Multi-target level rolling optimization scheduling method and system for off-grid wind-solar green hydrogen production

Abstract

The invention discloses a multi-objective hierarchical rolling optimization scheduling method and system for off-grid wind-solar green hydrogen production. The method comprises the following steps: firstly, establishing a total power balance model of an off-grid wind-solar hydrogen production scene, secondly, respectively establishing charging and discharging models of an energy storage device, introducing integer state variables of 0 and 1 to uniformly describe the independent charging and discharging models, and then respectively setting physical constraints. And finally, by establishing a multi-target hierarchical optimization proposition and according to the model and constraint established in the previous step, real-time wind and light power generation power and future power prediction data, performing optimization solution by adopting a nonlinear programming algorithm mixed with 0-1 integers. The system researched and developed based on the method can update a scheduling plan in a rolling manner through a prediction model and optimization according to high and low priority optimization targets set by an operator based on real-time and predicted wind-solar power generation power data, so that the dynamic optimization of a scheduling strategy is ensured, and reliable operation guidance is provided for wind-solar hydrogen production scheduling personnel in real time.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy system optimization and scheduling, and particularly to a multi-objective hierarchical rolling optimization scheduling method and system for off-grid wind-solar hydrogen production using green hydrogen. Background Art

[0002] In the off-grid scenario of wind-solar hydrogen production, the system usually becomes an "island", operating independently of the power grid and relying entirely on the fluctuating power supply of wind power and photovoltaic power generation. However, both wind power and photovoltaic power generation have obvious randomness, intermittency, and volatility, which directly affect the stability of hydrogen production by electrolyzers. Since the off-grid system cannot rely on the power grid to balance the fluctuating supply and demand, the power balance and distribution scheduling of the system itself are particularly important. During the hydrogen production process, the electrolyzer requires a relatively stable power input, and power fluctuations may affect the hydrogen production efficiency and equipment life. The introduction of energy storage devices can alleviate the volatility to a certain extent, but it also further increases the complexity of the system. In addition, due to the fluctuating characteristics of wind-solar power generation and the instability of power demand during the hydrogen production process, there may be a situation where the power supply is greater than the demand, resulting in some generated electricity being abandoned due to the insufficient absorption capacity of energy storage and hydrogen production devices, which is significantly insufficient in terms of efficient resource utilization.

[0003] Therefore, in order to maximize energy utilization efficiency and hydrogen production, it is necessary to establish a multi-objective hierarchical optimization model based on the dynamic characteristics of wind-solar power generation, rationally allocate the power balance of wind-solar power generation, energy storage systems, and electrolyzers, and on the basis of safe operation, achieve the minimization of abandoned electricity, the maximization of hydrogen production, and the optimal economy and energy efficiency of system operation. Summary of the Invention

[0004] The purpose of the present invention is to overcome the deficiencies of the prior art and provide a multi-objective hierarchical rolling optimization scheduling method and system for off-grid wind-solar hydrogen production using green hydrogen, so as to overcome the deficiencies of the off-grid wind-solar hydrogen production system in the utilization rate of fluctuating new energy and the stability of hydrogen production, and realize the dynamic optimization and efficient coordination of the entire chain of wind-solar power generation and green hydrogen production.

[0005] The purpose of the present invention is achieved through the following technical solutions: Provide a multi-objective hierarchical rolling optimization scheduling method for off-grid wind-solar hydrogen production using green hydrogen, including the following steps: (1) Establish a total power balance model for the off-grid wind-solar hydrogen production scenario, and this total power balance model is used to describe the off-grid wind-solar hydrogen production scenario; (2) Establish a dynamic model of the energy storage device, and introduce variables to describe the independent charge and discharge models as a unified mathematical model; (3) Combine the requirements of the actual wind-solar hydrogen production scenario to set physical constraints for the decision variables of the energy storage device involved in the model; (4)Set the objectives for optimizing the coverage level in combination with the actual needs of the hydrogen production scenario using wind and solar power, including the minimum curtailment power, the maximum hydrogen production, and high energy utilization, and set priorities for the optimization objectives; (5)Read the real-time wind and solar power generation and the future multi-step power prediction data provided upstream through the industrial communication protocol; (6)According to the data read in step (5), the optimization proposition set in step (4), the model in step (2), and the physical constraints of each decision variable in step (3), use the mixed 0-1 integer non-linear programming algorithm for optimization and solution. The obtained optimized scheduling instructions contain multiple beats, but only the first beat will be executed to obtain the optimal scheduling decision instruction for the first beat; (7)Execute the first-beat optimized scheduling instruction in step (6). In the next optimization cycle, re-optimize and solve according to the latest hydrogen production feedback data to achieve continuous optimization.

[0006] Further, the total power balance model in step (1) is used to describe the off-grid wind and solar hydrogen production scenario, specifically including the wind and solar power generation, the charge and discharge power of the energy storage device, the electrolyzer power, the curtailment power, and the charge and discharge state of the energy storage device in the off-grid wind and solar hydrogen production scenario.

[0007] Further, step (1) specifically includes the following sub-steps: (1.1)Establish a total power balance model for the off-grid wind and solar hydrogen production system, incorporating the wind and solar power generation, the charge and discharge power and state of the energy storage system, the electrolyzer power, and the curtailment power into the model to comprehensively describe the interaction relationships between each power node; (1.2)Form a dynamic power balance mechanism covering wind and solar power generation, energy storage, and hydrogen production systems by setting the power flow conditions under different operating conditions, enabling the system to meet the balance conditions of power input and consumption at each moment.

[0008] Further, step (2) is specifically as follows: (2.1)Establish a dynamic charging model and a dynamic discharging model for the energy storage device respectively. With the introduced discrete 0, 1 integer state variables, the dynamic charging model and the dynamic discharging model of the energy storage device can simultaneously describe the dynamic charging and discharging behaviors; (2.2)Realize the seamless switching of the energy storage device between charging and discharging behaviors through the switching of the state variables of the energy storage device, while ensuring that the model can describe the change path of the system's charging and discharging behaviors in a unified mathematical model.

[0009] Furthermore, the constraint in step (3) is a physical security constraint in the actual scenario. Based on the physical characteristics of the energy storage system and the electrolyzer equipment, key constraint conditions are set, which are non-relaxable hard constraints. The decision variables covered by the energy storage device include the upper and lower limits of the state of charge, the charge and discharge rate limits, the electrolyzer input power limit, and the maximum power change allowed by the electrolyzer within the safe range. These constraints are all incorporated into the scheduling model as inequality conditions to ensure that the scheduling optimization operates within the safety boundaries.

[0010] Furthermore, step (4) specifically includes the following sub-steps: (4.1) Based on power balance and physical constraints, a multi-objective hierarchical optimization model is constructed, with the core objectives of minimizing the curtailed power, maximizing the hydrogen production, and maximizing the energy utilization. (4.2) By assigning weights to each optimization objective and setting a hierarchical optimization structure with prioritization levels, the multi-objective model realizes the hierarchical allocation of priorities to meet the differentiated requirements of the scheduling objectives.

[0011] Furthermore, step (5) is specifically as follows: The current power data of wind and solar power generation is collected in real time through an industrial communication protocol interface, and combined with the power generation prediction information of multiple future time steps as the input data for rolling optimization. The real-time data and prediction data are used as the time series variables of the model, enabling the system to schedule the energy storage charge and discharge and the electrolyzer input power in advance according to the changing trend of the power generation.

[0012] Furthermore, step (6) is specifically as follows: The multi-objective optimization model is solved using a mixed 0-1 integer nonlinear programming algorithm. Since the nonlinear optimization problem involves 0 and 1 integer variables, a mixed integer nonlinear programming algorithm needs to be used for solving. The obtained optimal scheduling instructions contain multiple beats, but only the first beat will be executed. This algorithm simultaneously considers the power changes and constraint conditions at future times, and adopts a step-by-step solution method to optimize the objective function value at each time step according to the input power data, thereby ensuring the best match between the current scheduling decision and the real-time data.

[0013] Further, step (7) is specifically as follows. In a new optimization cycle, the optimization algorithm needs to re-solve the mixed 0-1 integer non-linear programming problem according to the latest wind-solar power data, energy storage device status, power, and real-time operating power information of the electrolyzer, and still only execute the scheduling optimization instruction of the first beat; that is, in each rolling cycle, execute the optimal scheduling instruction of the first time step of the current cycle; adjust the operation instructions of devices such as the energy storage charge-discharge power and the electrolyzer input power by executing the first scheduling instruction, and at the same time collect the latest data and repeat the solution steps in the next optimization cycle to adjust the scheduling plan according to the latest system status and data; through rolling optimization, realize the dynamic adjustment of the scheduling plan, enable the system scheduling to respond to real-time data changes in a timely manner, and achieve continuous optimization.

[0014] The present invention also discloses a multi-objective hierarchical rolling optimization scheduling system for off-grid wind-solar hydrogen production, including the following units: Total power balance model establishment unit: Establish a total power balance model for the off-grid wind-solar hydrogen production scenario, and this total power balance model is used to describe the off-grid wind-solar hydrogen production scenario; Dynamic and mathematical model establishment unit: Establish a dynamic model of the energy storage device, and introduce variables to describe the independent charge and discharge models as a unified mathematical model; Physical constraint setting unit: Combine the requirements of the actual wind-solar hydrogen production scenario to set physical constraints for the decision variables of the energy storage device involved in the model; Optimization proposition setting unit: Combine the requirements of the actual wind-solar hydrogen production scenario to set the goals covering hierarchical optimization propositions, including the least discarded power, the maximum hydrogen production amount, and high energy utilization, and set priorities for the optimization goals; Power prediction unit: Read the real-time wind-solar power generation power and the future multi-step power prediction data provided by the upstream through the industrial communication protocol; Optimization solution unit: According to the data read by the power prediction unit, the optimization propositions set in the optimization proposition setting unit, the model established by the dynamic and mathematical model establishment unit, and the physical constraints of each decision variable in the physical constraint setting unit, use the non-linear programming algorithm of mixed 0-1 integer to perform optimization solution. The obtained optimization scheduling instruction contains multiple beats, but only the first beat will be executed to obtain the optimal scheduling decision instruction of the first beat; Continuous optimization unit: Execute the first-beat optimization scheduling instruction in the optimization solution unit, and in the next optimization cycle, re-optimize and solve according to the latest hydrogen production feedback data to achieve continuous optimization.

[0015] The beneficial effects of the present invention are as follows: The method of the present invention is based on the real-time power and prediction data of wind and solar power generation, and intelligently and dynamically adjusts the charging and discharging power, state of charge of the energy storage system, and input power of the electrolyzer to achieve dynamic balance and efficient scheduling in green hydrogen production. The designed rolling optimization mechanism in the method can ensure that the system quickly responds to the fluctuations of wind and solar power generation, reduce curtailment losses, increase hydrogen production, and optimize energy utilization efficiency. Finally, it realizes the optimal scheduling decision in terms of economy and energy efficiency, providing strong technical support for off-grid green hydrogen applications. Description of the Drawings

[0016] By describing the preferred embodiments of the present invention in conjunction with the following drawings, the objects, features, and advantages of the present invention can be further understood. The present invention will be described in more detail with reference to the drawings of the present invention below. However, the present invention can be implemented in many different forms, so it should not be considered limited to the embodiments listed in the specification. On the contrary, providing such embodiments is to illustrate the implementation and completeness of the present invention, and to describe the specific implementation process of the present invention to those skilled in the art.

[0017] Figure 1 It is the algorithm logic framework diagram of the present invention; Figure 2 It is the power prediction data of new energy wind and solar power generation; Figure 3 It is the case simulation result diagram of the present method. Detailed Embodiments

[0018] The present invention will be described in detail below with reference to the drawings.

[0019] The present invention discloses a multi-objective hierarchical rolling optimization scheduling method and system for supporting off-grid wind and solar power to produce green hydrogen, as Figure 1 shown, including the following steps: Step 1: Construct a total system power balance model For the off-grid wind and solar power to hydrogen production scenario, a comprehensive power balance model including key nodes such as wind and solar power generation power, energy storage system, electrolyzer, and curtailment amount is designed. In this model, the fluctuations of wind power and photovoltaic power generation are dynamically balanced through the charging and discharging behavior of the energy storage system to maintain a stable power input. The balance model also introduces a curtailment variable of the system to describe the remaining power flow under the condition of power surplus in the system. The expression of this model is: ; Where represents time, is the new energy power generation, is the discharging power of the energy storage device, is the power consumed by the electrolyzer, is the charging power of the energy storage device, is the curtailment power, satisfying: ; Since the battery energy storage device cannot charge and discharge simultaneously, state variables are introduced , , which respectively represent the charging and discharging states of the battery. Further, the total power balance update model in formula (1) can be modified as: ; where , are binary variables that are either 0 or 1 respectively. In the charging state , ; in the discharging state, , . This model ensures that the system power flow achieves a closed-loop balance under different power input conditions.

[0020] Step 2: Construct a unified mathematical model for the charging and discharging behavior of energy storage: Construct the charging and discharging mathematical models of the energy storage system, and merge them into a single model expression by introducing 0, 1 integer state variables. At each time step, the charging and discharging behavior is regulated by the state variables to achieve seamless switching between the charging and discharging modes of the energy storage system, so that a single model is applicable to both the charging and discharging scenarios. The expressions for the charging, discharging, and unified charging and discharging models of the energy storage device are as follows, where the charging model is: ; The discharging model is: ; After introducing the 0, 1 state variables , the unified charging and discharging model is: ; where represents the state of charge of the energy storage device at the current time step , which is expressed as a percentage or a ratio relative to the total capacity of the battery and is used to describe the remaining battery charge; represents the time interval at the current time step , which is used to standardize the impact of the charging and discharging power on the state of charge; represents the rated capacity of the energy storage device at the current time step , which represents the total stored energy of the battery when it is fully charged. and respectively represent the charging and discharging efficiencies of the energy storage device, which are used to reflect the loss degree of energy conversion during the charging and discharging processes.

[0021] Step 3: Set the physical constraints of the energy storage and hydrogen production equipment; Combined with the characteristics of system equipment, key physical constraints are set for the energy storage device and the electrolyzer to ensure the safe and stable operation of the system. Specifically, it includes: Charge and discharge rate constraints of the energy storage device: ; where respectively represent the lower and upper limit constraints of the charging and discharging power.

[0022] State of charge constraint of the energy storage device: ; where respectively represent the lower and upper limit constraints of the state of charge of the energy storage device.

[0023] Input power limit of the electrolyzer: ; where respectively represent the lower and upper limit constraints of the power allowed under the safe operating state of the electrolyzer.

[0024] Input power fluctuation limit of the electrolyzer: ; where respectively represent the lower and upper limit constraints of the power fluctuation allowed under the safe operating state of the electrolyzer.

[0025] Step 4: Construct a multi-stage hierarchical optimization objective: Based on the power balance model and physical constraints, a multi-objective optimization model is constructed with the minimum curtailment rate, maximum hydrogen production, and maximized energy utilization efficiency as the core objectives; among them, the optimization objective is to minimize the new energy curtailment power in the future hours, and on this basis, maximize the power allocated to the electrolyzer (the power used for electrolytic water hydrogen production in the future is the most), then the corresponding objective function shows an optimization scenario, in the following form: ; ; where represents the sampling frequency, represents the objective of minimizing the curtailment power, represents the objective of maximizing the electrolyzer power generation, represents the current moment, represents the optimization time domain, and from the optimization objective function expressions and a hierarchical optimization problem is formed. The priority of the objective function expression is higher than that of the objective function expression , and solve the objective function expression The obtained solution is used as the objective function expression as the constraint representing the decision variables, the optimized objective function expression is as follows: ; In addition, the optimization objective of this step in the scheduling system can be adjusted at any time according to on-site requirements, including priorities.

[0026] Step Five: Real-time and predicted data collection of upstream wind and solar power generation Real-time collection of the current wind and solar power generation through industrial communication protocols (common protocols such as OPC and Modbus can be used), and obtain multi-step power prediction data for the future. The prediction data is derived from the power estimation of the upstream model and combined with real-time data as the key data input for rolling optimization scheduling. The future prediction data enables the model to have the ability to perceive future trends, making the power scheduling of energy storage and electrolyzers more predictive and forward-looking.

[0027] Step Six: Solve the rolling optimization algorithm; Use the mixed 0-1 integer nonlinear programming method to solve the hierarchical optimization problem in Step Four. In the current cycle, generate the scheduling instruction set for the current cycle based on the power prediction and real-time data of multiple future time steps, and only execute the optimal scheduling plan for the first time step. This method ensures the close connection between the scheduling strategy and the actual system requirements and performs real-time optimization through rolling updates. Taking the optimization objectives in formulas and as an example, the optimization proposition solved by the optimization algorithm can be expressed in the following form: Solution of the first-stage optimization problem: ; Denote the optimal objective of the first stage as , then the optimization proposition of the second stage is: ; ; Denote the decision instructions for multiple future steps of solving the above hierarchical optimization problem as : ; where is the scheduling decision instruction for actual execution. It should be noted that if the above multi-stage optimization problem can be successfully solved, the optimal instruction of the last stage is returned; if the solution of a certain stage fails, the optimal instruction of the previous stage is output; if the problems in the first stage all fail to be solved, the optimal instruction of the previous optimization moment is held at zero order.

[0028] Step Seven: Execute the scheduling and perform rolling updates; Execute the scheduling instructions for the first time step within each rolling period, and in the next period, re-solve the optimization problem based on the latest information such as the power generation data of wind and light, the energy storage state, and the power demand of the electrolyzer, and generate new scheduling instructions. Through rolling optimization, the system can update the scheduling decision in real time within each period and quickly respond to the fluctuations in wind and light power generation.

[0029] As Figure 2 shown, this figure is the power generation curve of wind and light of new energy read from the upstream. The curve shows the predicted data for the next 24 hours, which is given by the upstream (this figure takes photovoltaic power generation as an example). Figure 3 This is a simulation result figure, which shows the charging power curve, discharging power curve, electrolyzer power curve, curtailed power curve of the energy storage device, and the state of charge curve of the energy storage device scheduled by the present invention based on the new energy power generation data. From Figure 2 and Figure 3 it can be seen that during the low valley period of new energy power generation from 0:00 to 5:00, the system maintains operation by discharging the energy storage. At this time, the SOC (state of charge of the energy storage device) shows a reasonable downward trend and no curtailment occurs; while during the peak power generation period around 12:00, the system preferentially distributes the surplus electric energy to the electrolyzer and starts charging the energy storage, causing the SOC to rise steadily, and only a small amount of curtailment occurs when the power generation exceeds the system's consumption capacity. This dynamic scheduling mechanism realizes three optimization goals: firstly, minimizing the curtailment rate through precise control of the energy storage charging and discharging; secondly, maximizing the electrolyzer operation power during the abundant power generation period to improve the hydrogen production efficiency; finally, ensuring the continuous power supply capacity of the energy storage system through scientific management of the SOC. This shows that the present invention can reasonably schedule the power of each device in the hydrogen production system (the charging and discharging power of the energy storage device, the electrolyzer power) according to the wind and light power generation of new energy, so as to minimize the curtailment of new energy wind and light, and on this basis, maximize the power allocated to the electrolyzer.

[0030] The present invention also discloses a multi-objective hierarchical rolling optimization scheduling system for off-grid wind and light hydrogen production, including the following units: Unit for establishing the total power balance model: Establish the total power balance model for the off-grid wind and light hydrogen production scenario, and this total power balance model is used to describe the off-grid wind and light hydrogen production scenario; Unit for establishing the dynamic and mathematical model: Establish the dynamic model of the energy storage device, and introduce variables to describe the independent charging and discharging models as a unified mathematical model; Unit for setting physical constraints: Combine the requirements of the actual wind and light hydrogen production scenario to set physical constraints for the decision variables of the energy storage device involved in the model; Set up an optimization proposition unit: Combine the actual demand of the wind-solar hydrogen production scenario, set the goals for covering hierarchical optimization propositions, including the minimum curtailment power, the maximum hydrogen production, and high energy utilization, and set priorities for the optimization goals; Power prediction unit: Read the real-time wind-solar power generation power and the multi-step power prediction data provided by the upstream through the industrial communication protocol; Optimization solution unit: According to the data read by the power prediction unit, the optimization propositions set in the set optimization proposition unit, and the model established in the dynamic and mathematical model unit and the physical constraints of each decision variable in the set physical constraint unit, use the mixed 0-1 integer non-linear programming algorithm for optimization solution. The obtained optimization scheduling instruction contains multiple beats, but only the first beat will be executed to obtain the optimal scheduling decision instruction for the first beat; Continuous optimization unit: Execute the first-beat optimization scheduling instruction in the optimization solution unit. In the next optimization cycle, re-optimize and solve according to the latest hydrogen production feedback data to achieve continuous optimization.

[0031] This method is solid and innovative, with a clear step structure, easy to be programmed and implemented in a computer, and has strong flexibility, capable of adapting to a variety of off-grid wind-solar hydrogen production application scenarios. This embodiment is one of the preferred implementation manners of the present invention, but does not limit the present invention. For those skilled in the art, without departing from the spirit of the present invention, various equivalent changes and improvements can be made, and these changes and improvements are all within the protection scope of the present invention.

Claims

1. A multi-objective hierarchical rolling optimization scheduling method for off-grid wind-solar green hydrogen production, characterized in that: The following steps are involved: (1) Establishing a total power balance model for an off-grid wind-solar hydrogen production scenario, wherein the total power balance model is used to describe the off-grid wind-solar hydrogen production scenario; (2) Establish a dynamic model of the energy storage device and introduce variables to describe the independent charging and discharging models into a unified mathematical model; (3) In combination with the actual requirements of wind-solar hydrogen production scenarios, physical constraints are set for the decision variables of the energy storage devices involved in the model; (4) Based on the actual needs of wind and solar hydrogen production scenarios, set goals for coverage level optimization propositions, including minimum power abandonment, maximum hydrogen production, and high energy utilization, and set priorities for optimization goals; (5) Read the real-time wind and solar power generation and the future multi-step power forecast data provided by the upstream through the industrial communication protocol; (6) Based on the data read in step (5), the optimization proposition set in step (4), the model in step (2), and the physical constraints of each decision variable in step (3), a mixed 0-1 integer nonlinear programming algorithm is used to perform optimization and solution. The optimized scheduling instruction obtained contains multiple beats, but only the first beat will be executed, and the optimal scheduling decision instruction for the first beat is obtained; (7) Execute the first optimization scheduling instruction in step (6), and in the next optimization cycle, re-optimize and solve according to the latest hydrogen production feedback data to achieve continuous optimization.

2. According to claim 1, a multi-objective hierarchical rolling optimization scheduling method for off-grid wind-solar green hydrogen production is characterized in that: The total power balance model in step (1) is used to describe the off-grid wind-solar hydrogen production scenario, specifically including the wind-solar power generation power, the charging and discharging power of the energy storage device, the electrolyzer power, the abandoned power and the charging and discharging status of the energy storage device in the off-grid wind-solar hydrogen production scenario.

3. The multi-objective hierarchical rolling optimization scheduling method for off-grid wind-solar green hydrogen production according to claim 1 is characterized in that: The step (1) specifically includes the following sub-steps: (1.1) Establish a total power balance model for the off-grid wind-solar hydrogen production system, incorporating wind-solar power generation, the charging and discharging power and status of the energy storage system, the electrolyzer power and the abandoned power into the model to fully describe the interaction between various power nodes; (1.2) By setting the power flow conditions under different operating conditions, a dynamic power balance mechanism covering wind and solar power generation, energy storage and hydrogen production systems is formed, so that the system meets the balance conditions of power input and consumption at all times.

4. The multi-objective hierarchical rolling optimization scheduling method for off-grid wind-solar green hydrogen production according to claim 1 is characterized in that: The step (2) is specifically as follows: (2.1) Establishing a dynamic charging model and a dynamic discharging model of the energy storage device respectively. The dynamic charging model and the dynamic discharging model of the energy storage device can simultaneously describe the dynamic charging and discharging behaviors by introducing discrete 0 and 1 integer state variables; (2.2) By switching the state variables of the energy storage device, the energy storage device can seamlessly switch between charging and discharging behaviors, while ensuring that the model can describe the changing path of the system's charging and discharging behavior in a unified mathematical model.

5. The multi-objective hierarchical rolling optimization scheduling method for off-grid wind-solar green hydrogen production according to claim 1 is characterized in that: The constraints in step (3) are physical safety constraints in actual scenarios, which are based on the physical characteristics of the energy storage system and the electrolyzer equipment, and then set key constraints. They are hard constraints that cannot be relaxed, and the decision variables covering the energy storage device include upper and lower limits of the state of charge, charge and discharge rate limits, electrolyzer input power limits, and the maximum power change allowed for the electrolyzer within a safe range. These constraints are incorporated into the scheduling model as inequality conditions to ensure that the scheduling optimization operates within the safety boundary.

6. The multi-objective hierarchical rolling optimization scheduling method for off-grid wind-solar green hydrogen production according to claim 1 is characterized in that: The step (4) specifically includes the following sub-steps: (4.1) Based on power balance and physical constraints, a multi-objective hierarchical optimization model is constructed with the core objectives of minimizing curtailed power, maximizing hydrogen production and high energy utilization; (4.2) By assigning weights to each optimization objective and setting a priority-layered optimization structure, the multi-objective model can achieve hierarchical allocation of priorities to meet the differentiated needs of scheduling objectives.

7. The multi-objective hierarchical rolling optimization scheduling method for off-grid wind-solar green hydrogen production according to claim 1 is characterized in that: The step (5) is specifically as follows: the current power data of wind and solar power generation is collected in real time through the industrial communication protocol interface, and the power generation forecast information of multiple future time steps is combined as the input data of rolling optimization; the real-time data and the forecast data are used as the time series variables of the model, so that the system can schedule the energy storage charging and discharging and the electrolyzer input power in advance according to the power generation power change trend.

8. The multi-objective hierarchical rolling optimization scheduling method for off-grid wind-solar green hydrogen production according to claim 1 is characterized in that: The step (6) is specifically as follows: using a mixed 0-1 integer nonlinear programming algorithm to solve the multi-objective optimization model. Since the nonlinear optimization proposition involves 0 and 1 integer variables, it is necessary to use a mixed integer nonlinear programming algorithm to solve it. The solved optimization scheduling instruction contains multiple beats, but only the first beat will be executed. The algorithm also considers the power changes and constraints at future times, and adopts a step-by-step solution method to optimize the objective function value in each time step according to the input power data, thereby ensuring the best match between the current scheduling decision and the real-time data.

9. The multi-objective hierarchical rolling optimization scheduling method for off-grid wind-solar green hydrogen production according to claim 1 is characterized in that: Specifically, the step (7) is as follows: in a new optimization cycle, the optimization algorithm needs to re-solve the mixed 0-1 integer nonlinear programming problem based on the latest wind and solar power data, the energy storage device status, power, and the real-time operating power information of the electrolyzer, and still only execute the scheduling optimization instruction of the first beat; that is, in each rolling cycle, execute the optimal scheduling instruction of the first time step of the current cycle; adjust the equipment operation instructions such as the energy storage charging and discharging power and the electrolyzer input power by executing the first scheduling instruction, and at the same time re-collect the latest data in the next optimization cycle and repeat the solution steps, so as to adjust the scheduling plan according to the latest system status and data; through rolling optimization, the scheduling plan is dynamically adjusted, so that the system scheduling can respond to the real-time data changes in a timely manner and achieve continuous optimization.

10. A multi-objective hierarchical rolling optimization scheduling system for off-grid wind and solar green hydrogen production, characterized in that: The following units are included: Establishing a total power balance model unit: establishing a total power balance model for an off-grid wind-solar hydrogen production scenario, wherein the total power balance model is used to describe the off-grid wind-solar hydrogen production scenario; Establish dynamic and mathematical model units: Establish a dynamic model of the energy storage device and introduce variables to describe the independent charging and discharging models into a unified mathematical model; Setting physical constraint unit: Based on the actual requirements of wind-solar hydrogen production scenarios, set physical constraints for the decision variables of the energy storage devices involved in the model; Setting optimization proposition unit: Based on the actual needs of wind and solar hydrogen production scenarios, set the goals of coverage-level optimization propositions, including minimum power abandonment, maximum hydrogen production, and high energy utilization, and set priorities for the optimization goals; Power prediction unit: reads the real-time wind and solar power generation power and the future multi-step power prediction data provided by the upstream through the industrial communication protocol; Optimization solving unit: Based on the data read by the power prediction unit, the optimization proposition set in the optimization proposition unit, the model of the dynamic and mathematical model units, and the physical constraints of each decision variable in the physical constraint unit, the optimization solution is performed using a mixed 0-1 integer nonlinear programming algorithm. The solved optimization scheduling instruction contains multiple beats, but only the first beat will be executed, and the optimal scheduling decision instruction for the first beat is obtained; Continuous optimization unit: executes the first-shot optimization scheduling instruction in the optimization solution unit, and in the next optimization cycle, re-optimizes and solves according to the latest hydrogen production feedback data to achieve continuous optimization.

Images