A multi-time scale optimization scheduling method and system for hydrogen-electricity coupled microgrid

Through the combination of multi-time scale optimization scheduling method and dynamic window method, the slow convergence and local optimization problems of optimization scheduling algorithm in the hydrogen-electric coupled microgrid are solved, and efficient energy management of the hydrogen-electric coupled microgrid is realized, reducing energy costs and carbon emissions.

CN118783550BActive Publication Date: 2024-12-06SHANDONG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411251866.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-09
Publication Date
2024-12-06
Estimated Expiration
2044-09-09

AI Technical Summary

Technical Problem

When facing complex and changing hydrogen-electric coupled microgrids, the existing optimization scheduling algorithms have slow convergence speed, large calculation amount, and are prone to fall into local optimality, and are unable to effectively respond to the dynamic changes of the system and fluctuations in energy demand, resulting in insufficient system performance optimization.

Method used

Using multi-time scale optimization scheduling method, energy procurement and equipment scheduling are optimized through seasonal hydrogen storage optimization, recent economic scheduling and intraday dynamic optimization, combined with expert experience and knowledge base, and using a variable neighborhood search algorithm of dynamic window method to optimize energy procurement and equipment scheduling to achieve Energy supply and demand balance.

Benefits of technology

It improves optimization speed and accuracy, can dynamically adjust the system operation, adapt to real-time changes, achieve optimal operation of hydrogen-electrically coupled microgrids, reduce energy costs and carbon emissions, and improve the economic and environmental benefits of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118783550B_ABST
    Figure CN118783550B_ABST
Patent Text Reader

Abstract

The present invention proposes a multi-time-scale optimization scheduling method and system for a hydrogen-electricity coupled microgrid, which relates to the field of hydrogen-electricity coupling technology, including: according to the topological structure of the hydrogen-electricity coupled microgrid, a source-storage-load multi-time-scale coordinated optimization scheduling model is established, which is composed of seasonal hydrogen storage optimization with gradually shortened time scales, day-ahead economic scheduling and intraday dynamic optimization; on the basis of summarizing and analyzing historical optimization results, expert experience and best practices are introduced to establish an optimization strategy knowledge base to form constraint conditions for the optimization scheduling model; a variable neighborhood search solution algorithm based on a dynamic window method is adopted to solve the optimization scheduling model to obtain an optimal energy procurement and equipment scheduling plan for the optimal operation of the hydrogen-electricity coupled microgrid; when the present invention is connected to the main power grid, the balance between energy supply and load demand is optimized to achieve a balance between economy and environmental friendliness.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of hydrogen-electricity coupling, and in particular relates to a multi-time-scale optimization scheduling method and system for a hydrogen-electricity coupling microgrid. Background Art

[0002] With the development of the economy and society, energy conservation and emission reduction have become an irreversible trend in the energy field. Establishing an efficient, low-carbon, clean energy supply method has always been the focus of research. As an emerging energy system, the hydrogen-electricity coupled microgrid has received widespread attention. The hydrogen-electricity coupled microgrid refers to an energy network in which hydrogen energy and electric energy are converted into each other and efficiently coordinated. By organically combining the power system and the hydrogen energy system, efficient energy utilization and supply and demand balance are achieved. How to choose appropriate energy procurement and scheduling strategies to minimize energy costs and carbon emissions and improve the economic and environmental benefits of the system under the premise of meeting load demand is a difficult point.

[0003] In the hydrogen-electricity coupled microgrid, the production and consumption of energy are time-varying, and the conversion and storage between different energy sources also need to be adjusted in real time to adapt to the changes in load and renewable energy supply. Compared with the power supply chain, the hydrogen supply chain is more complex. Unlike the traditional electric energy microgrid, the preparation, transmission, storage and use of hydrogen have certain dynamic response time. In the optimal scheduling, the system operation plan needs to be dynamically adjusted to adapt to different environmental conditions and load requirements, and provide stable energy output. Therefore, the optimal scheduling problem of the hydrogen-electricity coupled microgrid needs to consider the operating status and energy conversion process of different equipment, which makes the traditional optimization model no longer applicable. It is necessary to consider the dynamic model of the equipment to accurately describe the energy conversion and scheduling process of the system, and dynamically optimize and adjust the system according to the energy demand and supply at any time. In addition, the energy supply and demand of the hydrogen-electricity coupled microgrid have complex fluctuations and uncertainties in time and space, which makes the collaborative optimization scheduling problem of the hydrogen-electricity coupled microgrid challenging. However, as the complexity of the model increases, it also poses challenges to the solution. In order to solve the problem of optimal scheduling of hydrogen-electricity coupled microgrids and improve the system operation efficiency, seeking an efficient optimization strategy and solution method is one of the current research hotspots.

[0004] The existing optimization scheduling solution algorithms often have problems such as slow convergence, large amount of calculation, long time consumption, and easy to fall into local optimality when facing complex and changing dynamic systems; with the increase of problem scale and complexity, the difficulty and computational complexity of establishing and solving mathematical models will also increase rapidly; mathematical programming methods are usually optimized at a discrete moment, which makes it difficult for traditional mathematical programming methods to be directly applied to the optimization of hydrogen-electric coupled microgrids. It is necessary to divide the problem into smaller time steps and jointly optimize the decisions of each time step; however, for some optimization problems with non-convex objective functions and non-differentiable points, the application effect of this method is not good; Patent application CN202210541150 proposes a method for optimizing the operating mode of a hydrogen-electric coupled system based on a mixed integer linear programming (MILP) model, which uses piecewise affine transformation to linearize the polarization curve of the hydrogen energy component. Although this processing method enables the mathematical programming method to be applied to the hydrogen energy system, it makes the model unable to reflect the dynamic characteristics of the hydrogen energy component, resulting in the optimization results based on this model deviating from the optimal solution, affecting the optimization of system performance.

[0005] Although traditional heuristic optimization algorithms can take dynamic change characteristics into account through iterative optimization, they often require a long iteration cycle to converge to a better solution, resulting in a slow solution speed. Hydrogen-electricity coupled microgrids usually require real-time or near real-time updates of operating plans, and traditional heuristic algorithms cannot meet system requirements.

[0006] In addition, the objectives and constraints of the hydrogen-electricity coupled microgrid optimization scheduling problem will change over time, which makes the search space have time-varying and complex characteristics, which makes traditional heuristic algorithms easily fall into local optimality and unable to find better results.

[0007] Therefore, the existing optimization algorithms are not adaptable and cannot cope with the complexity of the hydrogen-electricity coupled microgrid scheduling problem, nor can they optimize and adjust the equipment in the system according to the real-time energy demand and supply situation and the dynamic changes of the system status. Summary of the invention

[0008] In order to overcome the deficiencies of the above-mentioned prior art, the present invention provides a multi-time-scale optimization scheduling method and system for a hydrogen-electricity coupled microgrid, which optimizes the balance between energy supply and load demand when connected to the main power grid, thereby achieving a balance between economy and environmental friendliness.

[0009] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:

[0010] A first aspect of the present invention provides a multi-time scale optimization scheduling method for a hydrogen-electricity coupled microgrid.

[0011] A multi-time scale optimization scheduling method for a hydrogen-electricity coupled microgrid, comprising:

[0012] Obtaining the topological structure of the hydrogen-electricity coupled microgrid to be optimized;

[0013] According to the topological structure, an optimization scheduling model of source-storage-load multi-time scale coordination is established, which consists of seasonal hydrogen storage optimization with gradually shortened time scales, day-ahead economic scheduling and intraday dynamic optimization.

[0014] Based on the summary and analysis of historical optimization results, expert experience and best practices are introduced to establish an optimization strategy knowledge base and form the constraints of the optimization scheduling model;

[0015] The optimization scheduling model is solved by a variable neighborhood search solution algorithm based on a dynamic window method to obtain the optimal energy procurement and equipment scheduling plan, which is used to achieve the optimal operation of the hydrogen-electricity coupled microgrid under fluctuations in energy supply and demand.

[0016] Furthermore, the seasonal hydrogen storage optimization is a long-term optimization with an annual cycle and a monthly scale. By introducing seasonal grid price factors, the grid operation pressure in different seasons is characterized and evaluated. The goal is to minimize the annual operating cost. Based on the source and load data throughout the year, the seasonal hydrogen storage and release plan is optimized, including the hydrogen storage input and output power for each month.

[0017] Furthermore, the day-ahead economic dispatch is a medium-time scale optimization, with a day as the cycle and an hour as the scale. It optimizes the input and output plans of electrolyzers, fuel cells, hydrogen energy storage and electric energy storage while meeting load demand, and minimizes energy production costs while ensuring daily income.

[0018] Furthermore, the intraday dynamic optimization is a short-time scale optimization, with an hourly cycle and a 5-minute scale. When photovoltaic power generation and load demand fluctuate, it optimizes energy procurement and equipment scheduling plans, adjusts the real-time output of electrolyzers, fuel cells, hydrogen energy storage and electric energy storage, so that renewable energy, hydrogen energy and electricity demand are more matched, thereby achieving real-time energy balance within the microgrid.

[0019] Furthermore, the optimization strategy knowledge base is formed by integrating the optimization result data set and the expert knowledge data set;

[0020] The optimization result data set is a large amount of load demand and renewable energy power generation data generated through long-term monitoring and simulation. On this basis, optimization simulation is performed under different energy supply and demand scenarios, the optimization results of each scenario are recorded, and the scene features are analyzed and selected from the optimization results to form the optimization result data set;

[0021] The expert knowledge data set converts scheduling rules and historical experience into a computable parameter set, quantifies expert knowledge, and integrates the quantified expert knowledge into the optimization model to formulate system scheduling constraints to form an expert knowledge data set.

[0022] Furthermore, the variable neighborhood search solution algorithm is based on the day-ahead scheduling plan curve, determines the maximum variable range of the corresponding decision variables, forms an optimization window at the current moment, and selects the energy procurement and equipment scheduling plan with the highest evaluation score within the optimization window.

[0023] Furthermore, the determination of the maximum variable range of the corresponding decision variables to form the optimization window at the current moment is a dynamic window method, which adjusts and limits the parameter search space according to the real-time system status and dynamic response capability, and then dynamically reduces the feasible domain to generate a new feasible domain as the optimization window at the current moment.

[0024] A second aspect of the present invention provides a multi-time scale optimization scheduling system for a hydrogen-electricity coupled microgrid.

[0025] A multi-time scale optimization dispatching system for a hydrogen-electricity coupled microgrid includes an acquisition module, a modeling module, a constraint module and a solution module:

[0026] The acquisition module is configured to: acquire the topological structure of the hydrogen-electricity coupled microgrid to be optimized;

[0027] The modeling module is configured to: establish a source-storage-load multi-time scale coordinated optimization scheduling model based on the topological structure, which consists of seasonal hydrogen storage optimization with gradually shortened time scales, day-ahead economic scheduling, and intraday dynamic optimization;

[0028] The constraint module is configured to: introduce expert experience and best practices based on summarizing and analyzing historical optimization results, establish an optimization strategy knowledge base, and form constraint conditions for the optimization scheduling model;

[0029] The solution module is configured to: use a variable neighborhood search solution algorithm based on a dynamic window method to solve the optimization scheduling model to obtain an optimal energy procurement and equipment scheduling plan for achieving optimal operation of the hydrogen-electricity coupled microgrid under fluctuations in energy supply and demand.

[0030] A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps in a multi-time-scale optimization scheduling method for a hydrogen-electricity coupled microgrid as described in the first aspect of the present invention.

[0031] The fourth aspect of the present invention provides an electronic device, including a memory, a processor, and a program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the multi-time-scale optimization scheduling method for a hydrogen-electricity coupled microgrid as described in the first aspect of the present invention are implemented.

[0032] One or more of the above technical solutions have the following beneficial effects:

[0033] The present invention proposes a data- and knowledge-driven hybrid intelligent optimization strategy. On the basis of summarizing and analyzing historical optimization results, it introduces expert experience and best practices, formulates decision rules, and establishes an optimization strategy knowledge base, so as to improve the optimization speed and ensure the accuracy of optimization. A multi-time scale collaborative optimization scheduling framework is established, and through seasonal hydrogen storage, day-ahead economic scheduling and intraday dynamic optimization, the fluctuations of various energy inputs and load demands are balanced to achieve long-term and short-term supply and demand matching. The seasonal power grid price factor is introduced to characterize and evaluate the power grid operation pressure in different seasons, and according to the supply and demand relationship throughout the year, a seasonal hydrogen storage / desorption strategy is formulated through a sequential quadratic programming algorithm to solve the seasonal supply and demand mismatch problem. The dynamic window method (DWA) is applied to the optimal value search process, and a variable neighborhood search algorithm is developed. The optimization window is continuously adjusted, the feasible domain is dynamically reduced, and the real-time changes of the system are adapted to improve the solution efficiency and achieve optimal operation under fluctuating input.

[0034] The application of the method of the present invention has important industry contributions, providing a feasible and efficient solution approach, which can solve the optimization scheduling problem of hydrogen-electricity coupled microgrids; the method can improve the energy utilization efficiency of the system, reduce energy costs, and reduce carbon emissions, provide technical support for the development and application of the hydrogen energy industry, and help promote the development and promotion of the hydrogen energy industry; in addition, the method will promote technological innovation in the energy field, and also play an important role in promoting the transformation and sustainable development of the energy industry.

[0035] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0037] Figure 1 This is a multi-time scale optimization framework diagram of the first embodiment.

[0038] Figure 2 This is a flow chart of the optimization solution algorithm of the first embodiment.

[0039] Figure 3 This is a flowchart of dynamically reducing the feasible domain of the first embodiment. DETAILED DESCRIPTION

[0040] It should be noted that the following detailed descriptions are illustrative and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used in the present invention have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0041] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0042] Embodiment 1

[0043] In one embodiment of the present disclosure, a multi-time scale optimization scheduling method for a hydrogen-electricity coupled microgrid is provided, comprising the following steps:

[0044] Step S1: obtaining the topological structure of the hydrogen-electricity coupled microgrid to be optimized;

[0045] Step S2: According to the topological structure, a source-storage-load multi-time scale coordinated optimization scheduling model is established, which consists of seasonal hydrogen storage optimization with gradually shortened time scales, day-ahead economic scheduling, and intraday dynamic optimization;

[0046] Furthermore, the seasonal hydrogen storage optimization is a long-term optimization with an annual cycle and a monthly scale. By introducing seasonal grid price factors, the grid operation pressure in different seasons is characterized and evaluated. The goal is to minimize the annual operating cost. Based on the source and load data throughout the year, the seasonal hydrogen storage and release plan is optimized, including the hydrogen storage input and output power for each month.

[0047] Furthermore, the day-ahead economic dispatch is a medium-time scale optimization, with a day as the cycle and an hour as the scale. It optimizes the input and output plans of electrolyzers, fuel cells, hydrogen energy storage and electric energy storage while meeting load demand, and minimizes energy production costs while ensuring daily income.

[0048] Furthermore, the intraday dynamic optimization is a short-time scale optimization, with an hourly cycle and a 5-minute scale. When photovoltaic power generation and load demand fluctuate, it optimizes energy procurement and equipment scheduling plans, adjusts the real-time output of electrolyzers, fuel cells, hydrogen energy storage and electric energy storage, so that renewable energy, hydrogen energy and electricity demand are more matched, thereby achieving real-time energy balance within the microgrid.

[0049] Step S3: Based on the summary and analysis of historical optimization results, expert experience and best practices are introduced to establish an optimization strategy knowledge base and form constraint conditions for the optimization scheduling model;

[0050] Furthermore, the optimization strategy knowledge base is formed by integrating the optimization result data set and the expert knowledge data set;

[0051] The optimization result data set is a large amount of load demand and renewable energy power generation data generated through long-term monitoring and simulation. On this basis, optimization simulation is performed under different energy supply and demand scenarios, the optimization results of each scenario are recorded, and the scene features are analyzed and selected from the optimization results to form the optimization result data set;

[0052] The expert knowledge data set converts scheduling rules and historical experience into a computable parameter set, quantifies expert knowledge, and integrates the quantified expert knowledge into the optimization model to formulate system scheduling constraints to form an expert knowledge data set.

[0053] Step S4: using a variable neighborhood search solution algorithm based on a dynamic window method to solve the optimization scheduling model, and obtaining an optimal energy procurement and equipment scheduling plan for achieving optimal operation of the hydrogen-electricity coupled microgrid under fluctuations in energy supply and demand.

[0054] Furthermore, the variable neighborhood search solution algorithm is based on the day-ahead scheduling plan curve, determines the maximum variable range of the corresponding decision variables, forms an optimization window at the current moment, and selects the energy procurement and equipment scheduling plan with the highest evaluation score within the optimization window.

[0055] Furthermore, the determination of the maximum variable range of the corresponding decision variables to form the optimization window at the current moment is a dynamic window method, which adjusts and limits the parameter search space according to the real-time system status and dynamic response capability, and then dynamically reduces the feasible domain to generate a new feasible domain as the optimization window at the current moment.

[0056] The following is a detailed description of the implementation process of a multi-time scale optimization scheduling method for a hydrogen-electricity coupled microgrid in this embodiment.

[0057] In view of the asynchronous temporal and spatial distribution of renewable energy output and load demand, and taking into account factors such as energy supply, energy storage system, energy conversion and load demand, this embodiment proposes a multi-time scale optimization scheduling method for a hydrogen-electricity coupled microgrid with hydrogen-electricity combined energy storage:

[0058] Firstly, an optimal scheduling framework for source-storage-load multi-time scale coordination is proposed, such as Figure 1As shown, through seasonal hydrogen storage optimization, day-ahead economic scheduling and intraday dynamic optimization, a flexible energy scheduling strategy is adopted to balance the fluctuations of various energy inputs and load demand, optimize the balance between energy supply and load demand, and achieve a balance between economy and environmental friendliness.

[0059] Secondly, a data knowledge-driven hybrid intelligent optimization strategy is proposed. By integrating historical data and expert knowledge, it provides global optimal strategy guidance for optimization scheduling, avoids the solution process from falling into local optimality, and improves the speed and accuracy of optimization.

[0060] Finally, a variable neighborhood search solution algorithm based on the dynamic window method is developed to continuously adjust the optimization window and feasible domain search range to adapt to real-time dynamic changes, thereby achieving optimal operation of the system under input fluctuations.

[0061] Among them, the optimization strategies include optimizing energy procurement, energy load scheduling, energy interaction between power grids, etc. On the premise of meeting load demand, select appropriate energy procurement and equipment scheduling plans to minimize energy costs and carbon emissions, consider and coordinate the mutual influence with the main power grid, participate in the operation and management of the power grid through optimizing scheduling strategies, and improve the economic and environmental benefits of the system; therefore, the optimization scheduling of hydrogen-electricity coupled microgrids involves energy procurement, energy load scheduling, energy interaction between power grids, etc., and requires the coordinated management of multiple energy forms and energy conversion processes, including renewable energy power generation, water electrolysis to produce hydrogen, hydrogen energy storage, fuel cell power generation, etc.

[0062] The working state and performance of fuel cells and electrolyzers are greatly affected by external conditions (such as temperature, humidity, load changes, etc.), such as fluctuations in energy input, load changes, changes in environmental conditions, etc., and the two have different response times; in optimal scheduling, the system operation plan needs to be dynamically adjusted to adapt to different working conditions and provide stable energy output.

[0063] Seasonal hydrogen energy storage can achieve cross-seasonal energy complementarity and supply and demand balance, while electric energy storage has a fast response speed and can improve the volatility of renewable energy; hydrogen-electricity combined energy storage is complementary, but it also increases the complexity and dimension of the hydrogen-electricity coupled microgrid optimization scheduling problem. The essence of the energy storage scheduling problem is a time series optimization problem. Energy storage scheduling is highly interactive with hydrogen production and utilization. It is necessary to consider changes in energy supply and demand in the time dimension and schedule energy storage through the optimization of decision variables.

[0064] In order to solve the optimization scheduling problem of hydrogen-electric coupled microgrid, the originally complex optimization scheduling problem is decomposed into sub-problems of various time scales, which are described by different time scales. The objectives, variables and constraints of each model are determined by the physical characteristics of the equipment on a single time scale; in order to achieve the coordinated optimization of various energy conversion and storage devices at different time scales, a data and knowledge-driven multi-time scale optimization framework is established, such as Figure 1 As shown, the knowledge base (i.e., the optimization strategy knowledge base) provides strategy guidance for optimization, and the optimization results provide data updates for the knowledge base, so as to obtain more accurate optimization results.

[0065] 1. Optimization Strategy Knowledge Base

[0066] In the stage of establishing the knowledge base, a large amount of load demand and renewable energy generation data, namely source-load data, is generated through long-term monitoring and simulation; on this basis, optimization simulation is carried out under different energy supply and demand scenarios, and the optimization results of each scenario are recorded; cluster analysis is performed on the respective optimization results under different scenarios, and the clustering results are analyzed and interpreted according to the optimization objectives; the performance of each cluster is evaluated to find the cluster with the lowest operating cost, and the scenario characteristics in the cluster are analyzed, and the key characteristics are selected to form an optimization result data set, including a range of hourly input and output plans for electrolyzers, fuel cells, hydrogen energy storage and electric energy storage.

[0067] The scenarios here are spring, summer, autumn and winter. Specifically, the source and load data of a year are divided into 365 days, and the data of each day is given to the optimization program in turn. After the optimization results are run out, the optimization results are analyzed in a clustering way; the daily optimization results are the input and output plans of the electrolyzer, fuel cell, hydrogen energy storage and electric energy storage, as well as the hourly operation plan of the power grid's electricity purchase. The daily optimization results will correspond to the operating costs of that day; clustering the optimization results of each scenario is to cluster the daily optimization results according to different seasons, and find the hourly operation plan corresponding to the day with the best daily operating cost.

[0068] By converting scheduling rules and historical experience into computable parameter sets, expert knowledge is quantified and integrated into the optimization scheduling model. System scheduling constraints are formulated to form an expert knowledge data set, including a range of hourly input and output plans for electrolyzers, fuel cells, hydrogen energy storage, and electric energy storage, to ensure the rationality and stability of the scheduling plan. The effectiveness of the optimization strategy based on expert knowledge is verified through simulation and analysis, and the performance of the strategy in different scenarios is analyzed to ensure that it can adapt to various situations.

[0069] The expert knowledge data set and the optimization result data set are integrated to form an optimization strategy knowledge base to ensure the comprehensiveness and effectiveness of the knowledge base. The information in the knowledge base is used to guide the decision-making of the optimization scheduling strategy.

[0070] The guidance here is specifically:

[0071] The information in the knowledge base is specifically a range of hourly input and output plans for electrolyzers, fuel cells, hydrogen energy storage, and electric energy storage. This information is used to constrain parameters. For example, from historical optimization results, it is found that it is cheaper to charge energy storage batteries with low-priced electricity at night. When optimizing scheduling, the minimum power constraint for nighttime charging is set to a certain value, and the model is optimized according to this value. The electrolyzer needs to operate at low power when electricity prices are high, so the maximum power constraint of the electrolyzer during this time period is limited to a relatively small value. However, in order to avoid frequent starts and stops, the minimum power constraint is also set to a certain value. These specific constraint values ​​come from the knowledge base; other equipment is similar.

[0072] 2. Optimizing the Scheduling Model

[0073] The optimization scheduling model in the optimization scheduling framework performs energy optimization on long, medium and short time scales and complements each other. Information and data on different time scales can be transmitted and used collaboratively during the optimization process. The optimization results on the long time scale can provide a benchmark and reference for the optimization on the short time scale, while the optimization results on the short time scale can improve and supplement the optimization strategy on the long time scale.

[0074] Long-term optimization takes one year as the cycle and one month as the scale. By introducing seasonal grid price factors, the grid operation pressure in different seasons is characterized and evaluated. With the goal of minimizing the annual operation cost, the monthly hydrogen storage input and output power are optimized to solve the seasonal supply and demand mismatch problem and ensure the long-term stable operation of the system. The formula for annual operation cost is:

[0075]

[0076] in, and are the seasonal grid price factor and the basic grid price, The real-time amount of electricity purchased by the system from the power grid.

[0077] The medium-time scale optimization is based on days and hours. The goal of day-ahead economic dispatch is to optimize the input and output plans of electrolyzers, fuel cells, hydrogen energy storage, and electric energy storage while meeting load demand, and to minimize energy production costs while ensuring daily income. The formula for energy production costs is:

[0078]

[0079] in, and They are the real-time grid price and hydrogen price, is the real-time hydrogen load, and are the energy equipment operating cost and the start-stop penalty, For the collection of energy devices, and are the corresponding coefficients respectively.

[0080] The short-term optimization is based on an hourly cycle and a 5-minute scale. The purpose is to optimize the real-time output of electrolyzers, fuel cells, hydrogen energy storage, and electric energy storage under the condition of fluctuations in photovoltaic power generation and load demand, thereby reducing the deviation of energy supply and demand, making renewable energy, hydrogen energy and electricity demand more compatible, and thus achieving real-time energy balance within the microgrid. The formula for energy supply and demand deviation is:

[0081]

[0082] in, , and are the supply and demand deviations of hydrogen energy, electricity energy and thermal energy, , and They are the outputs of different energy devices. , and They are a collection of hydrogen energy, electrical energy and thermal energy equipment. , and They are real-time hydrogen load, electricity load and heat load. , and are the corresponding coefficients respectively.

[0083] By comprehensively considering the changes and demands on long, medium and short time scales, the overall optimization of the hydrogen-electricity coupled microgrid can be achieved, which can better cope with energy challenges on different time scales and provide flexible, efficient and reliable energy supply.

[0084] 3. Solution Algorithm

[0085] The search process for the optimal value needs to traverse the entire feasible domain space, which consumes a lot of time and computing resources on invalid parameter combinations, resulting in slow convergence of the optimization algorithm and easy to fall into the local optimum. This embodiment introduces the knowledge base to construct constraints, which can narrow the feasible domain of such complex optimization scheduling problems and limit the search scope to parameter combinations that meet various constraints or are nearby. This can avoid using the global parameter search space, speed up the solution process, reduce the possibility of falling into the local optimum, and thus find the optimal solution more accurately, improving the efficiency and feasibility of the optimization process. Figure 2 This is a flow chart of the multi-time scale optimization solution algorithm proposed in this embodiment.

[0086] Seasonal hydrogen storage optimization solution algorithm, such as Figure 2 As shown in the upper left part, the specific solution steps are as follows:

[0087] Step 1. Input the annual source load data;

[0088] Step 2. Establish a nonlinear programming model for seasonal hydrogen storage;

[0089] Step 3. Approximate the nonlinear programming model of seasonal hydrogen storage into a quadratic programming subproblem through Taylor expansion;

[0090] Step 4. Use the sequential quadratic programming algorithm to solve the quadratic programming sub-problem and solve the seasonal hydrogen storage problem;

[0091] Step 5. Obtain and output the seasonal hydrogen storage / release plan, i.e., the hydrogen storage input and output power for each month.

[0092] The optimization algorithm for day-ahead economic dispatch is as follows: Figure 2 As shown in the upper right part, the specific solution steps are as follows:

[0093] Step 1. Initialize parameters according to the seasonal hydrogen storage / release plan and optimization strategy knowledge base. The parameters initialized here are the parameters to be solved, that is, the hourly input and output plans of the electrolyzer, fuel cell, hydrogen energy storage and electric energy storage;

[0094] Step 2. Use the particle swarm algorithm to optimize and solve, and update the particle velocity and position at the current moment according to the state constraints. Here, the state constraints are the adjustable ranges of the electrolyzer, fuel cell, hydrogen energy storage, and electric energy storage at the current moment.

[0095] Step 3. Calculate the particle fitness value, and then obtain the optimization objective function value;

[0096] Step 4. Determine whether the optimization target meets the end condition. If so, proceed to step 5. If not, proceed to step 3.

[0097] Step 5. Obtain and output the day-ahead economic dispatch plan, i.e., the hourly input and output plans of the electrolyzer, fuel cell, hydrogen energy storage, and electric energy storage, as well as the power grid purchase plan.

[0098] The variable neighborhood search solution algorithm for intraday dynamic optimization is based on the dynamic window method. Before solving each optimization cycle, it takes the day-ahead scheduling plan curve as the basis, evaluates the system change capability according to the system real-time state feedback estimation and dynamic response model, and then determines the maximum variable range of the corresponding decision variables to form the optimization window at the current moment; as time goes by, the maximum variable range of the decision variables is continuously adjusted to realize the dynamic optimization window, so that the solution space can be searched under different states, thereby improving the optimization effect more comprehensively, such as Figure 2 As shown in the following part, the specific solution steps are as follows:

[0099] Step 1. Input the day-ahead optimization scheduling plan;

[0100] Step 2. Sampling the system status at the current moment;

[0101] Step 3. Calculate the dynamic window at the current moment. First, limit the variable range according to the day-ahead plan, then predict the load demand at the next moment, and then determine the maximum limit range of system changes. Finally, form the optimization window at the current moment.

[0102] Step 4. Simulate the output plan, generate multiple two-dimensional power arrays within the optional power range, and then simulate the change curves of these power arrays within a certain time interval;

[0103] Step 5. Based on the real-time state feedback estimation and the system dynamic response model, the evaluation weight coefficient is first adaptively adjusted to the system changes. Then, the score of each output plan is comprehensively evaluated. Then, the comprehensive evaluation score is standardized. Finally, the output plan with the highest evaluation score is selected as the best adjustment plan.

[0104] Step 6. The system executes the optimal adjustment plan at the current moment, i.e., the input and output plan of the electrolyzer, fuel cell, hydrogen energy storage and electric energy storage and the power purchase plan of the power grid every 5 minutes.

[0105] Step 7: Determine whether the end condition is met. If the end condition is met, the optimization ends. If not, proceed to step 2.

[0106] The dynamic window method can quickly respond to dynamic environmental changes and continuously adjust the maximum variable range of decision variables, that is, the set of all possible states that the system may reach at the next moment, to dynamically reduce the feasible domain. The reduction of the feasible domain can more accurately search for the optimal solution and improve the efficiency and feasibility of the optimization process. Reducing the feasible domain means predicting and evaluating the adjustable range of a single device before each solution based on real-time state feedback estimation and the system dynamic response model, and dynamically reducing the search range of parameters to ensure that the parameter value is within a reasonable range and avoids exceeding the system's response capability; this allows the solution algorithm to adjust and limit the parameter search space based on the real-time system state and dynamic response capability, thereby dynamically reducing the feasible domain. The process is as follows: Figure 3 As shown, the specific steps are as follows:

[0107] Step 1. Collect real-time status feedback data: Build a real-time status feedback mechanism for the system to obtain the real-time status feedback data of the system, such as input, output, response time and other information. This information can be used to accurately estimate the current state of the system and optimize based on this state.

[0108] Step 2. Establish a system dynamic response model: Based on the collected real-time status feedback data, establish a data-driven system dynamic response model to describe the system's response capability and response speed under different environmental parameter values, and evaluate its performance in terms of the objective function.

[0109] Step 3. Analyze the dynamic characteristics of the system: Based on the real-time state feedback and dynamic response model, analyze the dynamic characteristics of the system, including indicators such as the system's response time, stability, and sensitivity.

[0110] Step 4. Set a reasonable range of parameter values: Set a reasonable range of parameter values ​​based on the dynamic characteristics of the system and the goal of reducing the feasible domain.

[0111] Step 5. Introduce constraints: According to the parameter range determined in step 4, introduce corresponding constraints to limit the decision variables and change the range of the parameter search space.

[0112] Step 6. Generate a new feasible domain: Through appropriate division and adjustment, determine the reasonable value range of the decision variables at the next moment, exclude the areas that cannot meet the constraints, and generate a new feasible domain, that is, the set of all states that the system may reach at the next moment, as the optimization window at the current moment.

[0113] As time goes by, the real-time status of the system is constantly updated, steps 3, 4, 5, and 6 are repeatedly executed, and the optimization window is constantly updated to form a dynamic window. During the solution process, variable neighborhood search is realized to improve the solution speed and accuracy.

[0114] Reducing the feasible domain can limit the search range to the parameter combination that just meets various constraints to avoid using the global parameter search space; before each solution, analyze the current system state and determine the maximum variable range of the corresponding decision variable based on the dynamic response capability of the device at the current moment; by setting the upper and lower bounds of the parameters, limit the parameters to a feasible range to avoid parameter values ​​that are too large or too small; by adding or removing constraints to reduce the feasible domain, the optimization process is more stable, the explosion or disappearance of parameter values ​​is prevented, and the oscillation during the training process is reduced; the parameter value range for each solution of the decision variable is expressed as follows:

[0115]

[0116] in, It is the system i The device is t The dynamic output at the moment is calculated by the established dynamic response model based on the state information; and are the maximum decrement and the maximum increment, respectively, which are determined by the dynamic response capability within the optimization cycle. The expressions are as follows:

[0117]

[0118]

[0119] in, To optimize the cycle.

[0120] A dynamic response model is a mathematical model used to describe the response of a system or object to external stimuli or changes. It is usually used to analyze and predict the behavior of a system, especially its evolution and change over time. In a dynamic response model, the response of a system or object is based on time, that is, the state and performance of the system change over time. The differential equation model describes the evolution of the system state in discrete time steps, and usually uses a recursive relationship to represent the relationship between the current state of the system and the previous state. The dynamic response model based on the differential form is expressed as follows:

[0121]

[0122] in, , , and They are dynamic output, dynamic input, white noise and offset; and are the coefficients of the autoregressive term, reflecting the impact of historical inputs and outputs on current outputs; , and They represent the polynomial order and delay parameters respectively.

[0123] Embodiment 2

[0124] In one embodiment of the present disclosure, a multi-time scale optimization scheduling system for a hydrogen-electricity coupled microgrid is provided, including an acquisition module, a modeling module, a constraint module, and a solution module:

[0125] The acquisition module is configured to: acquire the topological structure of the hydrogen-electricity coupled microgrid to be optimized;

[0126] The modeling module is configured to: establish a source-storage-load multi-time scale coordinated optimization scheduling model based on the topological structure, which consists of seasonal hydrogen storage optimization with gradually shortened time scales, day-ahead economic scheduling, and intraday dynamic optimization;

[0127] The constraint module is configured to: introduce expert experience and best practices based on summarizing and analyzing historical optimization results, establish an optimization strategy knowledge base, and form constraint conditions for the optimization scheduling model;

[0128] The solution module is configured to: use a variable neighborhood search solution algorithm based on a dynamic window method to solve the optimization scheduling model to obtain an optimal energy procurement and equipment scheduling plan for achieving optimal operation of the hydrogen-electricity coupled microgrid under fluctuations in energy supply and demand.

[0129] Embodiment 3

[0130] The purpose of this embodiment is to provide a computer-readable storage medium.

[0131] A computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the steps in a multi-time-scale optimization scheduling method for a hydrogen-electricity coupled microgrid as described in the first embodiment of the present disclosure.

[0132] Embodiment 4

[0133] The purpose of this embodiment is to provide an electronic device.

[0134] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, the steps in the multi-time-scale optimization scheduling method for a hydrogen-electricity coupled microgrid as described in the first embodiment of the present disclosure are implemented.

[0135] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A multi-time scale optimization scheduling method for a hydrogen-electricity coupled microgrid, characterized in that: include: Obtaining the topological structure of the hydrogen-electricity coupled microgrid to be optimized; According to the topological structure, an optimization scheduling model of source-storage-load multi-time scale coordination is established, which consists of seasonal hydrogen storage optimization with gradually shortened time scales, day-ahead economic scheduling and intraday dynamic optimization. Based on the summary and analysis of historical optimization results, expert experience and best practices are introduced to establish an optimization strategy knowledge base and form the constraints of the optimization scheduling model; The optimization scheduling model is solved by using a variable neighborhood search solution algorithm based on a dynamic window method to obtain an optimal energy procurement and equipment scheduling plan for achieving optimal operation of the hydrogen-electricity coupled microgrid under fluctuations in energy supply and demand; The seasonal hydrogen storage optimization is a long-term optimization with a year as the cycle and a month as the scale. By introducing seasonal grid price factors, the grid operation pressure in different seasons is characterized and evaluated. With the goal of minimizing the annual operation cost, the seasonal hydrogen storage and release plan is optimized based on the source and load data throughout the year, including the hydrogen storage input and output power of each month. The day-ahead economic dispatch is a medium-time scale optimization, with a day as the cycle and an hour as the scale. It optimizes the input and output plans of electrolyzers, fuel cells, hydrogen energy storage and electric energy storage while meeting load demand, and minimizes energy production costs while ensuring daily income. The intraday dynamic optimization is a short-time scale optimization, with an hourly cycle and a 5-minute scale. When photovoltaic power generation and load demand fluctuate, it optimizes energy procurement and equipment scheduling plans, adjusts the real-time output of electrolyzers, fuel cells, hydrogen energy storage and electric energy storage, so that renewable energy, hydrogen energy and electricity demand are more matched, thereby achieving real-time energy balance within the microgrid.

2. A multi-time scale optimization scheduling method for hydrogen-electricity coupled microgrid according to claim 1, characterized in that: The optimization strategy knowledge base is formed by integrating the optimization result data set and the expert knowledge data set; The optimization result data set is a large amount of load demand and renewable energy power generation data generated through long-term monitoring and simulation. On this basis, optimization simulation is performed under different energy supply and demand scenarios, the optimization results of each scenario are recorded, and the scene features are analyzed and selected from the optimization results to form the optimization result data set; The expert knowledge data set converts scheduling rules and historical experience into a computable parameter set, quantifies expert knowledge, and integrates the quantified expert knowledge into the optimization model to formulate system scheduling constraints to form an expert knowledge data set.

3. A multi-time scale optimization scheduling method for hydrogen-electricity coupled microgrid according to claim 1, characterized in that: The variable neighborhood search solution algorithm is based on the day-ahead scheduling plan curve, determines the maximum variable range of the corresponding decision variables, forms an optimization window at the current moment, and selects the energy procurement and equipment scheduling plan with the highest evaluation score within the optimization window.

4. A multi-time scale optimization scheduling method for hydrogen-electricity coupled microgrid according to claim 3, characterized in that: The method of determining the maximum variable range of the corresponding decision variables to form the optimization window at the current moment is a dynamic window method. According to the real-time system status and dynamic response capability, the parameter search space is adjusted and restricted, and the feasible domain is dynamically reduced to generate a new feasible domain as the optimization window at the current moment.

5. A multi-time scale optimization dispatching system for hydrogen-electricity coupled microgrid, characterized in that: Including acquisition module, modeling module, constraint module and solution module: The acquisition module is configured to: acquire the topological structure of the hydrogen-electricity coupled microgrid to be optimized; The modeling module is configured to: establish a source-storage-load multi-time scale coordinated optimization scheduling model based on the topological structure, which consists of seasonal hydrogen storage optimization with gradually shortened time scales, day-ahead economic scheduling, and intraday dynamic optimization; The constraint module is configured to: introduce expert experience and best practices based on summarizing and analyzing historical optimization results, establish an optimization strategy knowledge base, and form constraint conditions for the optimization scheduling model; The solution module is configured to: solve the optimization scheduling model using a variable neighborhood search solution algorithm based on a dynamic window method to obtain an optimal energy procurement and equipment scheduling plan for achieving optimal operation of the hydrogen-electricity coupled microgrid under fluctuations in energy supply and demand; The seasonal hydrogen storage optimization is a long-term optimization with a year as the cycle and a month as the scale. By introducing seasonal grid price factors, the grid operation pressure in different seasons is characterized and evaluated. With the goal of minimizing the annual operation cost, the seasonal hydrogen storage and release plan is optimized based on the source and load data throughout the year, including the hydrogen storage input and output power of each month. The day-ahead economic dispatch is a medium-time scale optimization, with a day as the cycle and an hour as the scale. It optimizes the input and output plans of electrolyzers, fuel cells, hydrogen energy storage and electric energy storage while meeting load demand, and minimizes energy production costs while ensuring daily income. The intraday dynamic optimization is a short-time scale optimization, with an hourly cycle and a 5-minute scale. When photovoltaic power generation and load demand fluctuate, it optimizes energy procurement and equipment scheduling plans, adjusts the real-time output of electrolyzers, fuel cells, hydrogen energy storage and electric energy storage, so that renewable energy, hydrogen energy and electricity demand are more matched, thereby achieving real-time energy balance within the microgrid.

6. An electronic device, comprising: a memory for non-transitory storage of computer readable instructions; a processor for executing the computer-readable instructions; Wherein, when the computer-readable instructions are executed by the processor, the method according to any one of claims 1 to 4 is executed.

7. A storage medium, characterized in that: The computer-readable instructions are non-transitorily stored, wherein when the computer-readable instructions are executed by a computer, the method according to any one of claims 1 to 4 is performed.