Offshore wind farm layout optimization method and system

Abstract

The application relates to the cross field of computer software and new energy engineering technology, and provides a sea wind farm layout optimization method and system, which comprises the following steps: generating a candidate layout scheme, the candidate layout scheme comprising unit selection and site coordinates; based on the candidate layout scheme, performing power generation calculation, sea cable scheme determination based on topology optimization and rapid engineering quantity determination based on a preset mapping relationship in parallel; calculating economic indicators based on the results of the power generation calculation, the sea cable scheme determination and the rapid engineering quantity determination; performing feedback iteration based on the economic indicators until a convergence condition is met, and outputting a globally optimized scheme. The application integrates evaluation models with different precisions into a closed-loop iteration driven by economic indicators, greatly improves the global optimization efficiency while ensuring the accuracy of the evaluation of key cost factors, and realizes the output of a globally optimal scheme considering efficiency and accuracy.

Description

Technical Field

[0001] This invention relates to the intersection of computer software and new energy engineering technology, specifically to a method and system for optimizing the layout of offshore wind farms. Background Technology

[0002] Currently, in the early design phase of offshore wind farms, turbine selection, micro-site selection, electrical design, and structural design typically rely on multiple independent professional software programs for step-by-step, manual design. Due to the fragmentation of these tools and the need for manual data transfer and conversion, the design process is inefficient, making it difficult to automate and achieve global economic optimization across a broad range of options. Forcibly integrating high-precision calculation models from all disciplines into a single optimization loop would lead to an exponential increase in computational load, excessively long optimization times, and difficulty in convergence, making it difficult to balance global optimization efficiency with evaluation accuracy. Therefore, there is an urgent need for an offshore wind farm layout optimization scheme that can balance evaluation accuracy and optimization efficiency, achieving optimal global economics. Summary of the Invention

[0003] To address the problem of balancing global optimization efficiency and evaluation accuracy in existing technologies, this application proposes a method and system for optimizing the layout of offshore wind farms. This method integrates evaluation models of varying accuracies into a closed-loop iteration driven by economic indicators, thereby achieving a globally optimal solution output that balances evaluation accuracy and optimization efficiency.

[0004] To achieve the above objectives, this application adopts the following technical solution: A method for optimizing the layout of offshore wind farms includes: generating candidate layout schemes, wherein the candidate layout schemes include turbine selection and turbine location coordinates; based on the candidate layout schemes, performing parallel calculations of power generation, determination of submarine cable schemes based on topology optimization, and rapid determination of engineering quantities based on preset mapping relationships; calculating economic indicators based on the results of the power generation calculation, submarine cable scheme determination, and rapid determination of engineering quantities; and performing feedback iterations based on the economic indicators until convergence conditions are met, and outputting a global optimization scheme.

[0005] The above scheme achieves a global optimal solution output that balances efficiency and accuracy by coordinating precise optimization and rapid estimation in the same closed-loop iteration and using economic indicators as feedback drivers. This ensures the accuracy of key cost factor assessments while significantly improving global optimization efficiency.

[0006] Optionally, the rapid determination of engineering quantities based on preset mapping relationships includes: interpolating and estimating based on a library of typical design parameters for basic structures to determine the basic engineering quantities; the determination of submarine cable schemes based on topology optimization includes: performing submarine cable topology optimization with the goal of minimizing the overall cost throughout the entire life cycle to determine the submarine cable connection scheme.

[0007] The above scheme quickly determines the basic engineering quantity through interpolation estimation and accurately determines the submarine cable scheme through topology optimization, achieving a complementary advantage of rapid estimation and accurate optimization. While ensuring the accuracy of submarine cable cost assessment, it avoids the computational consumption caused by repeated detailed design of the basic structure.

[0008] Optionally, the feedback iteration based on the economic indicators includes: adopting a two-layer collaborative optimization framework, where the upper layer optimizes and manipulates the unit selection and unit location coordinates, and the lower layer synchronously adapts the submarine cable and foundation costs for each candidate layout scheme generated by the upper layer and feeds them back to the upper layer, with the economic indicators serving as the fitness function of the upper layer to drive the iteration.

[0009] The above scheme, through a two-layer collaborative optimization framework, clarifies the data flow and variable manipulation logic between the upper and lower layers, avoids search space explosion caused by coupling of variables of different dimensions, and improves the convergence speed and stability of the optimization algorithm.

[0010] Optionally, the economic indicators include the internal rate of return on total investment or the levelized cost of electricity (LCOE).

[0011] The above scheme uses the internal rate of return on total investment or levelized cost of electricity as the fitness function to directly link the technical solution with economic indicators, ensuring that the output global optimization scheme meets the economic optimal goal of the actual project.

[0012] Optionally, the rapid determination of engineering quantities based on preset mapping relationships adopts a multi-parameter fusion interpolation algorithm; the determination of submarine cable schemes based on topology optimization adopts a simulated annealing combined with a minimum spanning tree algorithm.

[0013] The above scheme provides a specific engineering algorithm implementation path by adopting a multi-parameter fusion interpolation algorithm and a simulated annealing combined with a minimum spanning tree algorithm, which further ensures the accuracy of rapid estimation and the global optimization capability of submarine cable topology optimization.

[0014] Furthermore, this application also provides an offshore wind farm layout optimization system, including a rapid turbine location generation module, a power generation calculation module, a submarine cable layout optimization module, a foundation weight rapid estimation module, and an engineering cost and economic analysis module. The rapid turbine location generation module is configured to generate candidate layout schemes and control the iteration process, wherein the candidate layout schemes include turbine selection and turbine location coordinates. The power generation calculation module, the submarine cable layout optimization module, and the foundation weight rapid estimation module are configured to perform power generation calculation, submarine cable scheme determination based on topology optimization, and rapid determination of engineering quantities based on preset mapping relationships in parallel. The engineering cost and economic analysis module is configured to determine economic indicators based on the calculation results and feed them back to the rapid turbine location generation module to drive iteration until a global optimization scheme is output.

[0015] The above solution achieves seamless data transfer and collaborative computing across multiple disciplines by forming a closed-loop iterative feedback architecture under the control of the scheduler, breaking down the traditional design barriers of separate disciplines and software.

[0016] Optionally, the basic weight rapid estimation module is configured to perform interpolation estimation based on a library of typical design parameters for basic structures to quickly determine the basic engineering quantity; the submarine cable layout optimization module is configured to perform submarine cable topology optimization to determine the submarine cable scheme with the goal of minimizing the overall cost throughout the entire life cycle.

[0017] The above solution, through specific module configuration, achieves deep integration of rapid estimation and precise optimization within the same system architecture, supporting the efficient operation of the closed-loop iterative mechanism.

[0018] Optionally, the system includes an optimization decision layer, which includes: an upper-level optimization submodule configured to manipulate the unit selection and unit location coordinates; and a lower-level adaptation submodule configured to synchronously call the submarine cable layout optimization module and the basic weight rapid estimation module to perform cost adaptation and feedback for each candidate layout scheme generated by the upper-level optimization submodule.

[0019] The above scheme, by optimizing the sub-module division and interaction relationship of the decision-making layer, solidifies the two-layer collaborative optimization mechanism at the system structure level, ensuring the controllability and efficiency of the optimization process in complex solution space.

[0020] Optionally, the engineering cost and economic analysis module is configured to calculate the internal rate of return on total investment or the levelized cost of electricity (LCOE) and feed it back to the rapid machine station generation module as an optimization objective function.

[0021] The above scheme constructs a complete closed loop from technical calculation to economic evaluation by limiting the output objects and feedback objectives of the engineering cost and economic analysis module, ensuring that the system always iterates towards the direction of optimal economic efficiency.

[0022] Optionally, the system includes a library of typical design parameters for basic structures, which is constructed based on a five-dimensional design space of foundation type, water depth, geological parameters, hydrological parameters, and wind turbine load.

[0023] The above scheme constructs a library of typical design parameters for basic structures through a five-dimensional design space, providing high-density standard data support for rapid interpolation estimation and ensuring that the error in quickly determining the quantity of basic engineering work is controlled within an acceptable range for the project.

[0024] Beneficial effects: This application overcomes the contradiction between global optimization efficiency and evaluation accuracy in existing technologies by integrating turbine location and model adjustment, power generation calculation, precise submarine cable optimization, and rapid foundation estimation into a closed-loop iteration with economic indicators as feedback. On the one hand, the rapid determination of engineering quantities based on preset mapping relationships avoids the huge computational consumption of repeated detailed design of the foundation structure, significantly shortening the time of a single iteration. On the other hand, the determination of the submarine cable scheme based on topology optimization ensures the accuracy of the evaluation of submarine cable investment, which accounts for a large proportion of the project cost. The two work together in the same closed loop, driving iteration with economic indicators, which greatly improves the global optimization efficiency while ensuring the accuracy of cost evaluation. It can output a truly globally economically optimal solution, shortening the design comparison cycle from several weeks to several hours, and improving the project's economics from the source. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of a method for optimizing the layout of an offshore wind farm according to an embodiment of this application; Figure 2 This is a schematic diagram of a system architecture for optimizing the layout of an offshore wind farm, according to an embodiment of this application. Figure 3 This is a schematic diagram of another offshore wind farm layout optimization method according to an embodiment of this application; Figure 4 This is a schematic diagram comparing the site layout and key economic indicators of an embodiment of this application. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0027] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. In the description of this application, it should be understood that the terms “first,” “second,” etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated.

[0028] Example 1: This embodiment provides a fundamental method for optimizing the layout of offshore wind farms. This method establishes a closed-loop iterative mechanism driven by economic indicators, integrating evaluation models of varying accuracies into the same optimization loop to achieve a globally optimal solution output that balances evaluation accuracy and optimization efficiency. Combined with... Figure 1As shown, the method specifically includes the following steps: Step S110: Generate candidate layout schemes, which include unit selection and unit location coordinates.

[0029] Specifically, turbine selection refers to choosing one or more specific models of wind turbines from a pre-defined wind turbine model library for combination. For example, selecting turbines with a single unit capacity of 10MW, 15MW, or 20MW based on the wind resource conditions of a certain sea area. Turbine location coordinates refer to the geographical spatial distribution of wind turbines within the wind farm site, usually expressed in latitude and longitude or local planar coordinates. The process of generating candidate layout schemes can be based on a pre-defined regularized layout algorithm. Under site boundary constraints and turbine safety distance constraints, it automatically generates multiple feasible combinations of turbine models and coordinates as initial inputs for subsequent iterative optimization. Specifically, users can set the site range by importing site boundary files in KML or Shapefile format, and set parameters such as the turbine orientation angle (usually perpendicular to the prevailing wind direction), safety distance (e.g., 5 times the rotor diameter), and boundary avoidance distance. Based on these inputs, the system generates multiple compliant, regularized initial turbine location layout schemes within seconds.

[0030] Step S120: Based on the candidate layout scheme, the power generation calculation, the submarine cable scheme determination based on topology optimization, and the rapid determination of engineering quantities based on preset mapping relationships are performed in parallel.

[0031] Specifically, after generating candidate layout schemes, the system synchronously triggers the calculation tasks of three branches, rather than executing them sequentially. This parallel execution significantly shortens the time required for a single scheme evaluation. Specifically, the power generation calculation is based on wind resource data and a wind turbine wake model to assess the theoretical power generation and wake loss across the entire site. The submarine cable scheme determination based on topology optimization is a precise optimization process. For a given turbine location, it searches for the optimal submarine cable connection topology, cable type selection, and route, aiming for optimal electrical connections or lowest cost. The rapid determination of engineering quantities based on pre-established mapping relationships is a rapid estimation process. It does not perform complex structural finite element analysis but instead quickly estimates the engineering quantities of the foundation structure, such as the weight of foundation steel, by querying pre-established mapping relationship data. The "pre-established mapping relationship" referred to in this application refers to the mathematical correspondence or discrete data table association between the foundation engineering quantities and design parameters. By coordinating precise submarine cable optimization with rapid foundation estimation, the computational bottleneck caused by using high-precision models across all disciplines is avoided, while ensuring the accuracy of key cost factor assessments.

[0032] Step S130: Based on the results of power generation calculation, submarine cable scheme determination, and rapid determination of engineering quantity, calculate the economic indicators.

[0033] Specifically, economic indicators are quantitative standards for evaluating the merits of candidate layout schemes. The system receives power generation data, submarine cable cost data, and foundation engineering quantity data, and combines them with parameters such as market material unit prices, construction cost rates, and financial discount rates to calculate the economic indicators of the current candidate layout schemes. Economic indicators can be financial indicators reflecting the project's profitability, such as the internal rate of return on total investment, net present value, or levelized cost of electricity. By directly linking technical solutions with economic indicators, it ensures that the optimization direction always points towards the actual economic optimal goal of the project.

[0034] Step S140: Perform feedback iteration based on economic indicators until the convergence condition is met, and output the global optimization scheme.

[0035] Specifically, the calculated economic indicators are used as feedback signals to drive the iterative process. The system evaluates the merits of current candidate schemes based on the economic indicators and uses optimization algorithms to generate a new generation of candidate layout schemes, then executes steps S120 and S130 again. This feedback iteration mechanism allows the schemes to continuously evolve towards better economic efficiency. The convergence condition can be that the number of iterations reaches a preset maximum number of generations, or that the improvement in the economic indicators of the optimal schemes across multiple consecutive generations is less than a preset accuracy threshold. When the convergence condition is met, the system outputs the scheme with the best economic indicators in the historical iterations as the global optimization scheme. Through the above closed-loop iteration, this embodiment achieves a significant improvement in global optimization efficiency while ensuring the accuracy of cost assessment, shortening the design comparison cycle from several weeks to several hours.

[0036] Example 2: Based on Example 1, this example elaborates on the specific implementation methods of rapid estimation and accurate optimization, as well as the interaction logic of the two-layer collaborative optimization framework.

[0037] Rapid determination of engineering quantities based on preset mapping relationships includes: interpolation estimation based on a library of typical design parameters for basic structures to determine the basic engineering quantities.

[0038] Specifically, the typical design parameter library for basic structures is a discrete data set established in advance after detailed design of various typical working conditions using finite element analysis software. When it is necessary to determine the foundation engineering quantity for a certain machine site, the system extracts the water depth, geological parameters, hydrological parameters, and load parameters of the selected machine type corresponding to that machine site, and performs retrieval and interpolation estimation in the parameter library. The interpolation estimation referred to in this application is not limited to a specific mathematical interpolation method; it can be linear interpolation, polynomial interpolation, inverse distance weighted interpolation, or Kriging interpolation. By mapping the target engineering quantity from the high-density typical design parameter library, the need for complete finite element modeling and verification of the basic structure in each iteration is avoided. This mechanism reduces the time required for a single foundation engineering quantity assessment from several hours to seconds, eliminates the computational bottleneck in closed-loop iteration from a microscopic perspective, and controls the estimation error within an acceptable range for engineering, achieving a nonlinear balance between assessment accuracy and computational efficiency.

[0039] The determination of submarine cable schemes based on topology optimization includes: optimizing the submarine cable topology to determine the submarine cable connection scheme with the goal of minimizing the overall cost over the entire life cycle.

[0040] Specifically, the submarine cable connection scheme includes not only the topological connection between wind turbines, but also the selection of submarine cable cross-section type, route, and determination of the location of offshore substations. In this embodiment, the objective function of topology optimization is set to minimize the comprehensive cost over the entire life cycle, rather than simply pursuing the lowest initial construction cost. The comprehensive cost over the entire life cycle includes the initial construction cost of submarine cable material procurement and construction, as well as the present value of the energy loss cost caused by resistive heating during the operation of the submarine cable. If only the initial construction cost is the objective, the optimization algorithm tends to select submarine cables with smaller cross-sections and lower costs, but this will lead to a sharp increase in long-term energy loss, which will damage the overall economic efficiency of the project. By performing precise topology optimization with the goal of minimizing the comprehensive cost over the entire life cycle, the accuracy of the assessment of submarine cable investment, which accounts for a large proportion of the project cost, is ensured, so that the basic data fed back to the economic indicators has real engineering guidance value.

[0041] Feedback iteration is based on economic indicators, including: adopting a two-layer collaborative optimization framework, the upper layer optimizes the selection of control unit and the coordinates of the aircraft position, and the lower layer synchronously adapts the cost of submarine cables and foundations for each candidate layout scheme generated by the upper layer and feeds it back to the upper layer. The economic indicators serve as the fitness function of the upper layer to drive the iteration.

[0042] Specifically, in the complex design of offshore wind farms, if all variables, such as turbine selection, turbine location coordinates, submarine cable topology, and foundation engineering quantities, are optimized at the same level, a combinatorial explosion problem will arise, resulting in an extremely high dimensional search space that makes it difficult for the algorithm to converge or even get stuck in local optima. This embodiment adopts a two-layer collaborative optimization framework to decouple the variables. The upper-layer optimization submodule only manipulates discrete turbine selection variables and continuous turbine location coordinate variables, using intelligent optimization algorithms (such as genetic algorithms or particle swarm optimization) to generate candidate layout schemes. After receiving the candidate layout schemes from the upper layer, the lower-layer adaptation submodule synchronously calls the submarine cable layout optimization module and the foundation weight rapid estimation module to quickly adapt the submarine cable cost and foundation cost under the current layout, and feeds back these cost data along with the power generation calculation results to the upper layer. The economic index serves as the fitness function of the upper layer to evaluate the merits of the current candidate layout scheme and guide the generation of the next generation scheme. This interactive logic of decoupling, synchronous adaptation, and feedback between upper and lower layers effectively reduces the search dimension of the upper-layer algorithm, prevents the search space explosion caused by the coupling of variables of different dimensions, and greatly improves the convergence speed and stability of the optimization algorithm.

[0043] Example 3: Based on Example 1, this example elaborates on the specific physical meaning of the economic indicators and the specific execution logic of the underlying algorithm.

[0044] Economic indicators include the internal rate of return on total investment or the levelized cost of electricity (LCOE).

[0045] Specifically, the internal rate of return (IRR) is the discount rate that makes the cumulative present value of net cash flows of a project equal to zero over the calculation period. It reflects the project's profitability and is unaffected by external financing costs. The levelized cost per kilowatt-hour (LCOE) is the ratio of the present value of total costs over the project's entire lifecycle to the present value of total electricity generation, i.e., the comprehensive cost per kilowatt-hour. In this embodiment, either the IRR or the LCOE can be used as the fitness function for higher-level optimization. For example, when using the IRR as the optimization objective, the algorithm tends to search for schemes with high electricity generation and low construction costs; when using the LCOE as the optimization objective, the algorithm not only focuses on initial investment but also comprehensively considers operation and maintenance costs and total lifecycle electricity generation. It should be understood that economic indicators can also include other financial indicators such as net present value, as long as they reflect the project's economic viability. By directly linking the technical solution with economic indicators, it ensures that the optimization direction always points to the actual economically optimal goal of the project.

[0046] The engineering quantities are quickly determined based on the preset mapping relationship, and a multi-parameter fusion interpolation algorithm is used.

[0047] Specifically, the multi-parameter fusion interpolation algorithm is a fast estimation method that combines local retrieval and spatial weighting. During execution, the system first extracts feature vectors such as the foundation type, water depth, geological parameters, hydrological parameters, and wind turbine load of the target site. It then performs a K-nearest neighbor search in a database of typical design parameters for foundation structures to find the K most similar design conditions to the target site. Subsequently, an inverse distance weighting method is used to calculate the weighted average of the foundation quantities corresponding to these K conditions, which serves as the estimated foundation quantity for the target site. The core idea of ​​inverse distance weighting is that samples with greater distances contribute more to the estimation result, aligning with the continuity of the mechanical response of engineering structures. By narrowing the computational domain through K-nearest neighbor retrieval and ensuring estimation accuracy through inverse distance weighting, this algorithm reduces the time required for a single foundation quantity assessment from several hours in traditional finite element analysis to milliseconds, while keeping the estimation error within an acceptable range for engineering projects. This eliminates the computational bottleneck in closed-loop iteration from a microscopic perspective.

[0048] The submarine cable scheme was determined based on topology optimization, using simulated annealing combined with the minimum spanning tree algorithm.

[0049] Specifically, submarine cable topology optimization is a typical combinatorial optimization problem, with the search space growing exponentially with the number of wind turbines. In this embodiment, the minimum spanning tree algorithm is first used to generate a basic topology network connecting all wind turbines and offshore substations, ensuring the shortest total length of the submarine cable, thus obtaining a relatively optimal initial feasible solution. Subsequently, the simulated annealing algorithm is introduced to globally optimize the basic topology network. During the iteration process, the simulated annealing algorithm not only accepts solutions that improve the objective function (overall lifecycle cost), but also accepts solutions that worsen the objective function with a certain probability, thus effectively avoiding getting trapped in local optima. During the optimization process, the algorithm continuously adjusts the wind turbine grouping, submarine cable cross-section type, route direction, and offshore substation location, and recalculates the overall lifecycle cost. By quickly constructing an initial solution using the minimum spanning tree algorithm and then using the simulated annealing algorithm for global perturbation and probabilistic escape, this combined algorithm significantly improves the efficiency of topology optimization while ensuring the accuracy of submarine cable scheme evaluation, allowing precise submarine cable optimization to be seamlessly integrated into the closed-loop iterative process.

[0050] This embodiment further ensures the accuracy of rapid estimation and the global optimization capability of submarine cable topology through the specific implementation of the above-mentioned underlying algorithm, achieving a technical effect that balances optimization efficiency and evaluation accuracy.

[0051] Example 4: This embodiment provides a system for optimizing the layout of offshore wind farms. Combined with... Figure 2 As shown, the system includes a rapid site generation module, a power generation calculation module, a submarine cable layout optimization module, a foundation weight rapid estimation module, and an engineering cost and economic analysis module.

[0052] The rapid generator position generation module is configured to generate candidate layout schemes and control the iteration process. The candidate layout schemes include generator selection and generator position coordinates.

[0053] Specifically, the rapid turbine location generation module, acting as the system's control center, is responsible for coordinating the timing and logic of the entire optimization process. At the hardware or software level, this module can be a scheduling engine with a processor and memory, internally pre-programmed with rule-based layout algorithms and intelligent optimization algorithms. The rapid turbine location generation module not only generates initial turbine selections and location coordinate combinations but also generates a new generation of candidate solutions based on feedback signals in each iteration. For example, turbine selection can be a combination of turbines with single-unit capacities of 10MW, 15MW, or 20MW chosen from a turbine library, and location coordinates can be latitude and longitude grid points. It should be understood that the candidate solutions generated by the rapid turbine location generation module are not limited to a single turbine type but can also be mixed layout schemes of multiple turbine types with different capacities to adapt to the needs of complex wind resource distribution.

[0054] The power generation calculation module, submarine cable layout optimization module, and foundation weight rapid estimation module are configured to perform power generation calculation, submarine cable scheme determination based on topology optimization, and rapid determination of engineering quantities based on preset mapping relationships in parallel.

[0055] Specifically, these three calculation modules are logically independent, but their execution sequence is synchronously controlled by the rapid site generation module. After the rapid site generation module outputs a set of candidate layout schemes, the power generation calculation module, the submarine cable layout optimization module, and the foundation weight rapid estimation module simultaneously receive the schemes and each initiates its calculation task. The power generation calculation module calls the wind resource database and wake model (e.g., the Jensen model or the Gaussian model) to output the theoretical power generation for the entire field; the submarine cable layout optimization module calls the electrical topology algorithm to output the submarine cable connection scheme and cost; and the foundation weight rapid estimation module calls the preset mapping relationship data to output the foundation engineering quantities. This parallel architecture breaks through the data blocking bottleneck of traditional serial calculations. For example, in the traditional serial mode, submarine cable design can only proceed after the power generation calculation is completed, while in this embodiment, the three processes proceed synchronously, significantly shortening the wall clock time for a single scheme evaluation. There is no need for direct data interaction between the modules; instead, information integration is achieved by uniformly aggregating the results to the downstream module.

[0056] The engineering cost and economic analysis module is configured to determine economic indicators based on the calculation results and feed them back to the rapid machine position generation module to drive iteration until a global optimization solution is output.

[0057] Specifically, the engineering cost and economic analysis module, as a key node in the closed-loop feedback, receives power generation data from the power generation calculation module, submarine cable cost data from the submarine cable layout optimization module, and foundation engineering quantity data from the foundation weight rapid estimation module. This module integrates an engineering cost quota library and a financial evaluation model, calculating the economic indicators of the current candidate scheme by converting technical parameters into economic parameters. Subsequently, the engineering cost and economic analysis module feeds this economic indicator as an evaluation signal back to the rapid station location generation module. The rapid station location generation module uses this signal to judge the merits of the current scheme and decides whether to accept it or proceed to the next iteration. Through this modular data flow design, the system forms a complete closed-loop feedback control architecture. This architecture not only enables seamless flow of multi-disciplinary data within a unified framework but also solidifies the closed-loop iteration mechanism through clear module boundaries, providing a visual comparison object for subsequent infringement determination. When the convergence condition is met, the rapid station location generation module outputs the historically optimal global optimization scheme, completing the entire layout optimization process.

[0058] In addition, combined Figure 2 As shown, the system also includes a data layer and a human-computer interaction layer. The data layer provides standardized data support for the aforementioned calculation modules, including a wind resource and geographic information database, an equipment parameter database, and a market price and financial model database. Specifically, the wind resource and geographic information database stores and manages wind measurement tower data, reanalysis data, seabed topography data, and site boundary coordinates; the equipment parameter database includes a wind turbine model library (containing parameter fields such as power curves, thrust coefficient curves, rotor diameter, rated power, hub height range, nacelle weight, and load spectrum), a submarine cable model library (containing parameter fields such as voltage level, cross-section, resistance per unit length, current carrying capacity, and cost per unit length), and a library of typical design parameters for foundation structures; the market price and financial model database stores unit prices of building materials, offshore construction rates, and financial evaluation benchmark parameters. The human-computer interaction layer provides a graphical user interface, supporting functions such as project management, data import and management, parameter setting, optimization calculation control, scheme visualization and comparison, and report generation and export, enabling users to configure project parameters, monitor the optimization process, and visualize geographic information and optimization results.

[0059] Example 5: Based on Example 4, this example further refines the specific configuration of the basic weight rapid estimation module and the submarine cable layout optimization module, and elaborates on the internal structure and interaction relationship of the optimization decision layer.

[0060] The basic weight rapid estimation module is configured to perform interpolation estimation based on a library of typical design parameters for basic structures to quickly determine the quantity of basic engineering work.

[0061] Specifically, the rapid foundation weight estimation module integrates a data retrieval and interpolation algorithm execution unit. Upon receiving the turbine location coordinates and unit selection information from the rapid turbine location generation module, this module automatically extracts the environmental feature vector corresponding to the target turbine location and matches it against a local or remotely connected database of typical foundation design parameters. The interpolation estimation referred to in this application is not limited to a specific mathematical interpolation method; it can be linear interpolation, polynomial interpolation, inverse distance weighted interpolation, or Kriging interpolation. By mapping the target engineering quantity from a high-density database of typical design parameters, the rapid foundation weight estimation module avoids performing complete finite element modeling and verification of the foundation structure in each iteration. This modular configuration reduces the time required for a single foundation engineering quantity assessment from several hours to seconds, eliminating the computational bottleneck in closed-loop iteration from a microscopic perspective, while controlling the estimation error within an acceptable range for engineering applications, achieving a nonlinear balance between assessment accuracy and computational efficiency.

[0062] The submarine cable layout optimization module is configured to optimize the submarine cable topology to determine the submarine cable scheme with the goal of minimizing the overall cost throughout the entire life cycle.

[0063] Specifically, the submarine cable layout optimization module encapsulates a combined optimization algorithm engine. This module not only generates the topological connections between wind turbines but also simultaneously optimizes the selection of submarine cable cross-section types, route planning, and the location of offshore substations. In this embodiment, the objective function of the submarine cable layout optimization module is set to minimize the overall lifecycle cost, rather than simply pursuing the lowest initial construction cost. The overall lifecycle cost includes the initial construction cost of submarine cable material procurement and construction, as well as the present value of the energy loss cost caused by resistive heating during the operation of the submarine cable. If only the initial construction cost is considered, the optimization algorithm tends to select submarine cables with smaller cross-sections and lower costs, but this will lead to a sharp increase in long-term energy loss, thereby harming the overall economic efficiency of the project. By performing precise topology optimization with the goal of minimizing the overall lifecycle cost, the submarine cable layout optimization module ensures the accuracy of the assessment of submarine cable investment, which accounts for a large proportion of the project cost, making the basic data fed back to the project cost and economic analysis module have real engineering guidance value.

[0064] The system includes an optimization decision layer, which comprises: an upper-level optimization submodule configured to control the selection of the control unit and the coordinates of the control position; a lower-level adaptation submodule configured to simultaneously call the submarine cable layout optimization module and the basic weight rapid estimation module to perform cost adaptation and feedback for each candidate layout scheme generated by the upper-level optimization submodule; and a global optimization engine configured to perform optimization by combining the upper-level and lower-level optimization submodules through a two-layer coordinated optimization algorithm.

[0065] Specifically, in the complex design of offshore wind farms, if all variables, such as turbine selection, turbine location coordinates, submarine cable topology, and foundation engineering quantities, are optimized at the same level, a combinatorial explosion problem will arise, resulting in an extremely high dimensionality of the search space, making it difficult for the algorithm to converge or even get stuck in local optima. This embodiment decouples variables by optimizing the sub-module division of the decision-making layer. The upper-level optimization sub-module only manipulates discrete turbine selection variables and continuous turbine location coordinate variables, using intelligent optimization algorithms (such as genetic algorithms or particle swarm optimization) to generate candidate layout schemes. After receiving the candidate layout schemes from the upper-level optimization sub-module, the lower-level adaptation sub-module synchronously calls the submarine cable layout optimization module and the foundation weight rapid estimation module to quickly adapt the submarine cable cost and foundation cost under the current layout, and feeds back these cost data along with the power generation calculation results to the upper-level optimization sub-module. The economic index serves as the fitness function of the upper-level optimization sub-module to evaluate the merits of the current candidate layout scheme and guide the generation of the next generation scheme. This decoupling, synchronous adaptation, and feedback interaction logic between the upper and lower layers effectively reduces the search dimension of the upper-layer algorithm, prevents search space explosion caused by coupling of variables of different dimensions, and significantly improves the convergence speed and stability of the optimization algorithm. It should be understood that although the diagram shows an independent division between the upper-layer optimization submodule and the lower-layer adaptation submodule, in other implementations, they can also be integrated into the same processor core through different threads or function call stacks, as long as the interaction logic of upper-layer variable manipulation and lower-layer synchronous adaptation and feedback is satisfied.

[0066] Example 6: Based on Example 4, this example elaborates on the specific feedback path of the engineering cost and economic analysis module and the construction process of the library of typical design parameters for basic structures.

[0067] The engineering cost and economic analysis module is configured to calculate the internal rate of return on total investment or the levelized cost of electricity (LCOE) and feed it back to the rapid machine location generation module as the objective function for optimization.

[0068] Specifically, the engineering cost and economic analysis module integrates an engineering cost quota library and a financial evaluation model. In each iteration, this module receives power generation data from the power generation calculation module, submarine cable cost data from the submarine cable layout optimization module, and foundation weight estimation data from the foundation engineering quantity module. First, the engineering cost and economic analysis module combines the foundation engineering quantity and submarine cable cost data with market material unit prices and construction rates to calculate the static total investment of the current candidate scheme. Then, it inputs the static total investment, estimated operation and maintenance costs, total life-cycle power generation, and financial discount rate into the financial evaluation model to calculate the internal rate of return (IRR) or levelized cost of electricity (LCOE). After calculation, the engineering cost and economic analysis module uses this economic indicator as the fitness value of the optimization objective function via an internal bus or API interface and feeds it back to the rapid turbine location generation module. The rapid turbine location generation module evaluates the merits of the current candidate layout scheme based on this fitness value and guides the generation of the next generation scheme. This path design, which transforms technical parameters into economic parameters and directly feeds them back to the scheduling center, constructs a complete closed loop from multi-disciplinary collaborative calculations to economic evaluation, ensuring that the system always iterates towards optimal economic performance. It should be understood that economic indicators can also include other financial indicators such as net present value or payback period, as long as they reflect the project's economic viability.

[0069] The system includes a library of typical design parameters for basic structures, which is constructed based on a five-dimensional design space including foundation type, water depth, geological parameters, hydrological parameters, and wind turbine load.

[0070] Specifically, the typical design parameter library for basic structures is the core data asset supporting the rapid interpolation estimation of foundation weight by the rapid foundation weight estimation module. In offshore wind power foundation design, the engineering quantity of the foundation structure is affected by a variety of environmental and load factors. This embodiment constructs a design space through five dimensions: foundation type, water depth, geological parameters, hydrological parameters, and wind turbine load, and performs systematic sampling. Among them, the foundation type can include monopile foundations, jacket foundations, high-pile cap foundations, etc.; the water depth parameter covers the typical water depth range of the target sea area, discretized in 2-meter increments, with monopile foundations covering water depths of 15 to 51 meters and jacket foundations covering water depths of 35 to 75 meters; geological parameters include layered information such as soil shear strength and density; hydrological parameters include environmental load conditions such as waves and ocean currents; and wind turbine loads include mechanical indicators such as thrust and bending moment at the hub height. When constructing the parameter library, the system utilizes finite element analysis software (such as SACS or ANSYS) to perform refined design and verification on thousands of typical working points in the aforementioned five-dimensional spatial grid. It extracts the main foundation dimensions and steel weight corresponding to each working point and stores them discretizedly in the form of a multi-dimensional array or hash table. When the foundation weight rapid estimation module receives the feature vector of the target location, it can quickly retrieve and interpolate within the five-dimensional design space. These five dimensions were chosen because they directly determine the mechanical response and material usage of the foundation structure. High-density five-dimensional spatial sampling ensures that the error in rapidly determining the foundation engineering quantities is controlled within an acceptable engineering range, such as ±5%. This eliminates the potential accuracy loss caused by rapid estimation from a data support perspective, preventing competitors from circumventing the technical solution of this application by simply replacing database dimensions or reducing sampling dimensions.

[0071] This embodiment, through the construction of the aforementioned economic evaluation feedback path and five-dimensional design space database, further ensures the reliability of the closed-loop iterative mechanism and the accuracy of rapid estimation, achieving a technical effect that balances optimization efficiency and evaluation accuracy.

[0072] Example 7: To verify the operational effectiveness and economic value of the technical solution in a real engineering environment, this embodiment takes an offshore wind power project in a certain sea area as an example, combined with... Figure 3 The application process and output results of the system are explained in detail.

[0073] The workflow of the offshore wind farm turbine selection and layout optimization software system. This process, with optimal economic return (IRR) throughout the entire lifecycle as its core objective, achieves an automated closed loop from data input to solution output through a two-layer collaborative optimization architecture. The following is a detailed step-by-step analysis combining the original text and technical principles shown in the diagram: Step 1, Project Creation and Data Import. This includes creating the project, importing the site selection base map, and wind measurement data.

[0074] This forms the foundation for the system's operation. By importing actual geographical site selection base maps and wind measurement data, the system constructs a realistic physical environment model of the wind farm, providing data support for subsequent micro-site selection and power generation calculations.

[0075] Step 2, Parameter and Constraint Settings. This includes selecting the wind turbine model, setting the wind turbine orientation, turbine spacing, submarine cable voltage level, and economic constraints.

[0076] This step defines the boundary conditions for optimization. The system receives not only physical parameters such as the wind turbine model, but also "economic constraints." This determines that subsequent optimization algorithms will seek the solution with the highest "Levelized Cost of Electricity (LCOE)" or "Internal Rate of Return (IRR)" while "meeting basic engineering requirements."

[0077] Step 3: Initiate global optimization. This can be triggered by the user clicking the 'Optimize' button. Step 3 performs two-layer collaborative optimization.

[0078] Specifically, step 3 may include the following sub-steps: 3.1 Generating candidate solutions at the upper level. This includes generating unit selection and unit location coordinates.

[0079] Upper-level optimization mainly addresses the questions of "which aircraft to select" and "where to place them." The system typically employs genetic algorithms or particle swarm optimization to automatically combine aircraft from a pre-defined aircraft library (such as a mixed 20MW / 26MW configuration) and intelligently arrange aircraft positions on a map, avoiding no-fly zones and shallow water areas.

[0080] 3.2 Parallel Layout and Evaluation For each candidate solution generated by the upper layer, the system triggers three key sub-processes in parallel: Sub-process A: Power generation calculation. This is achieved by calling the power generation calculation module.

[0081] Sub-process B: Rapid cost estimation for submarine cables and foundations. This is achieved by calling the submarine cable layout optimization module and the foundation weight rapid estimation module.

[0082] Subprocess C: Subprocess C performs economic analysis. It can integrate the results from A and B and calculate the IRR.

[0083] The results of the first two sub-processes are integrated, along with the cost of the wind turbine itself and the revenue from power generation, and then substituted into the financial model to calculate the core indicators IRR and LCOE.

[0084] 3.3 Fitness Evaluation and Decision Making. This involves using economic indicators across the entire lifecycle of the system to select the optimal solution.

[0085] 3.4 Optimization Algorithm Iteration. The underlying optimization algorithm can be invoked to generate a new generation of solutions. If the maximum number of iterations is not reached or convergence is not achieved, the system will mutate or cross-generate a new generation of aircraft positions and models based on the characteristics of the current optimal solution, and then re-enter the 3.1 cycle.

[0086] Termination and Result Output Stage: "No: Return to 3.1", "Yes: Exit the loop".

[0087] The algorithm stops when it has run the set number of generations (e.g., 1000 generations) or when the IRR metric no longer improves significantly.

[0088] Step 4: Output the globally optimal solution. This includes: generator set, generator location, submarine cable, foundation selection model, and economic indicators.

[0089] Output the final deployment diagram, including the specific mixed configuration (e.g., 12 units of 20MW + 3 units of 26MW), precise planar coordinates, submarine cable route map, basic specifications, and final return on investment.

[0090] Step 5: Results Visualization and Report Generation. This includes: topographic map display, a comparison table of multiple options, and an optimal recommendation report.

[0091] Transform dry data into engineering drawings and Word reports that engineers can read, and support one-click export of CAD and bill of materials (BOM) to complete design delivery.

[0092] like Figure 4 As shown, in this application scenario, the original project site area is approximately 22.5 km², which, after being cut according to constraints such as waterways and water depth, results in a usable area of ​​approximately 13 km². The system receives project boundary inputs including the polygon coordinates of the cut site boundary, reanalysis wind data, and geological and hydrological parameters. The prevailing wind direction is east-northeast (ENE). The user selects two turbine models with single-unit capacities of 20MW (300m rotor diameter) and 26MW (313m rotor diameter) as candidate models from the equipment parameter database, and sets a safety distance of 5 times the rotor diameter, a boundary avoidance distance of 500m, and a sea area constraint of less than or equal to 15 hectares / MW for each unit capacity.

[0093] After completing parameter configuration, the user initiates global optimization. The rapid turbine location generation module, based on a two-layer collaborative optimization framework, manipulates turbine selection and location coordinate variables. The lower-level adaptation submodule synchronously calls the submarine cable layout optimization module and the foundation weight rapid estimation module for cost adaptation. The engineering cost and economic analysis module uses the internal rate of return on total investment as the fitness function to drive iteration. The system converges in approximately 3.5 hours, outputting the global optimization scheme. The optimal scheme has a total installed capacity of 318MW, specifically arranged with 12 20MW wind turbines and 3 26MW wind turbines. The turbine locations are arranged in a regular grid pattern, with the row direction perpendicular to the ENE wind direction. The spacing between turbine locations is adaptively adjusted at the boundaries to meet all constraints.

[0094] Combination Figure 4 As shown, the optimized scheme output by this system is compared with the traditional manual scheme from a design institute. In terms of economics, the internal rate of return (IRR) of the optimized scheme is approximately 0.15% higher than that of the traditional design institute scheme, and the levelized cost of electricity (LCOE) is reduced by approximately RMB 0.007 / kWh. This improvement is attributed to the decoupling and global optimization of unit selection, turbine location coordinates, submarine cable topology, and foundation cost through a two-layer collaborative optimization framework, avoiding the local optimum trap caused by traditional serial design by different disciplines. Regarding foundation cost estimation, the rapid foundation weight estimation module 204 performs multi-parameter fusion interpolation based on a typical design parameter library of foundation structures built in a five-dimensional design space. The rapid foundation weight estimation results for each turbine location are verified by traditional finite element analysis, with errors controlled within ±4%. This indicates that the rapid determination of engineering quantities based on preset mapping relationships significantly reduces computation time while ensuring that the evaluation accuracy remains within an acceptable range for the project. In terms of submarine cable cost, the submarine cable layout optimization module optimizes the topology with the goal of minimizing the overall cost throughout the entire lifecycle, and the total cost of the output submarine cable connection scheme is lower than the budget of the traditional manual design scheme from a design institute. This is because simulated annealing combined with the minimum spanning tree algorithm can effectively escape local optima in a huge combinatorial space, achieving a comprehensive optimization of the initial construction cost and long-term power loss cost of submarine cables.

[0095] Through the above application scenarios, the system of this application has shortened the design comparison cycle from several weeks to several hours, and is superior to the traditional solution in terms of economic indicators, basic estimation accuracy and submarine cable cost control. This proves the practicality and significant economic value of the closed-loop iterative mechanism that integrates rapid estimation and precise optimization in balancing optimization efficiency and evaluation accuracy.

[0096] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for optimizing the layout of offshore wind farms, characterized in that, include: Generate candidate layout schemes, which include unit selection and unit location coordinates; Based on the candidate layout schemes, the power generation calculation, the submarine cable scheme determination based on topology optimization, and the rapid determination of engineering quantities based on preset mapping relationships are performed in parallel. Based on the results of the power generation calculation, submarine cable scheme determination, and rapid engineering quantity determination, economic indicators are calculated. The feedback iteration is performed based on the aforementioned economic indicators until the convergence condition is met, and then the global optimization scheme is output.

2. The method according to claim 1, characterized in that, The rapid determination of engineering quantities based on preset mapping relationships includes: interpolation estimation based on a library of typical design parameters for basic structures to determine the basic engineering quantities; The determination of the submarine cable scheme based on topology optimization includes: performing submarine cable topology optimization to determine the submarine cable connection scheme with the goal of minimizing the overall cost over the entire life cycle.

3. The method according to claim 2, characterized in that, The feedback iteration based on the economic indicators includes: A two-layer collaborative optimization framework is adopted. The upper layer optimizes and manipulates the selection of the generator set and the coordinates of the generator position. The lower layer synchronously adapts the cost of submarine cables and foundations for each candidate layout scheme generated by the upper layer and feeds it back to the upper layer. The economic indicators serve as the fitness function of the upper layer to drive the iteration.

4. The method according to claim 1, characterized in that, The economic indicators include the internal rate of return on total investment or the levelized cost of electricity (LCOE).

5. The method according to claim 4, characterized in that, The rapid determination of engineering quantities based on preset mapping relationships employs a multi-parameter fusion interpolation algorithm. The submarine cable scheme based on topology optimization was determined by using simulated annealing combined with the minimum spanning tree algorithm.

6. A system for optimizing the layout of offshore wind farms, characterized in that, It includes a rapid site generation module, a power generation calculation module, a submarine cable layout optimization module, a foundation weight rapid estimation module, and an engineering cost and economic analysis module; The rapid generator position generation module is configured to generate candidate layout schemes and control the iteration process. The candidate layout schemes include generator selection and generator position coordinates. The power generation calculation module, submarine cable layout optimization module, and foundation weight rapid estimation module are configured to perform power generation calculation, submarine cable scheme determination based on topology optimization, and rapid determination of engineering quantities based on preset mapping relationships in parallel. The engineering cost and economic analysis module is configured to determine economic indicators based on the calculation results and feed them back to the rapid machine position generation module to drive iteration until a global optimization solution is output.

7. The system according to claim 6, characterized in that, The basic weight rapid estimation module is configured to perform interpolation estimation based on a library of typical design parameters for basic structures in order to quickly determine the amount of basic engineering work. The submarine cable layout optimization module is configured to perform submarine cable topology optimization to determine the submarine cable scheme with the goal of minimizing the overall cost throughout the entire life cycle.

8. The system according to claim 7, characterized in that, The system includes an optimization decision layer, which includes: The upper-level optimization submodule is configured to manipulate the unit selection and unit location coordinates; The lower-level adaptation submodule is configured to synchronously call the submarine cable layout optimization module and the basic weight rapid estimation module to perform cost adaptation and feedback for each candidate layout scheme generated by the upper-level optimization submodule.

9. The system according to claim 6, characterized in that, The engineering cost and economic analysis module is configured to calculate the internal rate of return on total investment or the levelized cost of electricity (LCOE) and feed it back to the rapid machine station generation module as the objective function for optimization.

10. The system according to claim 9, characterized in that, The system includes a library of typical design parameters for basic structures, which is constructed based on a five-dimensional design space including foundation type, water depth, geological parameters, hydrological parameters, and wind turbine load.

Images