optimal, and unilateral changes to the strategy by either party cannot improve their own benefits). The final power allocation strategy is then output, providing a scientific decision-making basis for the actual operation of the hydrogen-electric coupling system.

[0035] By combining real-time data such as illumination conditions, user load, and electricity price for algorithm initialization and model optimization, the power allocation strategy closely matches the actual operating state of the system, enhancing the practicality and effectiveness of the strategy and improving the operational stability and reliability of the hydrogen-electric coupling system.

[0036] The interactive feedback mechanism between the upper and lower layer models enables dynamic and coordinated adjustment of power grid operation strategies and user electricity consumption behavior, which can flexibly respond to fluctuations in electricity market prices and changes in user load, and improve the system's adaptability to complex environments.

[0037] A two-layer dynamic game optimization framework based on the improved Chaotic Evolutionary Optimization-Electric Eel Foraging Algorithm (CEO-EEFO) is proposed. A uniform initial population is generated through a two-dimensional discrete memristor hyperchaotic mapping. The Lyapunov exponent is used to dynamically adjust the exploration and development weights of the algorithm, and a chaotic perturbation factor is embedded to improve global convergence. Simultaneously, an Energy Efficiency-Comfort Synergy Index (ESSI) is designed to balance the economic benefits to the power grid with the perceived losses to users, constructing a two-layer game architecture with the power grid operator as the leader and the user group as the responders.

[0038] The eel foraging optimization algorithm originates from the cooperative foraging behavior among electric eel groups. Its core idea is to effectively explore and utilize the search space by simulating the interaction, resting, hunting, and migration behaviors of electric eels, thereby optimizing the solution. The specific framework of the EEFO algorithm is as follows: 1.1 Population Initialization: In the initial stage, eel population locations are created randomly. This characterizes the eel's initial location at an uncertain food source in the environment, while simultaneously establishing an energy factor. Behavioral assessment of eel populations was conducted based on energy factors. Predicting the next behavioral move of an eel population: (1) (2) In the formula: For the first The initial position of the electric eel represents a potential solution in the search space; , These are the upper and lower bounds of the search space, respectively, set according to the upper and lower limits of the solution variables; For the first The energy state indicators of individual eels at each iteration are used to determine their subsequent interaction, rest, hunting, and migration behaviors. and and represent the current iteration number and the total iteration number, respectively; s is a random value located in the interval [0,1].

[0039] 1.2 Interaction Strategy: During this stage, eels engage in information exchange by comparing their own position within the population. Specifically, they update their position by measuring the differences between eels randomly selected in the population and those randomly generated in the search space. This behavioral pattern effectively drives the population's exploration process. When the fitness of the i-th eel in the population is less than the fitness of a randomly selected eel, i.e. hour: (3) when hour: (4) (5) In the formula: To point to the first The first individual eel The vector representing the direction of movement of the position; For the first During the nth iteration The location of each individual eel; The first randomly selected from the population The location of each individual eel; The locations of eels are randomly generated within the search space; For the first The center value of the eel population position in the next iteration; v represents the random movement vector of the eels stirring in different directions during the encounter, which is a random value in the interval [-1,1]. The total number of individuals within the eel population; , , It is a random value located in the interval [0,1].

[0040] 1.3 Rest Strategy: In this stage, the eel constructs its resting region by projecting its position onto the main diagonal of the search space. To improve the algorithm's local search performance, the size of the resting region gradually decreases as the iteration progresses, thereby increasing the exploitation rate. First, the search space and the eel's position are normalized: (6) In the formula: For the first The second iteration The individual eels are in their initial positions in the rest area.

[0041] Eel resting area It can be determined by its behavior before it comes to rest: (7) In the formula: It is a random number between (0,1); This is the position vector of the current optimal solution.

[0042] Each individual eel is identified by its location in the resting area. With current location To update its velocity vector toward the rest area : (8) In the formula: , It is a random number between (0,1); This indicates that it is rounded to the nearest whole number.

[0043] 1.4 Migration Strategies: During this phase, the eels migrate from the resting area to the hunting area. To avoid them easily getting trapped in local optima, the levy flight function is used to complete the migration, and the specific process is as follows: (9) In the formula: The location of the hunting grounds formed by individual eels; It is a random number between (0,1).

[0044] Its velocity vector pointing to the next location in the hunting area is: (10) In the formula: The next location of the individual eel in the hunting area; This is the Levi flight function, with the function parameter set to 1.5.

[0045] 1.5 Hunting Strategy: When an electric eel detects prey, it forms a circular encirclement around the prey (electric circle) and continuously contracts this circle. During this process, the prey will scatter randomly within the hunting area, while the electric eel dynamically adjusts its own distribution based on the prey's real-time location to maintain the convergence of the encirclement. Equations (11) and (12) describe the eel's coiling habit during the hunting process: (11) (12) (13) In the formula: Location of hunting grounds formed by eel groups; For eel hunting curling factor. Let be the velocity vector of an individual eel pointing to the next position during its hunting behavior; It is a random number between (0,1).

[0046] The specific algorithm flow of EEFO is as follows: Step 1: Set the basic parameters of the algorithm according to the input parameters (gas turbine output power, electrolyzer allocated power, fuel cell output power, and grid electricity price), such as population size, maximum number of iterations, dimension, and upper and lower limits of the search space.

[0047] Step 2: Initialize each individual in the population and calculate the initial fitness of each individual based on the objective function.

[0048] Step 3: If the energy factor E>1, the eel enters the interactive behavior stage and updates its position.

[0049] Step 4: If energy factor If the eel enters a resting phase, it updates its position using a local search strategy and simultaneously updates the current optimal solution.

[0050] Step 5: If energy factor If the individual eel enters the migration phase, a global search strategy is used to update the individual's position and simultaneously update the current optimal solution.

[0051] Step 6: If energy factor If the eel enters the hunting behavior phase, it updates its position by simulating predation strategies and simultaneously updates the current optimal solution.

[0052] Step 7: Determine whether the algorithm's execution result meets the preset termination condition. If it does, proceed to Step 8.

[0053] Step 8: End.

[0054] The traditional EEFO algorithm uses a random initialization strategy, which leads to uneven population distribution, incomplete solution space coverage, difficulty in capturing the global optimum in high-dimensional scenarios, and lack of an effective perturbation mechanism during the migration phase, making it difficult to escape the local optimum trap. In addition, its energy factor adjustment method is rigid, which is prone to premature convergence in the later stages of iteration. Therefore, a two-dimensional discrete memristor hyperchaotic mapping is introduced to generate the initial population, replacing the random initialization of the population in EEFO, so that the initial population distribution is more uniform and more comprehensively covers the potential solution space.

[0055] Further, step S103, initializing the CEO-EEFO algorithm based on the input parameters to obtain the initialized CEO-EEFO algorithm, specifically includes the following steps: Step S1031: For the CEO-EEFO algorithm, generate an initial population based on the two-dimensional discrete memristor hyperchaotic mapping to obtain a homogenized population.

[0056] Step S1033: Obtain the initial CEO-EEFO algorithm based on the homogenized population and input parameters.

[0057] Generating the initial population using a two-dimensional discrete memristor hyperchaotic mapping significantly enhances the uniformity and randomness of the population, avoiding the uneven distribution and local clustering problems that easily occur with traditional random initialization, thus providing a better search starting point for subsequent algorithm iterations. Combining a uniformized population with the input parameters for initialization ensures that the initial state of the CEO-EEFO algorithm accurately matches the requirements of the actual problem, reducing the risk of slow convergence or getting trapped in local optima due to improper initialization, and improving the efficiency and stability of subsequent optimization.

[0058] Furthermore, such as Figure 2 As shown, step S1031, for the CEO-EEFO algorithm, generates an initial population based on a two-dimensional discrete memristor hyperchaotic mapping to obtain a homogenized population, specifically including the following steps: Step S10311: Select two different original individuals from the current eel population. and The original individuals are linearly mapped to specific attraction basin regions [-0.5, 0.5] and [-0.25, 0.25] of the chaotic map: ; In the formula, and It is an individual eel. , The initial position of the chaos after mapping; , These are the upper and lower bounds of the search space, respectively; Step S10313: Map the obtained and Using this as the initial value, iterate N times to generate N chaotic candidate individuals: ; Step S10315: Perform inverse mapping on the N chaotic candidate individuals to obtain the actual search space of the original problem: ; Step S10317: Based on the actual search space, the original individuals , Generate N evolutionary directions: ; Step S10319: Use the Lyapunov index The linear energy factor of the decision-making EEFO algorithm for rest, migration, and hunting measurements. Replaced with super-chaotic energy factor ; ; In the formula, when hour, ;when hour, ; Step S10321: When At the same time, optimize the interaction strategy: ; Step S10323: When Enhanced interaction strategies: ; In the formula, This represents the step size for the algorithm's iterative search. , This is the best solution in the population at the current iteration number.

[0059] A master-slave game framework is adopted, with the power grid operator (PGO) as the strategy leader and the user group as the responders, to construct the game model architecture.

[0060] Furthermore, the upper-level constraints include system energy flow balance constraints, gas turbine ramp-up constraints, electrolyzer operation constraints, fuel cell operation constraints, and hydrogen storage device SOC constraints. The expressions for the upper-level constraints are as follows: ; In the formula, , , They are respectively The output power of photovoltaic equipment, gas turbines, and fuel cells at all times; , They are respectively Real-time power grid load and power consumption of electrolytic cell equipment; This represents the difference in output power between the gas turbines. Set an upper limit for the power ramp-up of gas turbine equipment; , These are the minimum and maximum operating power limits for the electrolytic cell, respectively. The rated core power of the electrolytic cell; Rated operating power of fuel cell power generation equipment; Electrolysis efficiency of the electrolytic cell; The specific energy consumption for hydrogen production is the power required to produce one unit mass of hydrogen. For fuel cell power generation efficiency; The energy density of hydrogen is the energy released when a unit mass of hydrogen is completely converted. , For hydrogen storage equipment The hydrogen storage capacity status at the current moment and the previous moment; The rated hydrogen storage capacity of the hydrogen storage device; , These are the upper and lower limits of the capacity status of hydrogen storage equipment.

[0061] The expression for the objective function that yields the optimal overall system benefit is: ; In the formula, , , These are revenues from electricity sales, hydrogen sales, and carbon emission reduction, respectively. , , The costs are respectively the operating and maintenance costs of the photovoltaic system, the gas turbine, and the electrolyzer. This refers to the cost of the energy storage system.

[0062] Various constraints, such as system energy flow balance constraints and gas turbine ramping constraints, regulate the operation of photovoltaic, gas turbine, electrolyzer, fuel cell and hydrogen energy storage equipment from different dimensions, ensure that each device works within a reasonable power range and under reasonable conditions, avoid system failure caused by equipment operating beyond its limits, and improve the stability and reliability of the integrated energy system operation.

[0063] With the goal of maximizing the overall system benefit, the system is guided to optimize the operation strategies of photovoltaic, gas turbine and other equipment while satisfying various constraints. This fully utilizes clean energy sources such as photovoltaics, rationally allocates the working status of equipment such as electrolyzers and fuel cells, maximizes the system's revenue from electricity and hydrogen sales and carbon emission reduction, and effectively controls the operation and maintenance costs of each piece of equipment and the cost of the energy storage system, thereby achieving efficient resource utilization and maximizing economic benefits.

[0064] Incorporating carbon emission reduction benefits into the objective function encourages the system to focus more on low-carbon operation, promotes the consumption of clean energy and the efficient and low-carbon utilization of traditional energy, helps reduce carbon emissions, and provides technical support for achieving the "dual carbon" goals of carbon peaking and carbon neutrality.

[0065] Furthermore, the expression for the lower-level constraint is: ; In the formula, , These are the upper and lower limits of the adjustable range for the user's flexible load, respectively. The objective function for determining electricity costs and perceived user comfort is expressed as follows: ; In the formula, for Real-time electricity price on the power grid; The actual electricity consumption of the user at time t; Let t be the user's perceived loss cost.

[0066] The lower-level constraints limit the range of the user's actual electricity consumption, keeping it within the user's flexible load adjustment range. This prevents the user's electricity consumption from exceeding their own adjustment capacity, ensuring that electricity consumption is within a reasonable and controllable range, and maintaining electricity order and stability.

[0067] The objective function that combines electricity costs with user perceived comfort aims to minimize electricity costs while considering user perceived loss costs. This effectively balances economic costs with user comfort, allowing users to have a better electricity experience while reasonably controlling their electricity expenditures, thereby improving user satisfaction with electricity services.

[0068] Such constraints and objective functions can guide users to arrange their electricity consumption time and amount more rationally, which helps the power grid to optimize the allocation and scheduling of power resources based on users' flexible load adjustment, thereby improving the overall operating efficiency and resource utilization efficiency of the power system.

[0069] In traditional dual-entity supply and demand game optimization, there is often a conflict between system economy and user demand. To address this contradiction, this invention proposes the "Energy-Service Synergy Index" (ESSI), whose core objective is to achieve a balance of interests among multiple stakeholders by integrating system economic benefits and user-perceived quality.

[0070] Furthermore, based on the population size and iteration conditions, iterative calculations are performed to output a comprehensive performance evaluation, the expression of which is: ; In the formula, , , These are revenues from electricity sales, hydrogen sales, and carbon emission reduction, respectively. , , , These are the operating and maintenance costs of the photovoltaic system, gas turbine, electrolyzer, and fuel cell, respectively. This is due to the loss of hydrogen production efficiency; To assess the perceived cost of loss for users.

[0071] The comprehensive performance evaluation formula takes into account various factors such as the revenue from electricity and hydrogen sales and carbon emission reduction, the operation and maintenance costs of equipment such as photovoltaics, gas turbines, electrolyzers, and fuel cells, as well as hydrogen production efficiency loss and user perceived loss costs. It can comprehensively and multidimensionally evaluate the overall benefits of the system, avoiding the one-sidedness of evaluating solely from the perspective of revenue or cost, and making the evaluation results more scientific and accurate.

[0072] This comprehensive performance evaluation provides a clear understanding of the revenue contribution and cost consumption of each link in the system, identifies the advantages and disadvantages in the system's operation, and provides a strong basis for subsequent optimization of the system in terms of equipment configuration and operation strategy adjustment, helping the system to develop in a more efficient and economical direction.

[0073] Incorporating carbon emission reduction benefits into the formula can effectively reflect the system's achievements in low-carbon development. At the same time, combining various costs can encourage the system to take into account low-carbon goals while pursuing economic benefits, achieving synergistic progress between economic and low-carbon development, which is in line with the current requirements for sustainable development such as "dual carbon".

[0074] Furthermore, based on the user load's flexible and adjustable range (such as temperature control equipment thresholds, energy storage device charging and discharging time windows, etc.), a user demand deviation tolerance is defined. When the user's actual electricity consumption Unable to meet the required value At this time, a perceived loss cost is incurred, and the expression for the user's perceived loss cost is: ; In the formula, For users to perceive the cost of loss; This represents the user's actual electricity consumption. Electricity consumption to meet user needs; The user perception loss coefficient during the low deviation phase; The user perception loss coefficient during the high deviation phase; Set a maximum value for the cost of perceived loss to the user.

[0075] By defining the tolerance level for deviations from user needs Based on the varying degrees of deviation between actual and demanded electricity consumption, the system constructs calculation expressions for perceived user loss costs across different time zones. This allows for the precise quantification of the experience loss incurred by users due to unmet electricity demands. From low deviation to high deviation, and then to severe deviation, different calculation methods correspond to different levels of loss, making the calculation of loss costs more closely aligned with actual user experiences. This provides more accurate user-side cost data for subsequent comprehensive performance evaluations.

[0076] This expression considers the flexible and adjustable range of user load (such as temperature control equipment thresholds and energy storage device charging and discharging time windows), reflecting a focus on user power comfort when calculating loss costs. Simultaneously, this quantification method provides a basis for system optimization strategies, prompting the system to minimize deviations from user power demand during power scheduling. This ensures efficient and economical overall system operation while guaranteeing user power comfort, achieving a balance between user experience and system optimization.

[0077] Perceived user cost is a significant component of user-side costs in the operation of integrated energy systems. Accurately quantifying it and incorporating it into relevant evaluation systems (such as comprehensive performance evaluation) can improve the cost-benefit assessment dimensions of integrated energy systems, making the assessment results more comprehensive and accurate, and thus better guiding the planning, operation, and optimization decisions of integrated energy systems.

[0078] Furthermore, the hydrogen efficiency loss cost (HELC) is defined as the cost incurred when the input power of the electrolyzer is lower than the minimum operating power of the electrolyzer. At that time, the penalty cost for starting and stopping the electrolytic cell is triggered. The hydrogen production efficiency loss caused by the power allocation strategy is quantified, and the expression for the hydrogen production efficiency loss is: ; In the formula, Let t be the power consumed by the electrolytic cell during operation at time t; This is the minimum operating power of the electrolytic cell; The cost of starting and stopping the electrolytic cell.

[0079] By defining the cost of hydrogen production efficiency loss and constructing an expression based on the relationship between the input power of the electrolyzer and the minimum operating power, the efficiency loss in hydrogen production caused by factors such as power allocation strategies can be accurately quantified. Representing this loss in the form of cost makes efficiency losses in the hydrogen production process measurable and assessable, providing a clear quantitative indicator for the efficiency analysis of the hydrogen production stage in integrated energy systems.

[0080] When the input power of the electrolyzer is lower than the minimum operating power, a start-stop penalty cost is triggered. This mechanism can guide the system to ensure that the electrolyzer operates within a reasonable power range when allocating power, avoiding frequent start-stop or inefficient operation of the electrolyzer due to excessively low power, thereby ensuring the stability and efficiency of the electrolyzer operation and improving the overall performance of the hydrogen production system.

[0081] Hydrogen production efficiency loss cost is an important component of the cost of hydrogen production in integrated energy systems. Incorporating it into the cost assessment system can improve the cost accounting dimensions of integrated energy systems, making cost assessments more comprehensive and accurate. This, in turn, provides a more reliable basis for the economic optimization and operational strategy adjustment of integrated energy systems, helping the system achieve synergistic optimization of economy and efficiency.

[0082] This invention constructs a dynamic game framework with grid operators as leaders and user groups as responders. The upper layer optimizes equipment operation and maintenance costs and power allocation strategies to form new electricity pricing strategies. The lower layer receives the new electricity pricing strategies transmitted from the upper layer and optimizes load demand based on electricity costs and user perceived comfort. Through interactive iteration, the model achieves global collaboration and improves the overall economic indicators of the electric-hydrogen coupling system.

[0083] By employing a two-dimensional discrete memristor hyperchaotic mapping to generate the initial population for the EFFO algorithm, the algorithm's global coverage of the solution space is enhanced, avoiding the randomness defects of the traditional EEFO algorithm. Furthermore, a chaotic perturbation factor, Chaos, is introduced into the interaction and hunting phases of the EFFO algorithm. Combined with the Lyapunov exponent, the algorithm's exploration and development weights are dynamically adjusted, avoiding the premature convergence problem of the traditional EEFO algorithm. This effectively avoids the local optimal solution situation that is prone to occur in the early stage of the algorithm, thus improving the solution efficiency of the traditional EEFO algorithm for high-dimensional problems.

[0084] This invention proposes an Energy Efficiency-Comfort Synergy Index (ESSI), which defines the user's perceived loss cost by considering the deviation between the user's actual power and the power demand, considers the hydrogen production efficiency loss cost caused by insufficient power allocation to hydrogen production equipment, and combines the operating costs and benefits of each piece of equipment to define the system's comprehensive energy efficiency ratio (CER). By integrating the system's economic benefits and the user's perceived quality, it further achieves a balance of interests among multiple stakeholders.

[0085] On the other hand, the present invention proposes an electronic device, characterized in that it includes: at least one processor; a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute any one of the power allocation methods for a hydrogen-electric coupling system based on the CEO-EEFO optimization algorithm.

[0086] On the other hand, the present invention proposes a medium, which is a computer storage medium storing a computer program, wherein the computer program, when executed by a processor, implements any one of the power allocation methods for a hydrogen-electric coupling system based on the CEO-EEFO optimization algorithm.

[0087] Computer storage media may be simply referred to as media. Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Dual Data SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchlink DRAM (SLDRAM), Rambus Direct RAM (RDRAM), Direct Memory Bus Dynamic RAM (DRDRAM), and Memory Bus Dynamic RAM (RDRAM). The various embodiments described in this specification are presented in a progressive manner, with reference allowed to each other for similar or identical parts. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments for apparatuses, devices, and non-volatile computer storage media are described simply because they are substantially similar to the method embodiments; relevant details can be found in the descriptions of the method embodiments.

[0088] The above embodiments are merely illustrative examples and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for power distribution of a hydrogen-electric coupling system based on a CEO-EEFO optimization algorithm, characterized in that, include: The population size, iteration conditions, and variable dimensions of the CEO-EEFO algorithm are determined based on the system variables to be solved. The system variables include the gas turbine's momentary output power, the electrolyzer's momentary operating power, the fuel cell's momentary output power, and the grid electricity price curve. The CEO-EEFO algorithm is initialized based on the input parameters to obtain the initialized CEO-EEFO algorithm. The input parameters include illumination condition data, user load data, and initial electricity price data. Based on the initialization CEO-EEFO algorithm, a two-layer dynamic model is constructed. The two-layer dynamic model includes an upper-layer grid operator model and a lower-layer user model. The upper-layer grid operator model includes upper-layer constraints and an objective function for optimizing the overall system revenue. The lower-layer user model includes lower-layer constraints and objective functions for electricity cost and user perceived comfort. Based on the aforementioned variable dimensions, the upper-level grid operator model is solved to obtain the electricity pricing strategy and power allocation strategy. The lower-level user model receives the electricity pricing strategy to adjust load demand and feeds it back to the upper-level grid operator model. Based on the population size and the iteration conditions, iterative calculations are performed to output the final power allocation strategy. The termination conditions include the maximum number of iterations or the Stackelberg equilibrium state.

2. The method of claim 1, wherein the method is based on a CEO-EEFO optimization algorithm. The CEO-EEFO algorithm is initialized based on the input parameters to obtain the initialized CEO-EEFO algorithm, which specifically includes: For the CEO-EEFO algorithm, an initial population is generated based on a two-dimensional discrete memristor hyperchaotic mapping to obtain a homogenized population; The initialization CEO-EEFO algorithm is obtained based on the homogenized population and the input parameters.

3. The method of claim 2, wherein the method is based on a CEO-EEFO optimization algorithm. The CEO-EEFO algorithm generates an initial population based on a two-dimensional discrete memristor hyperchaotic mapping to obtain a homogenized population, specifically including: two different original individuals are selected from the current eel population and and linearly mapping the original individuals to the specific basin of attraction of the chaotic map in the intervals [-0.5, 0.5] and [-0.25, 0.25]: ； wherein and is an eel individual , mapped chaotic initial position; , are upper and lower bounds of the search space, respectively The mapping results are and As initial values, N chaotic candidate individuals are generated by iterating N times: ； By performing an inverse mapping on the N chaotic candidate individuals, we obtain the actual search space of the original problem: ； based on the actual search space, the original individual 、 generate N evolution directions: ； Adopting Lyapunov exponent , the decision EEFO algorithm rest, migration and hunting measurement line energy factor , replaced by hyperchaotic energy factor ; ； wherein when , when , ; When Optimize interaction strategy: ； When , enhance the interaction strategy: ； wherein is the step size of the algorithm iteration search; , is the best solution in the population at the current iteration number.

4. The method of claim 1-3, wherein, The upper-level constraints include system energy flow balance constraints, gas turbine ramp-up constraints, electrolyzer operation constraints, fuel cell operation constraints, and hydrogen storage device SOC constraints. The expressions for the upper-level constraints are as follows: ； wherein , , are respectively photovoltaic device, gas turbine, fuel cell output power at time t; , are respectively grid load, electrolyzer device consumption power at time t; is the gas turbine output power difference; is the gas turbine device power ramping set upper limit; , are the minimum and maximum operating power limit parameters of the electrolyzer, respectively; is the rated power of the electrolyzer; is the rated operating power of the fuel cell power plant; is the electrolysis efficiency of the electrolyzer; is the specific energy consumption for hydrogen production, i.e. the power required to produce a unit mass of hydrogen; is the fuel cell power plant efficiency; is the energy density of hydrogen, i.e. the energy that can be released by a unit mass of hydrogen when completely converted; , is the hydrogen storage device is the state of the hydrogen storage capacity at the current and previous time instant; is the rated hydrogen storage capacity of the hydrogen storage device; , are the upper and lower limits of the state of the hydrogen storage capacity; The expression for the objective function of the system's overall optimal benefit is: ； wherein, , , are the system electricity, hydrogen and carbon emission reduction sales revenues, respectively; , , are the system photovoltaic, gas turbine and electrolyzer operation and maintenance costs, respectively; is the cost of the energy storage system.

5. The method of claim 4, wherein the method is based on a CEO-EEFO optimization algorithm. The expression for the lower-level constraint is: ； In the formula, , are respectively the upper and lower limits of the user flexible load adjustable interval; The objective function for determining electricity cost and perceived user comfort is expressed as follows: ； In the formula, is The electricity price of the power grid at time t; is the actual electricity consumption of the user at time t; is the user's perceived loss cost at time t.

6. The method of claim 1-3, wherein the method is a method of power distribution for a hydrogen power coupled system based on a CEO-EEFO optimization algorithm. Based on the population size and the iteration conditions, iterative calculations are performed, and a comprehensive performance evaluation is output. The expression for the comprehensive performance evaluation is: ； In the formula, , , Respectively, the system sells electricity, sells hydrogen and carbon emission reduction income; , , , Respectively, the system photovoltaic, gas turbine, electrolytic cell and fuel cell operation and maintenance cost; Hydrogen production efficiency loss; User perceived loss cost.

7. The method of claim 6, wherein the method is based on a CEO-EEFO optimization algorithm. The expression for the user-perceived loss cost is: ； In the formula, is the user perceived loss cost; is the user actual power consumption; is the user demand power consumption; is the low deviation stage user perceived loss coefficient; is the high deviation stage user perceived loss coefficient; sets a maximum value for the user perceived loss cost.

8. The method of claim 6, wherein the method is based on a CEO-EEFO optimization algorithm. The expression for the hydrogen production efficiency loss is: ； wherein P(t) is the power consumed by the electrolyzer at time t; Pmin is the minimum power at which the electrolyzer can operate; Pstart is the start-up penalty cost of the electrolyzer.

9. An electronic device, comprising: include: At least one processor; A memory that is communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the power allocation method for a hydrogen-electric coupling system based on the CEO-EEFO optimization algorithm as described in any one of claims 1 to 8.

10. A medium characterized by, The computer program is stored and is executed by the processor to realize the hydrogen-electric coupling system power distribution method based on the CEO-EEFO optimization algorithm in any one of claims 1 to 8.

Images