Island reef micro-energy network distribution robust optimization scheduling method and system based on flexible operation domain

Abstract

The application discloses a kind of island micro-energy network distribution robust optimization scheduling method and system based on flexible operation domain, comprising the following steps: S1, for the time sequence coupling association characteristics of island micro-energy network power supply system and water supply system, establish the island micro-energy network electricity-water collaborative flexible operation domain representation model with hydrogen energy preparation system, seawater desalination system and water reservoir as key adjustable resources;S2 analyzes island micro-energy network electricity-water collaborative flexible operation domain representation model, obtains the flexible operation domain boundary of island micro-energy network electricity-water collaboration;S3 introduces boundary constraint for the flexible operation domain boundary of island micro-energy network electricity-water collaboration, constructs and solves island micro-energy network distribution robust economic dispatching model.The present application can greatly improve the approximation accuracy of flexible operation domain while ensuring the calculation efficiency, effectively support the economicity and robustness of island micro-energy network scheduling scheme under the condition of extremely high proportion of renewable energy output fluctuation and uncertainty.

Description

Technical Field

[0001] This invention belongs to the field of microgrid optimization scheduling, and in particular relates to a method and system for optimizing the scheduling of island and reef microgrids based on flexible operating domains. Background Technology

[0002] Remote islands and reefs are important territories of my country, and it is necessary to build independent microgrids (IMEGs) to ensure energy and freshwater supply, replacing the traditional fossil fuel energy supply mode with high energy consumption and high emissions. At present, the industry has formed a multi-energy coupled island and reef microgrid construction concept based on renewable energy sources such as marine wind energy, solar energy, and wave energy. Existing research has solved the basic scheduling problems of water and heat supply and multi-energy complementarity of islands and reefs through deep reinforcement learning, multi-energy coupling modeling and other methods.

[0003] However, existing technologies have significant shortcomings. Island and reef microgrids exhibit strong random fluctuations in both source and load, and energy prediction errors cannot be eliminated, making existing deterministic scheduling schemes prone to suboptimal or even failed performance. Currently, the academic community mainly optimizes system operation through uncertainty-based scheduling and flexible supply-demand balance theory. Analysis methods based on the Flexible Operating Domain (FOR) can effectively characterize the system's uncertainty adaptation boundary and have become the mainstream research direction. While existing FOR construction methods can quantify system flexibility and ensure scheduling reliability, they are still limited in their effectiveness.

[0004] However, this type of method has limited adaptability. For island and reef microgrids with a high proportion of renewable energy and multiple types of flexible resource coupling, the system operation dimension is high and the number of operation points to be analyzed is huge, which will lead to a surge in computing power burden. Traditional FOR modeling and optimization methods cannot be directly applied and are difficult to accurately adapt to the complex operation characteristics of island and reef microgrids. There is an urgent need to optimize the scheduling scheme to solve the above technical bottlenecks. Summary of the Invention

[0005] To address the shortcomings and deficiencies of existing technologies, this invention proposes a flexible operating domain-based distributed bar optimization scheduling method for island and reef microgrids.

[0006] Specifically, the technical solution adopted in this invention is as follows: A method and system for optimizing the scheduling of island and reef microgrids based on flexible operating domains, comprising the following steps: S1: To address the temporal coupling and correlation characteristics of the power supply and water supply systems of island and reef microgrids, a characterization model of the flexible operation domain of the electricity-water coordinated system of island and reef microgrids is established, with hydrogen production systems, seawater desalination systems, and reservoirs as key adjustable resources.

[0007] The core of the island and reef microgrid system architecture proposed in 101) is to build an independent energy system that can meet the combined needs of electricity, heat, fresh water and hydrogen energy through the integration of diversified energy production, conversion and storage equipment. As well as multi-energy load demand, the island and reef microgrid system architecture mainly consists of the following parts: (1) Marine renewable energy power generation system, specifically including wind power generation system, photovoltaic power generation system, wave power generation system and tidal power unit, as the energy input of the island and reef microgrid system; (2) Energy conversion system, specifically including seawater desalination system, wastewater recycling system, electrolysis hydrogen production system and hydrogen fuel cell, responsible for the conversion between different forms of energy and the production of specific energy forms; (3) Multi-type energy storage units, specifically including electric energy storage, thermal storage, water storage tank and hydrogen storage tank, used to smooth supply and demand fluctuations and realize the transfer of energy in the time dimension; (4) Multi-energy load demand, covering the electricity load, heat load, fresh water load and hydrogen load required for the life and production of island residents. The island and reef microgrid system architecture diagram is attached. Figure 2 As shown.

[0008] 102) Construct the characteristic function of marine renewable energy power generation: output power of wind turbine Depends on real-time wind speed Its piecewise function is shown in (1); the output power of the photovoltaic panel With solar irradiance and ambient temperature Related, as shown in (2); the output power of the wave energy converter. With significant wave height The square of the tidal energy is proportional, as shown in (3); based on the dynamics of the flow velocity and the hydrodynamic coefficient, the tidal energy is proportional to the square of the tidal energy. Modeling is performed as shown in (4).

[0009] (1), (2), (3), (4), In the formula, and These represent the cut-in and cut-out wind speeds, respectively. This refers to the rated power of a single fan. Rated wind speed; The total area of ​​the photovoltaic panels. and These are all parameters for photovoltaic panels. This is a temperature correction factor; Indicates the overall efficiency of the converter. The density of seawater, For wave cycles, This represents the total wave contact length. Indicates the instantaneous velocity of the tide. This is a fluid dynamics coefficient.

[0010] Constructing the energy conversion coupling function: water production power consumption rate of the seawater desalination system Total power consumption and water production flow rate The constraints are shown in (5); the input electrical power for hydrogen production by electrolysis - Hydrogen production The relationship and minimum / maximum operating range are shown in (6); the hydrogen-electric conversion, electro-thermal synergistic output and operating constraints of the hydrogen fuel cell are shown in (7); the coupling constraints of water production flow and energy consumption of the wastewater recycling system and the dynamic balance constraints of the untreated wastewater storage tank are shown in (8).

[0011] (5), (6), (7), (8), In the formula, and For the parameters of the empirical function, This refers to the influent salt concentration. For water recovery rate, For reverse osmosis membrane resistance, Indicates the maximum water production flow rate of a single unit; For the efficiency of the electrolytic cell, The electro-hydrogen conversion coefficient is... This refers to the rated power of a single electrolytic cell; To output electrical power for hydrogen fuel cells, For hydrogen consumption, Because of the low calorific value of hydrogen, and These are power generation efficiency and waste heat recovery efficiency, respectively. , , and These represent the upper and lower limits of output power and the upper and lower limits of climbing power, respectively. For the production flow rate of reclaimed water, The electrical power consumed, This indicates the storage capacity of the untreated wastewater reservoir. Indicates the system's start / stop status. The efficiency coefficient for electricity-to-water conversion. and These are the minimum and maximum processing volumes, respectively. This represents the wastewater generation coefficient (the proportion of total water demand). This represents the total capacity of the untreated wastewater storage tank.

[0012] Constructing the physical energy storage response function: The response function of all energy storage systems is consistent. Taking battery energy storage as an example, the charging and discharging power constraints, energy storage state (State of Charge, SOC) constraints, charging and discharging rate constraints, SOC range and scheduling cycle start and end SOC constraints are shown in (9)-(13) respectively.

[0013] (9), (10) (11), (12), (13) In the formula, This refers to the net charge / discharge power. and These are the discharge and charge power, respectively. For battery energy storage capacity, This represents the initial energy storage level. and For charging and discharging efficiency, This refers to the charge / discharge rate.

[0014] Constructing a comprehensive demand response function and energy balance function: This invention considers that electricity, heat, hydrogen, and freshwater loads each have the ability to transfer energy over time within a scheduling cycle, that is, part of the load during peak energy consumption periods can be transferred to off-peak periods, which is called horizontal demand response capability; at the same time, multiple loads can be substituted for each other during the same period, and users can choose different energy supplies to meet their energy needs, which is called vertical demand response capability. The mathematical principles of these two types of demand response functions are similar, and their mathematical expressions are shown in equation (14).

[0015] (14) In the formula, Representing each moment The k Basic load; Indicates a fixed load; and They represent the first k Type of load in t Load after demand response and the amount of data participating in demand response; Indicates the first demand response k Type of load in t Values ​​for the time period; and These represent the incoming and outgoing power of the demand response load, respectively; and The variable is 0-1. If only one variable is 1 at any given time, it means that this type of load is transferred in or out during the demand response process. The total transfer amount of demand response load within a scheduling cycle is 0. and They represent the first k Type of load in t The upper and lower limits for participation in demand response during time period t.

[0016] For each moment Electricity after comprehensive demand response ,heat ,freshwater and hydrogen The load and the fixed load together constitute the corresponding base load, from which the energy supply and demand balance of the system can be obtained as shown in (15)-(18).

[0017] (15) (16) (17) (18) Constructing time-series operation constraint functions for devices and systems: The time-series operation constraints of all devices and the overall system can generally be abstracted into state transition type dynamic constraint functions. f Its general standard expression is as follows: (19) In the formula, Represents the system state vector. Represents the control decision vector. This is an external random disturbance.

[0018] 103) Based on the differentiated regulation capability model of island and reef microgrid energy conversion system, energy storage system and load demand flexible adjustment resources, it is accurately aggregated into a multi-energy coupled flexible operation domain through the Minkowski sum method. The specific representation of its mathematical model is shown in (20).

[0019] (20) In the formula, Let the vector of decision variables represent all flexible and adjustable resources. Then, the physical conservation and equilibrium constraints of energy flow can be represented in matrix form as follows: ,in P andW These represent net electrical power and net water flow rate, respectively. For a fixed energy transmission and conversion coefficient matrix, This represents the energy flow optimization vector. All operational constraints are uniformly represented as... ,in Represents linear constraints. This includes the corresponding upper and lower limits.

[0020] S2: Proposes an analytical method for the boundary geometric properties of a flexible operating domain in high-dimensional space based on the Hausdorff distance optimality criterion and the fast approximation capability of Zonotope.

[0021] (201) This invention provides a high-precision, computationally efficient analytical method for the boundary of the flexible operating domain of island and reef microgrids. To address the problem described in S1 that the flexible operating domain of island and reef microgrids has high dimensionality, complex geometry, and is difficult to directly apply to optimal scheduling, this invention proposes a Zonotope with good geometric and algebraic properties for analytical approximation of the flexible operating domain. This section will elaborate on the parameterized representation of the Zonotope, the optimal Zonotope approximation model based on Hausdorff distance, and the iterative solution algorithm for the optimal Zonotope approximation model. The algorithm flowchart is attached. Figure 2 As shown.

[0022] First, for approximate flexible operating domains A Zonotope is defined as a zonotope consisting of a central vector. and generator matrix A uniquely parameterized set of affine transformations. Its mathematical expression is (23): (twenty three) In the formula, for The center vector of defines the geometric center of the approximate domain; The generator matrix consists of a set of generator vectors. Composition, each generator vector A direction and length are defined geometrically, and the linear combination of these generators, namely Minkowski and, constitutes the shape of the Zonotope; For the first i The weight factors of each generator are limited to the range [-1, 1]. Generator matrix The shape, size, and orientation of the approximate domain are determined, which is key to characterizing the adaptability of the flexible operating domain.

[0023] 202) To ensure that Zonotope can accurately approximate the flexible operating domain, this invention proposes to construct an approximation model based on the optimal Hausdorff distance criterion. The Hausdorff distance measures the maximum global mismatch between two geometric sets, and compared to traditional metrics such as Euclidean distance, it is better able to capture the worst-case boundary deviations that are crucial to system safety. Specifically, the optimization objective of this part is to find the optimal center. and generator matrix To minimize Hausdorff distance between Zonotope and The optimization problem can be specifically described as (24): (twenty four) Among them, Hausdorff distance Support functions of two sets In all unit directions The maximum difference is quantified as shown in (25): (25) In the formula, The support function for the original flexible operating domain is essentially a linear programming problem. The support functions of the Zonotope have closed analytical expressions. By minimizing the global maximum difference of the support functions, it is possible to drive the boundaries of the Zonotope to fit uniformly and closely to the boundaries of the original flexible operating domain in all directions.

[0024] 203) Due to the objective function Relative to optimization variables Since the surface is neither convex nor smooth, it is difficult to obtain an analytical solution directly. To address this, an iterative optimization algorithm combining subgradient descent and a greedy heuristic is proposed for efficient solution. In each iteration, this algorithm adjusts the center vector collaboratively. and generator matrix This allows the boundaries of the Zonotope to dynamically and faithfully approximate the geometric boundaries of the original flexible operating domain. A single iteration of the algorithm (from step...) arrive It includes the following core mathematical steps: ① Subgradient calculation: The subgradient provides guidance for the descent direction of a non-smooth function. In iteration... In this process, calculating the subgradient of the Hausdorff distance is a prerequisite for all update steps. First, by solving an optimization problem (26), we find the unit direction with the largest difference between the support functions of the two domains under the current approximation. : (26) In actual calculations, the solution It is a For a linear programming problem with parameters, and It has an analytical form. This optimization problem can be solved using a nonlinear optimization method that performs a discretized search on the unit circle.

[0025] Using the obtained maximum deviation direction Hausdorff is about the center and generator The subgradient of can be analytically derived, as shown in (27): (27) These subgradient vectors indicate the direction in which the center and generator should be adjusted to minimize the current maximum boundary deviation as quickly as possible.

[0026] ② Iterative update of the center vector and generator matrix: After obtaining the subgradient, the algorithm updates the Zonotope parameters through the following cooperative steps. First, the center vector... Perform a standard subgradient descent update: ,in For the first The step size or learning rate of each iteration.

[0027] Then, to fine-tune the shape of the Zonotope, the generator matrix is ​​optimized to dynamically adjust its orientation and size. Specifically, each generator... Decoupling is its direction (unit vector) ) and size (scaling factor) ),Right now Keep your direction The scaling factor vector remains unchanged. Perform a gradient descent operation to further reduce the Hausdorff distance. For each scaling factor... The update rule is (28): (28) This step ensures the generator's contribution in high-dimensional space and The geometric anisotropy is matched.

[0028] To optimize the direction composition of the generator set, a greedy strategy is adopted to introduce the direction that is most effective in reducing the current maximum error.

[0029] The first step is to find In the direction of the current maximum deviation The poles (i.e., boundary points) on the surface. : (29) Then, construct a new center. Candidate generator pointing to the critical boundary point .

[0030] The second step is to find the current set of generators. After the generator is replaced, the generator that minimizes the Hausdorff distance is the one with the "minimum contribution". That is, solve (30): (30) In the formula, It is The Middle Replace each generator with The new Zonotope is then formed. Finally, an update is performed. .

[0031] ③ Achieve optimal approximation through iterative optimization: The above update steps are continuously performed in a main loop until the change in Hausdorff distance is less than the preset convergence threshold. or reaching the maximum number of iterations. The final output That is, the flexible operating domain The optimal Zonotope approximation.

[0032] S3: Considering the boundary constraints of the flexible operation domain of island and reef microgrids with electricity-water synergy, a distributed bar optimization scheduling method for island and reef microgrids is proposed, as shown in the appendix. Figure 3 As shown, this paper demonstrates an efficient solution for the optimal scheduling strategy that balances operational economy and extreme uncertainty.

[0033] 301) The Zonotope approximation method based on the Hausdorff distance optimality criterion is applied to a scenario that simultaneously considers multi-energy coordinated conversion constraints, unit ramping constraints, and cross-time dynamic coupling constraints with energy storage units and reservoirs. The cross-time dynamic coupling constraints are described by the following unified matrix form: (31) In the formula, for The system state vector for a given time period includes at least the unit output. With energy storage state of charge ;matrix and It is a coupling coefficient matrix, whose internal structure encodes all cross-time physical relationships, including dynamic changes in energy storage SOC and unit ramp rate. To constrain the boundary vectors. This is achieved through the flexible operating domain. By approximating, an approximate analytical flexible operation domain is ultimately constructed that can uniformly characterize the comprehensive flexible supply capability of the island and reef microgrid under multiple energy and time dimensions. .

[0034] 302) Construct and solve a distributed bar economic scheduling framework, which specifically includes: ① Construct a mixed-integer quadratic programming model with the objective of minimizing the overall operating cost over all time periods. The objective function is: (32) In the formula, For the unit exist Contributions during a given period , and Its cost coefficient, For its start-up costs, and These are 0-1 variables representing its running and startup states, respectively. For energy storage units exist Cost per kilowatt-hour of electricity degradation over a given period; ② The Zonotope is embedded into the model as a unified, system-level, flexible operating domain constraint, which is achieved by introducing a set of auxiliary variables. The specific form of implementation is as follows: (33) In the formula, the left side represents the decision made by the scheduling department. The high-dimensional vector that determines the net power exchange trajectory of multiple energy sources throughout the entire time period. and They are respectively The center and generator matrix, Ensured auxiliary variables The range of values ​​for ; ③ Combining the aforementioned data-driven uncertainty sets Construct a partial blue bar chance constraint to ensure that the energy-power balance is maintained with high probability: (34) In the formula, For the rescheduled variable vector, and These are the predicted and actual outputs of renewable energy, respectively. For uncertainty set Compatible probability distribution fuzzy sets, Risk tolerance; ④ Solve the model using mixed-integer linear programming to obtain the scheduling strategy that is the optimal solution to the optimization problem in the feasible space of all scheduling decision variables.

[0035] 303) Considering the boundedness and asymmetry of the uncertainty fluctuations of marine renewable energy, and the scarcity or unreliability of data, a data-driven polyhedron set reconstruction method is adopted to model the uncertainty, specifically including the following steps: 1) Record the predicted value as The historical prediction error sequence is denoted as ; 2) Using the mean of the error sequence Covariance Matrix Construct an ellipsoidal uncertainty set: (twenty one) In the formula, For statistical significance level and The determined threshold is used to guarantee the event. The probability is at least .

[0036] 3) Based on the ellipsoidal set, and according to the linear structure of the island and reef microgrid scheduling problem, reconstruct the polyhedral uncertainty set: (twenty two) In the formula, This represents scheduling decision variables that include historical data input. With uncertainty The correlation function, y For scheduling decision variables based on uncertain inputs, To address the cost function of uncertainty, This represents the upper limit of historical data statistics.

[0037] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention proposes and constructs a robust optimization scheduling method and system for island and reef microgrids based on a flexible operating domain. It aggregates dynamically coupled flexible resources into a unified and analyzable flexible operating domain and achieves high-fidelity geometric approximation and boundary analysis, providing a key theoretical basis for achieving robust optimization scheduling. (2) The present invention adopts a flexible operating domain approximation method that takes into account both geometric fidelity and computational feasibility. With the global worst-case deviation (Hausdorff distance) as the optimization target, the center and generator of the Zonotope are iteratively optimized. While ensuring high computational efficiency, it avoids the excessive conservatism of the traditional inner approximation method and the safety risks of the outer approximation method, and significantly improves the approximation accuracy of the flexible boundary. (3) This invention proposes a data-driven method for constructing uncertainty sets and robust scheduling applications. It does not require prior assumptions about the probability distribution of uncertainty. It directly learns the fluctuation boundary from historical data and solves it efficiently with the analytical flexible operating domain under a unified robust optimization framework, thereby enhancing the scheduling scheme's ability to cope with extreme uncertainty disturbances. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the microgrid system architecture for islands and reefs. Figure 2 The flowchart shows the iterative solution algorithm for the optimal approximation model of Zonotope. Figure 3 A robust optimization scheduling framework for island and reef microgrids; Figure 4 (a) The results of the flexible operation domain characterization of island and reef microgrids in time periods 4 and 8; Figure 4 (b) The results of the flexible operation domain characterization of island and reef microgrids in time periods 12 and 16; Figure 4 (c) The results of the flexible operation domain characterization of island and reef microgrids in time periods 20 and 24; Figure 5 The convergence process of the iterative solution algorithm for the optimal approximation model of Zonotope; Figure 6 A multi-time-period three-dimensional flexible operating domain with cross-time-period coupling constraints; Figure 7 This is a power balance diagram for island and reef microgrids based on robust optimization scheduling; Figure 8 This is a water flow balance diagram for an island and reef microgrid based on robust optimization scheduling. Detailed Implementation

[0039] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0040] In a preferred embodiment of the present invention, a typical island and reef microgrid system is constructed. This system includes a photovoltaic system with a maximum installed capacity of 0.8MW, a 2.5MW wind power system consisting of five 0.5MW units, and a wave energy conversion device with a designed rated power of 800kW as the main renewable energy supply. To achieve cross-time energy transfer and multi-energy synergy, the system is equipped with a 32.5MW / 130MWh electrical energy storage system, a 3.6MW alkaline electrolyzer cluster consisting of three 1.2MW units, and a 4814m³ capacity... 3The system includes a high-pressure gaseous hydrogen storage tank and a hydrogen fuel cell with a rated power of 1.2 MW. It also includes an 800 kW reverse osmosis seawater desalination system and a storage tank with a capacity of 13,400 m³. 3 The system includes a reservoir, serving as a key water-electricity coupling unit and adjustable load for the seawater desalination system. In addition, to meet the heat load requirements of the islands and reefs, the system is equipped with a thermal storage tank with a capacity of 0.8 MW / 1.6 MWh.

[0041] This invention proposes a flexible operating domain-based sub-Bruker optimal scheduling method for island and reef microgrids. This example compares and analyzes three different flexible operating domain approximation methods: the traditional internal approximation method (IA-based); the traditional external approximation method (OA-based); and the proposed Hausdorff optimal Zonotope-based approximation method (HOZ-based). In the IA-based method, the approximate flexible operating domain is constructed to completely encompass the island and reef microgrid's flexible operating domain, ensuring absolute operational safety but sacrificing some flexibility potential. In the OA-based method, the approximate flexible operating domain is constructed to completely enclose the island and reef microgrid's flexible operating domain, encompassing all flexibility potential but introducing infeasible operational risks. The proposed HOZ-based method aims to achieve a high-fidelity approximation of the island and reef microgrid's flexible operating domain by minimizing global geometric deviations.

[0042] The simulation in this example is based on the appendix. Figure 1 The aforementioned island and reef microgrid system architecture, attached Figure 4 The results of characterizing the flexible operating domain of island and reef microgrids using three different approximation methods at different time periods (time periods 4, 8, 12, 16, 20, and 24) are presented. (Appendix) Figure 5 The convergence process of the iterative solution algorithm for the HOZ-based method proposed in this invention is demonstrated. (Appendix) Figure 6 This demonstrates the multi-time-period three-dimensional flexible operating domain constructed by the method of this invention, which considers cross-time-period coupling constraints. (Appendix) Figure 7 and attached Figure 8 The system's electrical and water balance results are obtained after robust optimization scheduling based on the flexible operating domain constructed by the HOZ-based method of this invention.

[0043] According to the appendix Figure 4 Appendix Figure 5 and attached Figure 6 Analysis shows that the HOZ-based method proposed in this invention exhibits a significant advantage in approximation accuracy within the flexible operating domain. (See attached diagram.) Figure 4As shown, the IA-based method constructs an overly conservative region. For example, in time period 4, its upper limit for power regulation is only about 75MW, compared to the 95MW upper limit of the flexible operation domain of the island and reef microgrid, resulting in a loss of over 20MW of flexible regulation capability. This creates a significant gap with the boundary of the flexible operation domain, resulting in a substantial loss of feasible flexibility space. The OA-based method constructs an overly aggressive region, introducing numerous physically infeasible risk areas. Its boundary deviation at the vertices is particularly pronounced, overestimating the power regulation range by more than 15MW and the water flow regulation range by 10kgal / h in multiple time periods, posing significant operational risks to the system. In contrast, the HOZ-based method proposed in this invention achieves a high degree of fit with the boundary of the flexible operation domain of the island and reef microgrid in all time periods, effectively balancing operational safety and economy while ensuring extremely high geometric fidelity. (Appendix) Figure 5 The smooth loss curve verifies the effectiveness and stability of the iterative optimization algorithm proposed in this invention. The algorithm can converge quickly within a finite number of iterations, with over 90% of the loss function reduction completed within 1000 iterations, and finally converges to a stable low value after approximately 2500 iterations, demonstrating its feasibility in engineering applications. (Appendix) Figure 6 This intuitively demonstrates that the method of the present invention can characterize the cross-time coupling characteristics caused by factors such as energy storage and ramping, forming a flexible "pipeline" with dynamic evolution. Its maximum power adjustment range has been expanded from less than 60MW at night (time period 24) to more than 90MW during the midday photovoltaic peak (time period 12), providing a precise dynamic boundary for multi-time period optimization scheduling.

[0044] According to the appendix Figure 7 and attached Figure 8 Further analysis reveals that the precise and flexible operating domain constructed based on the HOZ-based method of this invention can effectively support robust and economically optimized scheduling of island and reef microgrids. (Appendix) Figure 7 Power balance diagram and appendix Figure 7 The water flow balance diagram illustrates the dynamic supply and demand balance achieved by the system during the 24-hour scheduling cycle. During peak renewable energy generation periods (10:00-16:00), with total renewable energy output reaching nearly 60MW, the system charges at up to 20MW via a battery energy storage system (BESS), simultaneously driving adjustable loads such as the electrolyzer and seawater desalination system, effectively achieving 100% local consumption of clean energy. During evening peak load periods (18:00-20:00), the BESS and hydrogen fuel cells work together, providing over 15MW of flexibility support to compensate for power shortages. (Appendix) Figure 8In this study, the operation of the seawater desalination system is highly correlated with electricity prices and renewable energy output. As a key water-electricity coupling unit and adjustable load, it produces over 20 kgal / h of water during midday (12-2 PM) while maintaining low-power operation at night, demonstrating the economic efficiency of the scheduling strategy. The numerical results of the overall scheduling show that the flexible operating domain model constructed in this invention can accurately guide the coordinated optimization of multiple energy flows, effectively addressing source-load uncertainties while maximizing the overall operational benefits of the system.

[0045] On the other hand, this application also provides a flexible operating domain-based island and reef microgrid sub-bar optimization scheduling system, including: a flexible operating domain modeling module, a flexible operating domain boundary resolution module, and a sub-bar optimization scheduling module; The flexible operation domain modeling module is used to construct a multi-energy collaborative flexible operation domain model of island and reef microgrids that considers energy conversion systems, energy storage systems and multi-energy load demands. By establishing marine renewable energy power generation characteristic functions, energy conversion coupling functions, physical energy storage response functions, comprehensive demand response functions and multi-energy balance constraints, it describes the adjustability of multiple energy systems such as electricity, fresh water and hydrogen energy in island and reef microgrids. The flexible operating domain boundary analysis module is used to perform geometric boundary analysis on the high-dimensional flexible operating domain of island and reef microgrids based on the Hausdorff distance optimality criterion and the Zonotope fast approximation method. By optimizing the Zonotope center vector and generator matrix, it can make it approximate the original flexible operating domain boundary with high fidelity, thereby realizing the analytical representation of complex multidimensional flexible operating capabilities. The Zonotope-based robust optimization scheduling module is used to construct a Zonotope-based robust optimization scheduling model for island and reef microgrids based on flexible operating domain boundary constraints. It embeds the Zonotope-form flexible operating domain constraints into the energy management model and combines them with the fuzzy set of uncertain distribution of marine renewable energy. The module uses a mixed integer optimization method to solve for a robust optimization scheduling scheme that meets the requirements of safe system operation and economic objectives.

[0046] This application also provides an electronic device, which may include at least one processor and at least one memory. The processor may include one or more processing cores. The processor connects to various parts of the server using various interfaces and lines, and performs various functions of the server and processes data by running or executing instructions, programs, code sets, or instruction sets stored in the memory, and by calling data stored in the memory.

[0047] This application also provides a computer-readable storage medium storing instructions. When executed by one or more processors, these instructions cause an electronic device to perform one or more of the methods described in the above embodiments.

[0048] Furthermore, the functional units in this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0049] The serial numbers in the above embodiments are for descriptive purposes only and do not represent the chronological order of assembly or use of the components. Those skilled in the art will understand that the embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0050] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0051] Unless otherwise specifically stated, the relative steps, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of the invention.

[0052] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Those skilled in the art will readily conceive of other embodiments of this disclosure upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described in this disclosure. The specification and embodiments are considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. A method for optimized scheduling of island and reef microgrids based on flexible operating domains, characterized in that, Includes the following steps: S1 addresses the temporal coupling characteristics of the power supply and water supply systems of island and reef microgrids, constructs an energy system for island and reef microgrids including a hydrogen production system, a seawater desalination system, and a water storage tank, and establishes a flexible operation domain characterization model for the electricity-water synergy of island and reef microgrids. The S2 analytical flexible operating domain representation model uses a high-dimensional space flexible operating domain boundary geometric property analytical method based on the Hausdorff distance optimal criterion and the Zonotope fast approximation capability to obtain the flexible operating domain boundary. S3 introduces boundary constraints to the flexible operating domain boundary, constructs and solves the island and reef microgrid distributed bar economic scheduling model.

2. The island and reef microgrid distributed bar optimization scheduling method based on flexible operating domain as described in claim 1, characterized in that, The energy system architecture of the island and reef microgrid described in S1 includes an energy conversion system consisting of a marine renewable energy power generation system, a hydrogen fuel cell, a hydrogen production system, a seawater desalination system, and a wastewater recycling system, an energy storage system consisting of batteries, a thermal energy storage system, hydrogen storage tanks, and water storage ponds, as well as multi-energy load requirements.

3. The island and reef microgrid distributed bar optimization scheduling method based on flexible operating domain as described in claim 2, characterized in that, The flexible operation domain characterization model described in S1 includes the energy conversion system, energy storage system, and multi-energy load demand of the island and reef microgrid energy system. It aggregates the flexible operation capabilities of adjustable resources using the Minkowski method and applies it to the flexible operation domain of the island and reef microgrid. The mathematical model is expressed as follows: ， In the formula, The vector representing all adjustable resource decision variables, and the physical conservation and equilibrium constraints of energy flow represented in matrix form are as follows: , P and W Indicates net electrical power and net water flow rate. For a fixed energy transmission and conversion coefficient matrix, The energy flow optimization vector is represented as follows: all operational constraints are uniformly represented as ,in Represents linear constraints. Represents the upper and lower limits of the linear constraint.

4. The island and reef microgrid distributed bar optimization scheduling method based on flexible operating domain as described in claim 1, characterized in that, The specific steps in S2 based on the Hausdorff distance optimality criterion are as follows: Define a vector with a center vector c and generator matrix The parameterized Zonotope is, ， In the formula, The center of the Zonotope is the geometric center of the flexible operating domain; Characterizes the adjustment direction of the flexible operating domain; The generator weights, with values ​​ranging from [-1, 1]; The flexible operating domain of the Zonotope and island microgrid is optimized using a subgradient-based iterative optimization algorithm. Hausdorff distance between To enable Zonotope to quickly approximate a flexible operating domain Hausdorff distance The optimal representation is, ， in, Characterization support function In all unit directions The biggest difference, ， In the formula, These are support functions for Zonotope.

5. The island and reef microgrid distributed bar optimization scheduling method based on flexible operating domain as described in claim 4, characterized in that, The analytical method for the boundary geometric characteristics of the high-dimensional flexible operating domain in S2 specifically employs a subgradient-based iterative optimization algorithm. In each iteration, this algorithm adjusts the center vector and generator matrix to dynamically and faithfully approximate the geometric boundary of the original flexible operating domain. The optimization process of the generator includes: S21 optimizes the scaling factor of the generator based on the gradient descent method, so that the contribution of the generator in the high-dimensional space is consistent with the geometric boundary of the flexible operating domain. S22 selects and updates the generator set based on a greedy algorithm. In each iteration, it selects the generator that contributes the most to the Hausdorff distance and has the lowest computational complexity, and then updates it. S23 obtains the optimal Zonotope for the flexible operating domain of the approximate island and reef microgrid by iteratively optimizing the generator set and scaling coefficient.

6. The island and reef microgrid distributed bar optimization scheduling method based on flexible operating domain as described in claim 4, characterized in that, The specific steps for constructing the island and reef microgrid distributed energy bar economic dispatch model as described in S3 include: S31 constructs an energy management model for island and reef microgrids that considers electricity-water synergy. The objective function for minimizing the overall operating cost throughout the entire time period is: ， In the formula, For the unit exist Contributions during a given period , and Its cost coefficient, For its start-up costs, and These are 0-1 variables representing its running and startup states, respectively. For energy storage units exist Cost per kilowatt-hour of electricity degradation over a given period; The boundary constraint expression form of the flexible operation domain of the S32 island and reef microgrid with electricity-water synergy is as follows: ， In the formula, the left side is determined by the scheduling variable. The determined high-dimensional vector of net power from multiple energy sources throughout the entire time period. and These are the center and generator matrices of the Zonotope, respectively. Ensured auxiliary variables The range of values ​​for; The S33 construction design and the electro-hydraulic synergistic island and reef microgrid's distributed bar chance constraint is expressed as follows: ， In the formula, For the scheduling variable vector, and These are the predicted and actual outputs of renewable energy, respectively. To address the uncertainty of polyhedral shapes Compatible probability distribution fuzzy sets, Risk tolerance; S34 solves the energy management model of island and reef microgrids that incorporates electricity-water coordination using a mixed-integer linear programming method, and obtains the scheduling strategy for all scheduling decision variables while satisfying the boundary constraints of the flexible operating domain.

7. The island and reef microgrid distributed bar optimization scheduling method based on flexible operating domain as described in claim 6, characterized in that, In step S33, the uncertainty set of the multi-faceted polyhedron shape under the constraint of the Bruker bar is divided. The construction of this system needs to take into account the uncertainties and fluctuations of marine renewable energy, specifically including the following steps: S331 denotes the predicted value of marine renewable energy as The prediction error sequence is denoted as ; S332 utilizes the mean of the prediction error sequence Covariance Matrix Construct an ellipsoidal uncertainty set: ， In the formula, For statistical significance level and The determined threshold is used to guarantee the event. The probability is at least ; Based on the ellipsoidal uncertainty set, S333 constructs a polyhedral uncertainty set according to the island and reef microgrid energy management model that considers electricity-water synergy: ， In the formula, This represents scheduling decision variables that include historical data input. With uncertainty The correlation function, y For scheduling decision variables based on uncertain inputs, To address the cost function of uncertainty, This represents the upper limit of historical data statistics.

8. A distributed bar optimization scheduling system for island and reef microgrids based on flexible operating domains, characterized in that, The method for performing the method as described in any one of claims 1 to 7 includes: an electric-water collaborative flexible operation domain modeling module, a high-dimensional operation domain boundary resolution module, and a distributed bar economic scheduling optimization module; The flexible operation domain modeling module is used to construct an energy system for the island and reef microgrid, including a hydrogen production system, a seawater desalination system, and a water storage tank, based on the temporal coupling and correlation characteristics of the power supply system and water supply system of the island and reef microgrid, and to establish a flexible operation domain characterization model for the electricity-water synergy of the island and reef microgrid. The high-dimensional operating domain boundary analysis module is used to connect to the electric-water collaborative flexible operating domain characterization model. It adopts a high-dimensional spatial geometric analysis method based on the Hausdorff distance optimal criterion and combined with the Zonotope fast approximation capability to solve the boundary characteristics of the high-dimensional flexible operating domain and accurately obtain the complete flexible operating domain boundary of the island and reef microgrid under the electric-water collaborative mode. The sub-bulb-bar economic scheduling optimization module is used to construct a sub-bulb-bar economic scheduling model adapted to the multi-energy coupling characteristics of islands and reefs by using the solved boundary of the electric-water coordinated flexible operation domain as the system constraint condition, and to solve the model to achieve stable, economical, and robust optimization scheduling of island and reef microgrids under uncertain source load fluctuations.

9. An electronic device, characterized in that, At least one processor and at least one memory, wherein the processor may include one or more processing cores capable of performing the method as described in any one of claims 1 to 7.

10. A computer-readable storage device, characterized in that, The storage device contains computer program instructions, which are invoked and loaded by the processor and executed to perform the method as described in any one of claims 1 to 7.

Images