Offshore wind farm collector circuit optimization method and system in conjunction with booster station location
By combining two-dimensional continuous coordinate search of the booster station with the optimization of the collector topology, the problem of decoupling the location of the booster station and the collector topology in offshore wind farms was solved. This method achieved the globally optimal station location and zoning results, improving the economic efficiency and feasibility of the project.
Patent Information
- Application Number
- CN202611404245.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-10
- Publication Date
- 2026-11-17
AI Technical Summary
In existing methods for planning offshore wind farm power collection systems, the location of the booster station and the power collection topology are often decoupled sequentially, which makes it impossible to apply the optimization scheme to actual offshore wind farms. It is also impossible to find the globally optimal station coordinates and power collection zoning results, which affects the economic efficiency and feasibility of the project.
A joint optimization method combining two-dimensional continuous coordinate search of the booster station and topology routing optimization of the collector is adopted. Through multi-starting point adaptive step-size random walk search, agent evaluation screening and precise evaluation confirmation, combined with system cost weighted minimum spanning tree algorithm and geometric obstacle bypass path planning, a submarine cable polyline path and submarine cable specifications that meet the constraints are generated.
The system achieved globally optimal substation locations and power collection zoning results, improving the engineering economy and feasibility of the power collection system planning scheme, optimizing submarine cable routes and specifications, and enhancing the overall planning scheme's feasibility.
Smart Images

Figure CN123121798A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of offshore wind farm technology, and in particular to a method and system for jointly optimizing the circuit composition of an offshore wind farm and the location of a booster station. Background Technology
[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.
[0003] The offshore wind farm's power collection system is the core electrical system that collects the electricity generated by each wind turbine via submarine cables to an offshore substation. The planning and design of this system involves two highly coupled sub-problems: determining the optimal location of the substation, and planning the topology and path of the submarine cable loops. The substation location constrains the overall geometry of the power collection network and the cable length distribution of each loop, while the submarine cable topology determines whether a particular location is truly economical; the two are mutually dependent.
[0004] Most existing methods treat the location of the booster station as a fixed prior known quantity, or limit its candidate range to a number of manually defined discrete coordinate points, grid nodes, or one-dimensional line segments between two points. The site selection is completed by enumerating discrete candidates and then evaluating the optimal topology one by one. The location of the booster station and the collection topology are often decoupled sequentially. The optimal grouping is obtained by fixing the topology first, or the optimal station location is obtained by fixing the group first, which makes the optimization scheme unapplicable to actual offshore wind farms. Summary of the Invention
[0005] To address the aforementioned technical problems, this invention provides a method and system for joint optimization of offshore wind farm collector routing and booster station location. This method and system can deeply integrate two-dimensional continuous coordinate search of booster stations with collector topology routing optimization, thereby improving the overall engineering economy and feasibility of the collector system planning scheme.
[0006] To achieve the above objectives, the present invention adopts the following technical solution: The first aspect of the present invention provides a method for jointly optimizing the location of the offshore wind farm collection circuit and the booster station.
[0007] In one or more embodiments, a method for jointly optimizing the distribution network of an offshore wind farm with the location of a booster station is provided, including: Based on the boundary polygon of the site area, the set of obstacles, and the preset safety distance constraints of the booster station, a feasible domain determination function and geometric constraint environment for the booster station are constructed. Using the two-dimensional coordinates of the booster station as the decision variable, and based on the feasible region determination function and geometric constraints of the booster station, a multi-starting-point adaptive step-size random walk search is performed. With the cooperation of surrogate evaluation screening and actuarial evaluation confirmation, the candidate optimal booster station location is obtained. By taking the location coordinates of each wind turbine in the set of wind turbine location coordinates as poles, and sorting and encoding them based on the polar angle of the tangent vector, and with the cooperation of proxy evaluation screening and actuarial evaluation confirmation, the wind turbine collection partitioning results that simultaneously meet the loop size constraints and comprehensive cost are obtained. Within each wind turbine power collection zone, based on the system cost weighted minimum spanning tree algorithm and combined with preset electrical and cost parameters, a power collection loop topology with each candidate optimal booster station as the root node is generated. The geometric obstacle avoidance path planning algorithm is called on each collector loop edge segment in the collector loop topology to generate a submarine cable polyline path that meets the set constraints; The submarine cable specifications and models are automatically selected based on the downstream capacity of each segment of the submarine cable, and relevant cost information is summarized to obtain a joint optimization scheme. The joint optimization scheme includes the candidate optimal booster station location, the results of each wind turbine power collection zone and its corresponding power collection circuit topology, the submarine cable broken line path, the submarine cable specifications and models, and a detailed breakdown of the total construction cost.
[0008] As one implementation method, during the multi-starting point adaptive step-size random walk search, uniform random sampling is performed within the boundary polygon of the field area to select several initial candidate points that satisfy the feasible region determination function. Proxy evaluation is performed on all initial points, and they are sorted in ascending order according to proxy cost. Points with higher rankings are selected according to a set ratio to perform actuarial evaluation to obtain the actuarial cost of each initial point, which is used as the starting cost of each random walk.
[0009] As one implementation method, during the multi-starting point adaptive step-size random walk search, each round of neighborhood candidate points is first quickly screened by proxy cost, and only candidate points with proxy costs better than the current best are submitted for actuarial evaluation; the actuarial evaluation determines whether the corresponding candidate point is accepted or rejected.
[0010] As one implementation method, the process of wind turbine power collection zone division includes: By taking the position coordinates of each wind turbine as the pole and sorting and encoding the polar angle of the tangent vector, the corresponding number of tangent position vectors are obtained, which are the candidate individuals; A reference cut point vector is generated by uniform cutting and offset in the positive and negative directions respectively to generate several initial individuals with geometric diversity. Random integer perturbation is applied to the cut point of each initial individual and the population is filled to the preset population size. Based on the total power collection cost of each candidate individual, a two-stage screening process is adopted within a preset population size. The process is as follows: first, all candidate individuals are approximated by straight-line distance to replace the actual obstacle bypass path, and a proxy cost assessment is run. The candidates are ranked according to their proxy costs, and only the top-ranked candidates are subjected to a precise assessment.
[0011] As one implementation method, the process of generating the collector loop topology with each candidate optimal booster station as the root node is as follows: Within the power collection zone of each wind turbine, with the substation as the root node, the wiring within the power collection loop is performed using the weighted minimum spanning tree algorithm based on system cost, with the increment of the total system cost as the edge weight. Based on the required current flow of each collector circuit, the submarine cable is selected segment by segment. Based on the selected submarine cable segment by segment, for each collector circuit, the cumulative voltage drop is calculated link by link on the series link from the step-up station to each terminal node. The results are compared with the preset voltage drop threshold to determine whether the series link is qualified. Otherwise, the submarine cable specification with the smallest cross-sectional area on the corresponding series link is automatically upgraded, the cumulative voltage drop is recalculated, and the iteration continues until the cumulative voltage drop of all series links does not exceed the preset voltage drop threshold or the submarine cable specification reaches the upper limit of the available specification.
[0012] As one implementation method, the agency evaluation screening and actuarial evaluation confirmation adopt the same cost model and submarine cable selection rules.
[0013] As one implementation method, in the agent evaluation and screening process, the total agent cost of the entire field is the sum of the outgoing cable agent cost and the geometric default penalty; the outgoing cable agent cost is estimated based on the straight-line distance from the booster station to the landing point; in the agent evaluation and screening process, only the initial population is evaluated, and genetic operator iteration is not performed; During the actuarial assessment and confirmation process, the total actuarial cost of the entire site is the sum of the precise cost of the outgoing cables and the precise cost increments of the wiring for all wind turbine collection zones and all substations; the precise cost of the outgoing cables is estimated based on the obstacle-bypassing broken line distance from the substation to the landing point; and the actuarial assessment and confirmation process involves the complete multi-generation evolution.
[0014] A second aspect of the present invention provides a system for jointly optimizing the circuitry of an offshore wind farm with the location of a booster station.
[0015] In one or more embodiments, an offshore wind farm collector circuit is jointly optimized with a booster station location system, including: The module for constructing the feasible region and constraints of the booster station is used to construct the feasible region determination function and geometric constraint environment of the booster station based on the site boundary polygon, the set of obstacles, and the preset safety distance constraints of the booster station. The candidate optimal booster station location generation module is used to perform a multi-starting-point adaptive step-size random walk search based on the booster station's two-dimensional coordinates as decision variables, the booster station's feasible region determination function, and the geometric constraint environment. With the cooperation of proxy evaluation screening and actuarial evaluation confirmation, the candidate optimal booster station location is obtained. The wind turbine power collection zoning result generation module is used to take the position coordinates of each wind turbine in the wind turbine position coordinate set as the pole, sort and encode them based on the polar angle of the tangent vector, and obtain the wind turbine power collection zoning result that simultaneously meets the loop size constraint and the overall cost under the cooperation of proxy evaluation screening and actuarial evaluation confirmation. The collector loop topology generation module is used to generate a collector loop topology with each candidate optimal substation as the root node, based on the system cost weighted minimum spanning tree algorithm and in combination with preset electrical and cost parameters, within the collector zone results of each wind turbine. The submarine cable polyline path planning module is used to call the geometric obstacle bypass path planning algorithm on each collector loop segment in the collector loop topology to generate a submarine cable polyline path that meets the set constraints. The joint optimization scheme generation module is used to automatically select the submarine cable specifications and models based on the downstream capacity of each submarine cable segment and summarize relevant cost information to obtain the joint optimization scheme. The joint optimization scheme includes the candidate optimal booster station location, the power collection zone results of each wind turbine and its corresponding power collection loop topology, the submarine cable broken line path, the submarine cable specifications and models, and a detailed breakdown of the total construction cost.
[0016] A third aspect of the present invention provides a computer-readable storage medium.
[0017] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method for joint optimization of offshore wind farm collection circuits with booster station locations.
[0018] A fourth aspect of the present invention provides an electronic device.
[0019] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the above-described method for joint optimization of offshore wind farm collection circuits with booster station locations.
[0020] Compared with the prior art, the beneficial effects of the present invention are: This invention uses the two-dimensional coordinates of the booster station as the decision variable, combined with the feasible region and geometric constraints of the booster station, to perform a multi-starting-point adaptive step-size random walk search. With the cooperation of surrogate evaluation screening and actuarial evaluation confirmation, the candidate optimal booster station locations are obtained. This breaks through the limitations of existing technologies that fix or restrict the station location to discrete candidate points / one-dimensional line segments, and can find the truly globally optimal station location coordinates. Taking the coordinates of each wind turbine location in the wind turbine location coordinate set as the pole, and sorting and encoding based on the polar angle of the tangent vector, with the cooperation of surrogate evaluation screening and actuarial evaluation confirmation, the wind turbine collection zone results that simultaneously satisfy the loop size constraints and comprehensive cost are obtained. Based on the system cost weighted minimum spanning tree algorithm and combined with preset electrical and cost parameters, the collection loop topology with each candidate optimal booster station as the root node is generated, and then the submarine cable polyline path that meets the set constraints is generated and the submarine cable specifications and models are determined. This realizes the joint optimization of deep integration between the two-dimensional continuous coordinate search of the booster station and the collection topology routing optimization, thereby improving the engineering economy and feasibility of the collection system planning scheme as a whole. Attached Figure Description
[0021] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.
[0022] Figure 1 This is a flowchart of the method for jointly optimizing the location of the offshore wind farm collector circuit and the booster station according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the system structure of the offshore wind farm collector circuit and the booster station location jointly optimized according to an embodiment of the present invention; Figure 3 This is a schematic diagram of an electronic device according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the final optimized collector network according to an embodiment of the present invention. Detailed Implementation
[0023] The present invention will be further described below with reference to the accompanying drawings and embodiments.
[0024] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, 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 invention pertains.
[0025] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.
[0026] Figure 1 A schematic diagram of the joint optimization method for offshore wind farm circuitry and booster station location according to an embodiment of the present invention is provided. Figure 1 The offshore wind farm collector circuit of this embodiment is jointly optimized with the location of the booster station, and may include the following steps S101~S106.
[0027] The specific implementation process of steps S101 to S106 is as follows: Step S101: Based on the boundary polygon of the site area, the set of obstacles, and the preset safety distance constraints of the booster station, construct the feasible region determination function and geometric constraint environment of the booster station.
[0028] Establish a feasible region determination function If point polygon at the boundary of the site Internally, and the Euclidean distance from all wind turbines is not less than the safety distance of the booster station. ,but (Feasible), otherwise 0 (infeasible).
[0029] Step S102: Using the two-dimensional coordinates of the booster station as the decision variable, and based on the feasible region determination function and geometric constraints of the booster station, perform a multi-starting-point adaptive step-size random walk search. With the cooperation of proxy evaluation screening and actuarial evaluation confirmation, the candidate optimal booster station location is obtained.
[0030] During the multi-starting point adaptive step-size random walk search, uniform random sampling is performed within the boundary polygon of the field area to select several initial candidate points that satisfy the feasible region determination function. Proxy evaluation is performed on all initial points, and they are sorted in ascending order according to proxy cost. Points with the highest ranking are selected according to a set ratio to perform actuarial evaluation to obtain the actuarial cost of each initial point, which is used as the starting cost of each random walk.
[0031] During the multi-starting point adaptive step-size random walk search, each round of neighborhood candidate points is first quickly screened by proxy cost, and only candidate points with proxy costs better than the current best are submitted for actuarial evaluation; the actuarial evaluation determines whether the corresponding candidate point is accepted or rejected.
[0032] Specifically, using the two-dimensional coordinates of the substation location For the decision variables, a global search is performed within the feasible region of the field boundary polygon. The process is as follows: Multi-starting point initialization: Uniformly random sampling is performed within the bounded rectangle of the field boundary, and samples are selected to meet the requirements. of Initial candidate points The agent evaluation was performed on all initial points, sorted in ascending order of agent cost, and then proportionally... Select the top-ranked points for further actuarial evaluation to obtain the actuarial cost of each initial point, which will be used as the starting cost of each random walk.
[0033] Adaptive step-size random walk main loop: using the current best point Current step size and current actuarial costs Given the state, perform the following iterations: Step a. At the current point radius Uniformly randomized within the range Angle of direction ,have to Candidate points:
[0034] Filter to meet Candidate points; Step b. Run a proxy evaluation for each feasible candidate point, filter points whose proxy costs are better than the current value, and run an actuarial evaluation for the points that are not included in the spatial exclusion list; Step c. If a certain actuarial result Less than the current actuarial cost ,Right now If so, accept that point as the new current point and increase the step size. Times:
[0035] And reset the no-improvement counter ; It is a positive integer greater than or equal to 2; Step d. If all candidate points fail the actuarial verification in this round, reduce the step size. Times:
[0036] No Improved Counter ; Step e. When Or the total number of assessments exceeds the limit. When the current movement ends; This is the preset counter cutoff number.
[0037] Spatial exclusion of rejected actuated points: Maintain a spatial exclusion list, recording all station coordinates that have undergone actuation but whose results are not better than the current optimal. For each newly generated candidate point... If its distance from any point in the list is less than the repulsion radius If so, the actuarial calculation is skipped directly to prevent repeated evaluation of neighboring areas that have been proven to be suboptimal, thus reducing ineffective actuarial calls.
[0038] The agency evaluation and screening process and the actuarial evaluation and confirmation process both use the same cost model and submarine cable selection rules. The core difference lies in the method of obtaining the path length and the depth of evolutionary iteration.
[0039] During the agent evaluation and selection process, the total agent cost is the sum of the outgoing cable agent cost and the geometric default penalty; the outgoing cable agent cost is estimated based on the straight-line distance from the booster station to the landing point; during the agent evaluation and selection process, only the initial population is evaluated, and genetic operator iteration is not performed.
[0040] Specifically, during the agent evaluation and screening process, for a given booster station location... With partition scheme Agent evaluation for each partition Perform the following steps independently: Step a1. Geometric minimum spanning tree construction: using Euclidean distance For border rights, in Run Prim's minimum spanning tree algorithm on all internal wind turbine nodes to generate an undirected, rootless tree. Select the fan closest to the booster station. As the root node, Transform into Given a directed tree with the root node, add the connecting edges from the root node to the booster station. (length is) ).
[0041] Step b1. Calculate the cost of each edge in the directed tree: For each directed edge in the directed tree... ( As the parent node, (as child nodes), with Euclidean distance as the path geometric length:
[0042] Will Substitute into the installation length formula Statistics Total installed capacity of downstream wind turbines of the root subtree Automatically select submarine cable model according to preset rules Call the unified cost accounting function to calculate the cost increment for that side. (Including all cost items that have already been implemented). Among them, The laying length factor (considering actual laying process factors such as seabed undulations and binding allowances of submarine cables, it is usually taken as 1.03 to 1.08). The length of the vertical section of the submarine cable at the substation and wind turbine (in meters) is used for cost calculation.
[0043] Step c1. Summary of Partition Proxy Costs: The basic proxy cost for each partition is:
[0044] Step d1. Geometric Default Penalty: Taking all line segments of the geometric minimum spanning tree of each partition as the object, count the total number of the following four types of geometric default events. (1) A straight line segment extends beyond the boundary polygon of the field area; (2) A straight line segment intersects with any obstacle polygon in the configuration; (3) A straight line segment intersects with the path of the outgoing cable; (4) An intersection occurs between any two straight line segments. Let the penalty coefficient be... ,when At that time, a severe penalty will be imposed on the overall basic agency cost:
[0045] in ;when The agency cost is the basic cost and is not penalized; however, once a first default occurs ( The penalty multiplier jumps from 1 to [a higher value]. The penalty mechanism allows the evolutionary algorithm to identify and eliminate geometrically flawed partitioning schemes during the surrogate evaluation phase, without needing to call Theta. Path planning. The penalty coefficient here. A value set by the user, for example, it can be set to a value between 0 and 1.
[0046] Step e1. Total Cost of Outsourcing: Add the cost of outgoing cable outsourcing. (Estimated based on the straight-line distance from the substation to the landing point, and derived from the distance and cable specifications):
[0047] In proxy evaluation mode, the evolutionary algorithm is fixed to run only one generation (evaluating only the initial population) and does not perform genetic operator iterations, thereby reducing the computational load of each site evaluation to 1 / 10 to 1 / 100 of that of actuarial calculations. This is used for rapid screening of a large number of candidate sites during the random walk phase. The scope of cost items (material costs, construction costs, marine usage fees, cross-surcharges, switchgear fees) is independently configured and controlled, regardless of whether proxy mode is in effect.
[0048] During the actuarial assessment and confirmation process, the total actuarial cost of the entire site is the sum of the precise cost of the outgoing cables and the precise cost increments of the wiring for all wind turbine collection zones and all substations; the precise cost of the outgoing cables is estimated based on the obstacle-bypassing broken line distance from the substation to the landing point; and the actuarial assessment and confirmation process involves the complete multi-generation evolution.
[0049] During the actuarial assessment confirmation process, the actuarial assessment is based on a given zoning scheme. For each partition, as input Perform the following steps: Step a2. System cost-weighted Prim cabling: System cost increment Using edge weights, run the improved Prim algorithm to construct a rooted tree with the booster station as the root. This allows topological edge selection to directly correspond to engineering economic objectives (rather than geometric shortest).
[0050] Step b2. Theta Actual path planning: For Each edge Call Theta Obstacle avoidance algorithms search for obstacles under obstacle constraints. arrive A polyline path, the path is composed of a sequence of inflection points. The actual geometric length is described as follows:
[0051] Step c2. Calculate the cost edge by edge: Substituting the installation length formula and calculating the downstream capacity, we can perform submarine cable selection and complete cost accounting to obtain the accurate cost increment. Simultaneously, pressure drop verification and automatic upgrades are performed on each terminal node.
[0052] Step d2. Calculate the total cost of the entire operation:
[0053] Precise cost of outgoing cables The distance was estimated based on the obstacle bypass distance from the booster station to the landing point and the cable specifications.
[0054] The actuarial evaluation runs through a complete multi-generational evolution (including genetic operators, local search, and early stopping mechanisms) to accurately confirm a small number of high-quality candidate sites that have passed the agent screening, and to perform global fine-grained optimization of the final power collection scheme.
[0055] The key differences between the two models are compared in Table 1: Table 1. Comparison of key differences between the two models
[0056] Step S103: Using the location coordinates of each wind turbine in the set of wind turbine location coordinates as poles, sort and encode them based on the polar angle of the tangent vector. With the cooperation of proxy evaluation screening and actuarial evaluation confirmation, the wind turbine collection partitioning result that simultaneously satisfies the loop size constraint and the overall cost is obtained.
[0057] In step S103, the process of determining the wind turbine collector zoning results includes: By taking the position coordinates of each wind turbine as the pole and sorting and encoding the polar angle of the tangent vector, the corresponding number of tangent position vectors are obtained, which are the candidate individuals; A reference cut point vector is generated by uniform cutting and offset in the positive and negative directions respectively to generate several initial individuals with geometric diversity. Random integer perturbation is applied to the cut point of each initial individual and the population is filled to the preset population size. Based on the total power collection cost of each candidate individual, a two-stage screening process is adopted within a preset population size. The process is as follows: first, all candidate individuals are approximated by straight-line distance to replace the actual obstacle bypass path, and a proxy cost assessment is run. The candidates are ranked according to their proxy costs, and only the top-ranked candidates are subjected to a precise assessment.
[0058] Specifically, the collector circuit partitioning problem can be formalized as: given Typhoon generator, and number of collector circuits The wind turbine assembly is divided into Each subset (partition) corresponds to a collector circuit, minimizing the total collector cost. This invention transforms the combinatorial optimization problem into a continuous evolutionary search problem of an ordered sequence of cut points.
[0059] Coding design (tangent vector polar angle sorting): based on the booster station Calculate the polar angle of each wind turbine, taking the pole as the reference point. Arrange in ascending order of polar angle The typhoon generators are arranged in an ordered sequence. .by The vector of each tangent point position (satisfy ) represents a partitioning scheme: the Each zone contains fans. (Agreement) This encoding inherently guarantees that each partition is a sector with continuous polar angles, avoiding geometrically non-convex partition shapes that are unfavorable for routing, while compressing the search space from an exponential number of combinations to... An ordered vector space of 3D integers.
[0060] Initialization (based on bidirectional rotation of angle): First, a reference tangent vector is generated using a uniform cutting method. (Right now Then, the individuals are offset in both the forward and reverse directions (equivalent to rotating and grouping sectors) to generate several initial individuals with geometric diversity; and random integer perturbations are applied to the tangent points of each initial individual to fill the population to the preset size. .
[0061] Fitness assessment (two-stage screening within the population): for each candidate individual Calculate the objective function value—total power collection cost. When using a two-stage in-population screening approach, first, all candidate individuals are approximated by straight-line distances to represent actual obstacle avoidance paths. A proxy cost assessment is then performed, and individuals are ranked according to proxy cost. Only the top-ranked individuals are selected. Actuarial calculations for candidate individuals (including Theta) (By using obstacle avoidance paths), the total computational cost of population evaluation is significantly reduced while ensuring that no optimal candidate is overlooked. The precise evaluation of each candidate individual is executed in parallel through a process pool.
[0062] Parallel evaluation: The actuarial evaluation of each candidate individual is executed in parallel through a process pool. Each working process independently holds a model snapshot and passes position parameters through a temporary configuration file to avoid competition caused by sharing database connections between processes.
[0063] Specifically, the genetic operator design process includes: Step a3. Elite Preservation: In each generation, directly preserve the one with the best fitness. Each individual enters the next generation, ensuring that the optimal solution is monotonically non-degenerate.
[0064] Step b3. Nearest neighbor-guided crossover operator: Given an individual from parent generation 1 Individuals from parent generation 2 ,exist a certain point Nearby, from Find the nearest geometrically distant (by polar distance) A number of candidate cut points, guided by this, from... The extraction ratio is Insertion of a segment of the cutpoint subsequence This operator generates offspring. It prioritizes recombination between geometrically adjacent partition boundaries to avoid generating meaningless combinations of tangent points with large polar distance spans.
[0065] Step c3. Boundary mutation operator: with probability Trigger mutation. For each current partition, extract the set of leaf nodes (terminal wind turbines connected to only one uplink edge) from its proxy MST as movable candidate nodes; for each candidate leaf node, select it from all other partitions according to Euclidean distance. For each nearest neighbor target partition, two types of action candidates are generated: Move — Move the leaf node to the target partition (must meet the current partition size). Target partition size ); Swap—Swap this node with a leaf node of the target partition (partition size remains unchanged). All actions are sorted in descending order of agency cost improvement, and each action is assigned a soft acceptance probability (improvement amount). Time takes the base probability Improvement amount The decision to execute an action is based on an exponential decay (the action decreases exponentially). The accepted action updates the partition set progressively. After execution, the partitions are rearranged by polar angle and the tangent vector is updated.
[0066] Step d3. Destroy-Repair Mutation Operator: with probability trigger.
[0067] Destruction Phase: Select from the current partitioning scheme The typhoon generator is temporarily removed from its partition (becoming a "free" node); during selection, the leaf nodes of each partition's proxy MST are used as the bias candidate pool, based on probability. Prioritize extraction from leaf nodes (with boundary offset), and randomly extract the rest from the entire field to ensure that the removed nodes are concentrated at the partition connectivity boundary rather than inside, thus avoiding damage to the partition core topology.
[0068] Repair phase: Randomly shuffle the detached nodes and then greedily redistribute them one by one—for each detached node, traverse all target partitions that meet the capacity feasibility, and use triples to... The nodes are sorted (in order of incremental agent cost to add to the partition, deviation from the target size, and shortest distance to each leaf node of the target partition) and assigned to the highest-ranked target partition. After all nodes are assigned, the partition size repair function is called to gradually transfer leaf nodes between partitions to ensure that the size of each partition meets the target partition requirements. Constraints are applied to generate legitimate candidate individuals.
[0069] Step e3. Scale partitioning mutation operator: For the partition scale vector Apply a perturbation. With a probability of 0.25, when the condition is met... , Under the constraint of full resampling of the entire field; otherwise, a boundary between adjacent partitions is randomly selected. (Partition) and (between), check whether the two transfer directions are feasible ( and If both directions are feasible, one is randomly selected; if only one direction is feasible, it is executed; if neither is feasible, the scale vector remains unchanged. With the new scale vector as the target, the leaf nodes are gradually transferred between adjacent partitions through the partition repair function so that the actual number of units in each partition is consistent with the new target, and a legal new individual is generated.
[0070] Step f3. Local Improvement: For the current optimal individual after refinement, perform several rounds of local search with single-cutting point movement, enumerate the positions of the nearest cutting points in the neighborhood of each cutting point and select the optimal movement to achieve final refinement.
[0071] Early Stopping Mechanism: Monitoring the optimal fitness improvement amount in each generation When continuous Substitute to satisfy At that time, evolution is prematurely terminated. Among them, and The first and Optimal fitness of generation; This represents the initial optimal fitness. The preset ratio threshold is a value set by those skilled in the art based on experience, and its value is between 0 and 1.
[0072] Intergenerational Incremental LRU Cache: The evolutionary algorithm maintains a four-layer in-memory LRU cache during iteration, storing the proxy cost, partition boundary leaf nodes, proxy topology tree, and cutoff point pair scores of historical partitioning schemes, and retaining an effective window across generations. This allows the same partitioning scheme to directly hit the cache when repeatedly evaluated across different generations, significantly improving the evaluation throughput of each generation. Taking the statistics from generation 2 onwards as an example, the number of cache hits per generation can reach more than 150,000, with a hit rate exceeding 70%, effectively reducing a large amount of redundant computation.
[0073] Step S104: Within the results of each wind turbine power collection zone, based on the system cost weighted minimum spanning tree algorithm and combined with preset electrical and cost parameters, generate a power collection loop topology with each candidate optimal booster station as the root node.
[0074] The process of generating the collector loop topology with each candidate optimal booster station as the root node is as follows: Step S1041: Within the power collection zone results of each wind turbine, with the substation as the root node, run the system cost weighted minimum spanning tree algorithm to perform wiring within the power collection loop, using the system total cost increment as the edge weight; assume the currently constructed rooted tree... Candidate edges (Connect nodes already in the tree) wind turbines that have not yet been connected The edge weight is defined as the increase in the system cost of the tree after adding that edge:
[0075] in This represents the total cost of the current tree, including material costs, construction costs, marine usage fees, cross-fees, and switchgear costs.
[0076] Starting from the booster station, a rooted tree is constructed using a recursive depth-first search. At each level of recursion, the following operations are performed: Step a4. Candidate edge enumeration and constraint filtering: Traverse the set of currently connected nodes and find all edges whose degree has not reached the upper limit. The Cartesian product of the nodes and the remaining unconnected wind turbines is used to enumerate all candidate directed edges. For each candidate edge, check in sequence: (1) Bifurcation angle constraint: the angle between the edge and the existing edge of the same node is not less than 1. (2) Minimum clearance constraint: The minimum geometric distance between the new edge and the existing edge is not less than (3) Boundary / obstacle constraints: If the direct connection line is unobstructed within the field area, it can be used directly; otherwise, Theta is called. (3) Obstacle bypass path search for the actual feasible path of the edge; (4) Topological conflict constraint: does not intersect with the outgoing cable path or existing loop. Candidate edges that do not satisfy any constraint are skipped.
[0077] Step b4. Global Cost Sorting and Greedy Priority Trial: Sort all candidate edges that have passed the constraint test in this layer according to the system cost increment. Sort the edges globally in ascending order and try them one by one. Find the edge with the minimum cost. Add to diagram The algorithm recursively enters the next layer to continue connecting the remaining wind turbines; if the recursion of the next layer returns successfully, the current edge selection is confirmed, and the algorithm exits.
[0078] Step c4. Backtracking and recovery: If... If the recursive search starting from the first edge fails to connect all wind turbines (i.e., a wind turbine becomes an isolated node in a deeper layer because all its neighbors have reached the degree limit), then the edge is removed from the graph, the corresponding node degree change and the sink node mark are revoked, and the next candidate edge with the lowest cost in this layer is tried. If all candidate edges in this layer have failed, the failure is returned upwards, triggering backtracking in the upper layer.
[0079] The above process prioritizes the edge with the lowest system cost at each level, and only triggers backtracking if the greedy choice leads to global infeasibility. This ensures constraint feasibility while approximating the greedy optimal solution as closely as possible. The upper limit of the recursion depth is set to... If the time limit is exceeded, return failure directly to prevent the search space from being exhausted.
[0080] Step S1042: Select the appropriate type of submarine cable segment by segment based on the required current carrying capacity of each collector circuit.
[0081] Suppose that there are a total of [number] downstream sections of a certain submarine cable. Typhoon turbines, with a total installed capacity of (Unit: MW), then the required current for this section is... (Unit is) )for:
[0082] in The rated line voltage of the collector (in kV). This is the current design factor (usually taken as 1.0 to 1.05). Refer to the submarine cable specification table and select the submarine cable model with the smallest cross-sectional area that meets the following conditions. :
[0083] in Model Rated current carrying capacity, This is the current margin (in A).
[0084] Step S1043: Based on the selected submarine cable segment by segment, for each collector circuit, calculate the cumulative voltage drop on each link in the series connection from the step-up station to each terminal node. Compare the result with a preset voltage drop threshold to determine if the series connection is qualified. Otherwise, automatically upgrade the submarine cable specification with the smallest cross-sectional area on the corresponding series connection, recalculate the cumulative voltage drop, and iterate until the cumulative voltage drop of all series connections does not exceed the preset voltage drop threshold. Or until the submarine cable specifications reach the upper limit of available specifications.
[0085] Among them, the preset voltage drop threshold These are values set manually based on experience. Those skilled in the art can set specific values according to actual circumstances, which will not be elaborated here.
[0086] Let the link be the first The current carrying capacity of the submarine cable segment is The actual installation length is (Unit: km), resistance per unit length is (Unit: Ω / km), inductive reactance per unit length is (Unit: Ω / km), then the voltage drop in this section for:
[0087] voltage drop The unit is If the cumulative voltage drop at a certain terminal node satisfies If the cable meets the voltage drop constraint, the link is considered qualified; otherwise, the specification of the submarine cable with the smallest cross-sectional area on the link is automatically upgraded, the calculation is recalculated, and the iteration continues until all links meet the voltage drop constraint. Or until the submarine cable specifications reach the upper limit of available specifications.
[0088] Step S105: Call the geometric obstacle bypass path planning algorithm on each collector loop segment in the collector loop topology to generate a submarine cable polyline path that satisfies the set constraints.
[0089] For each edge in the Prim topology, an improved Theta method is used. The algorithm performs geometric path planning within the gridded field area, replacing the original straight-line connections with polyline paths.
[0090] Step 1 (Obstacle Modeling): Model the laid submarine cable segments according to the buffer width. The obstacles are expanded into rectangular polygons, and the locations of each wind turbine foundation are expanded into circular obstacles according to a safe radius. The area outside the site boundary is used as a global boundary constraint, and the obstacles are merged to generate an obstacle map.
[0091] Step Two (Fixed Obstacle Grid Cache): For fixed obstacles (area boundaries, planned submarine cables) in this path planning task, store the corresponding grid occupancy map in the cache using the hash signature of the obstacle set as the key. Subsequent Theta operations on the same fixed obstacle map will then be performed. Calling the cache directly avoids the computational overhead of repeatedly building obstacle grids.
[0092] Step 3 (Theta) Grid path search): at grid resolution The field area is gridded, starting from the starting point. (Wind turbine or booster station) and endpoint The raster where the element is located serves as the start and end node. The Theta algorithm is then run on the map occupying that raster. Search. The algorithm maintains a list of... For an open set of priorities (min-heap), where From To the current node The known shortest actual path length (cumulative Euclidean distance). The heuristic distance to the endpoint is the straight-line distance. Each round, elements are removed from the heap. smallest node Expand and traverse its 8 neighboring nodes (horizontal, vertical, and diagonal). For each neighbor, perform the following judgment: Step a5. Line of sight priority (Path2): Let... The current parent node is , test from arrive Is the straight segment unobstructed? If so, calculate... ;like Then let skip Directly Connected to Instead of via a shorter diagonal line The broken line segment.
[0093] Step b5. Standard expansion (Path1): If If the line of sight is obstructed, then revert to standard A. Steps: Inspection Is there any obstacle? If so, calculate. If it is better, then let ,Will Connected to .
[0094] In both cases, the updated version will be used. by Push it into the pile. After the search reaches its destination, follow... By tracing back the link using pointers, each parent node is converted to world coordinates in turn, resulting in a sequence of vertices. .
[0095] Step 4 (Path Acceptance and Fallback): If Theta Returns a polyline path whose length does not exceed the straight-line distance. The original straight line connection is replaced with a sequence of vertex coordinates, and the geometric length and cost of the corresponding edge are updated; if the search fails or the length limit is exceeded, the straight line connection is retained.
[0096] Step 5 (Laying Length Coefficient Correction): For all paths (whether straight or broken), the actual installation length is calculated as follows: (Unit is) ):
[0097] in The geometric length of the path (in meters). The laying length factor (considering actual laying process factors such as seabed undulations and binding allowances of submarine cables, it is usually taken as 1.03 to 1.08). The length of the vertical section of the submarine cable at the substation and wind turbine (in meters) is used for cost calculation.
[0098] Step S106: Automatically select the submarine cable specifications and models based on the downstream capacity of each submarine cable segment and summarize relevant cost information to obtain a joint optimization scheme; the joint optimization scheme includes the candidate optimal booster station location, the results of each wind turbine power collection zone and its corresponding power collection circuit topology, the submarine cable broken line path, the submarine cable specifications and models and a detailed breakdown of total construction costs.
[0099] Define a complete engineering cost model, total cost of the power collection system It consists of the following components:
[0100] in, For material costs; For construction costs; For sea (land) use fees; For cross-fees; For switch cabinet fees.
[0101] (1) Material costs:
[0102] in For the assembly of submarine cable segments, For the first The selected submarine cable specifications, The corresponding material unit price (yuan / km) Installation length (km).
[0103] (2) Construction cost:
[0104] in The unit price for construction is (yuan / km).
[0105] (3) Sea (land) use fee: The cost of submarine cable corridors is calculated based on the length of the route and the width of the corridor. The product is used to calculate the occupied area, discounted to the construction period:
[0106] in The annual fee is per unit area (RMB / ha / year). The annuity discount factor is... The discount rate is... The service life (in years).
[0107] (4) Crossing surcharge: Count the number of crosses between each collector submarine cable route and known pipelines / cables and other constraints within the site. Additional fees for each type of cross-category single transaction Grand total:
[0108] (5) Switchgear fee:
[0109] in The number of collector circuits, Cost of a single-circuit switchgear (in yuan).
[0110] Each cost item can be independently enabled or disabled via a configuration switch, allowing for flexible selection of the scope of inclusion based on the accuracy requirements of each engineering stage, and is independent of the agency / actuarial evaluation mode.
[0111] Assume the total installed capacity of the wind turbine is (Unit: MW), the voltage of the collector cable is... (Unit: kV), power factor: The current design factor is The current carrying capacity corresponding to the maximum rated current specification is (Unit: A), then the maximum number of wind turbines that can be connected to each circuit. for:
[0112] in Unit capacity (in MW). Number of feasible circuits. satisfy:
[0113] in The constraint is the minimum number of fans per loop. For all loops that satisfy this constraint... Values constitute a candidate set It independently performs a wandering search, comprehensively compares and selects the globally optimal combination. Simultaneously, it calculates the main transformer capacity level and number of units based on the total capacity.
[0114] For step-up substations with multiple main transformers, this invention uses the Hungarian algorithm (a linear assignment problem) to evenly distribute the installed capacity of the collector circuit to each main transformer, minimizing the maximum deviation in the total connected capacity of each main transformer. Assume there are a total of... strip collector circuit and The main transformer in Taiwan has an installed capacity of [missing information - likely related to power generation]. Define the assignment matrix The optimization problem then becomes:
[0115]
[0116] in To achieve a balanced objective, when the capacity deviation of each main transformer exceeds the user-defined tolerance, the partitioning scheme is marked as infeasible, guiding the evolutionary algorithm to search for and avoid such schemes.
[0117] Figure 4The cross-sectional specifications of submarine cables are distinguished by five line types: solid thick line for 3×500mm², double dotted line for 3×400mm², single dotted line for 3×240mm², long dashed line for 3×120mm², and thin dotted line for 3×95mm²; the outgoing cable (3×630mm²) is separately marked with a medium-pitch dashed line. A solid black square indicates the optimal location for an offshore substation. The solid black triangle represents the landing point. Hollow circles represent the locations of each wind turbine (including their numbers). The site boundary is drawn with sparse dashed lines, and the four existing submarine pipelines are drawn with thin dotted lines. The rectangular restricted area in the southern part of the site is marked with a solid black rectangle (500m wide and 1200m long, located at coordinates). Nearby); the two submarine cable paths for collecting power (A43→A56, A45→A53) on either side of the restricted area are located by Theta The algorithm plans a three-fold bypass path via the flanks of the restricted area, increasing the obstacle bypass length by 172.74m (+6.5%) and 160.41m (+6.7%), respectively. Figure 4 The deflection of the straight line connection is visible, and the five existing pipeline intersections are distributed between the wiring loops.
[0118] like Figure 2 As shown, the offshore wind farm collector circuit optimization system and the booster station location optimization system provided in this embodiment of the invention can be implemented in software. The offshore wind farm collector circuit optimization system and the booster station location optimization system include the following software modules: The feasible region and constraint construction module 201 for the booster station is used to construct the feasible region determination function and geometric constraint environment of the booster station based on the site boundary polygon, the set of obstacles and the preset safety distance constraint of the booster station; The candidate optimal booster station location generation module 202 is used to perform a multi-starting-point adaptive step-size random walk search based on the booster station's two-dimensional coordinates as decision variables, the feasible region determination function of the booster station, and the geometric constraint environment. With the cooperation of proxy evaluation screening and actuarial evaluation confirmation, the candidate optimal booster station location is obtained. The wind turbine power collection zoning result generation module 203 is used to take the position coordinates of each wind turbine in the wind turbine position coordinate set as the pole, sort and encode based on the polar angle of the tangent vector, and obtain the wind turbine power collection zoning result that simultaneously meets the loop size constraint and the overall cost under the cooperation of proxy evaluation screening and actuarial evaluation confirmation. The collector loop topology generation module 204 is used to generate a collector loop topology with each candidate optimal substation as the root node based on the system cost weighted minimum spanning tree algorithm and the preset electrical parameters and cost parameters within the collector zone results of each wind turbine. The submarine cable polyline path planning module 205 is used to call the geometric obstacle bypass path planning algorithm on each collector loop segment in the collector loop topology to generate a submarine cable polyline path that meets the set constraints. The joint optimization scheme generation module 206 is used to automatically select the submarine cable specifications and models based on the downstream capacity of each submarine cable segment and summarize relevant cost information to obtain the joint optimization scheme. The joint optimization scheme includes the candidate optimal booster station location, the power collection zone results of each wind turbine and its corresponding power collection loop topology, the submarine cable broken path, the submarine cable specifications and models, and a detailed breakdown of the total construction cost.
[0119] It should be noted that the modules in the offshore wind farm collector circuit of the present invention and the joint optimization system of the booster station location correspond one-to-one with the steps in the offshore wind farm collector circuit and the joint optimization method of the booster station location in the above embodiments, and their specific implementation processes are the same, so they will not be repeated here.
[0120] The structure of the electronic device according to an embodiment of the present invention will be described in detail below. Figure 3 This is a schematic diagram of the composition structure of an electronic device provided in an embodiment of the present invention. It can be understood that... Figure 3 The diagram shows only an exemplary structure of the electronic device, not the entire structure. Some or all of the structures shown may be implemented as needed.
[0121] The electronic device provided in this embodiment of the invention includes: at least one processor 301, a memory 302, a user interface 303, and at least one network interface 304. The offshore wind farm circuit is coupled together by various components in a joint optimization system with a booster station location via a bus system 305. It is understood that the bus system 305 is used to realize communication between these components. In addition to a data bus, the bus system 305 also includes a power bus, a control bus, and a status signal bus. However, for clarity, in… Figure 3 The general designated all buses as Bus System 305.
[0122] The user interface 303 may include a monitor, keyboard, mouse, trackball, click wheel, buttons, touchpad, or touch screen.
[0123] It is understood that memory 302 can be volatile memory or non-volatile memory, or both. In this embodiment of the invention, memory 302 is capable of storing data to support the operation of the terminal. Examples of this data include any computer programs used to operate on the terminal, such as operating systems and applications. The operating system includes various system programs, such as framework layers, core library layers, driver layers, etc., used to implement various basic services and handle hardware-based tasks. Applications can include various applications.
[0124] In some embodiments, the offshore wind farm circuitry optimization system provided by this invention can be implemented using a combination of hardware and software. For example, the offshore wind farm circuitry optimization system provided by this invention can be a processor in the form of a hardware decoding processor, programmed to execute the offshore wind farm circuitry optimization method provided by this invention. For instance, the hardware decoding processor can employ one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), or other electronic components.
[0125] As an example, processor 301 can be an integrated circuit chip with signal processing capabilities, such as a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., wherein the general-purpose processor can be a microprocessor or any conventional processor, etc.
[0126] As an example of the hardware implementation of the offshore wind farm integration circuit and the booster station location co-optimization system provided in this embodiment of the invention, the device provided in this embodiment of the invention can be directly executed by a processor 301 in the form of a hardware decoding processor. For example, it can be implemented by one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), or other electronic components to implement the offshore wind farm integration circuit and booster station location co-optimization method provided in this embodiment of the invention.
[0127] The memory 302 in this embodiment of the invention is used to store various types of data to support the operation of the offshore wind farm collection circuit by the joint optimization system with the booster station location, or to store data for execution. Figure 1The program code for the method shown. Examples of this data include: any executable instructions for operation on an offshore wind farm collector circuit co-optimization system with booster station location, such as executable instructions, and programs implementing the offshore wind farm collector circuit co-optimization method with booster station location embodiments of the present invention may be included in executable instructions.
[0128] Specifically, according to embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program including functions for executing... Figure 1 The program code for the method shown. In such an embodiment, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by the central processing unit, it performs the various functions defined in the apparatus of this application.
[0129] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as 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.
[0130] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A method for jointly optimizing the circuit layout of an offshore wind farm and the location of a booster station, characterized in that, include: Based on the boundary polygon of the site area, the set of obstacles, and the preset safety distance constraints of the booster station, a feasible domain determination function and geometric constraint environment for the booster station are constructed. Using the two-dimensional coordinates of the booster station as the decision variable, and based on the feasible region determination function and geometric constraints of the booster station, a multi-starting-point adaptive step-size random walk search is performed. With the cooperation of surrogate evaluation screening and actuarial evaluation confirmation, the candidate optimal booster station location is obtained. By taking the location coordinates of each wind turbine in the set of wind turbine location coordinates as poles, and sorting and encoding them based on the polar angle of the tangent vector, and with the cooperation of proxy evaluation screening and actuarial evaluation confirmation, the wind turbine collection partitioning results that simultaneously meet the loop size constraints and comprehensive cost are obtained. Within each wind turbine power collection zone, based on the system cost weighted minimum spanning tree algorithm and combined with preset electrical and cost parameters, a power collection loop topology with each candidate optimal booster station as the root node is generated. The geometric obstacle avoidance path planning algorithm is called on each collector loop edge segment in the collector loop topology to generate a submarine cable polyline path that meets the set constraints; The submarine cable specifications and models are automatically selected based on the downstream capacity of each segment of the submarine cable, and relevant cost information is summarized to obtain a joint optimization scheme. The joint optimization scheme includes the candidate optimal booster station location, the results of each wind turbine power collection zone and its corresponding power collection circuit topology, the submarine cable broken line path, the submarine cable specifications and models, and a detailed breakdown of the total construction cost.
2. The method for jointly optimizing the location of offshore wind farm collector circuits and booster station as described in claim 1, characterized in that, During the multi-starting point adaptive step-size random walk search, uniform random sampling is performed within the boundary polygon of the field area to select several initial candidate points that satisfy the feasible region determination function. Proxy evaluation is performed on all initial points, and they are sorted in ascending order according to proxy cost. Points with the highest ranking are selected according to a set ratio to perform actuarial evaluation to obtain the actuarial cost of each initial point, which is used as the starting cost of each random walk.
3. The method for jointly optimizing the location of offshore wind farm collector circuits and booster station as described in claim 1, characterized in that, During the multi-starting point adaptive step-size random walk search, each round of neighborhood candidate points is first quickly screened by proxy cost, and only candidate points with proxy costs better than the current best are submitted for actuarial evaluation; the actuarial evaluation determines whether the corresponding candidate point is accepted or rejected.
4. The method for jointly optimizing the location of offshore wind farm collector circuits and booster station as described in claim 1, characterized in that, The process of wind turbine collector zoning results includes: By taking the position coordinates of each wind turbine as the pole and sorting and encoding the polar angle of the tangent vector, the corresponding number of tangent position vectors are obtained, which are the candidate individuals; A reference cut point vector is generated by uniform cutting and offset in the positive and negative directions respectively to generate several initial individuals with geometric diversity. Random integer perturbation is applied to the cut point of each initial individual and the population is filled to the preset population size. Based on the total power collection cost of each candidate individual, a two-stage screening process is adopted within a preset population size. The process is as follows: first, all candidate individuals are approximated by straight-line distance to replace the actual obstacle bypass path, and a proxy cost assessment is run. The candidates are ranked according to their proxy costs, and only the top-ranked candidates are subjected to a precise assessment.
5. The method for jointly optimizing the location of offshore wind farm collector circuits and booster station as described in claim 1, characterized in that, The process of generating the collector loop topology with each candidate optimal booster station as the root node is as follows: Within the power collection zone of each wind turbine, with the substation as the root node, the wiring within the power collection loop is performed using the weighted minimum spanning tree algorithm based on system cost, with the increment of the total system cost as the edge weight. Based on the required current flow of each collector circuit, the submarine cable is selected segment by segment. Based on the selected submarine cable segment by segment, for each collector circuit, the cumulative voltage drop is calculated link by link on the series link from the step-up station to each terminal node. The results are compared with the preset voltage drop threshold to determine whether the series link is qualified. Otherwise, the submarine cable specification with the smallest cross-sectional area on the corresponding series link is automatically upgraded, the cumulative voltage drop is recalculated, and the iteration continues until the cumulative voltage drop of all series links does not exceed the preset voltage drop threshold or the submarine cable specification reaches the upper limit of the available specification.
6. The method for jointly optimizing the location of offshore wind farm collector circuits and booster station as described in claim 1, characterized in that, The same cost model and submarine cable selection rules were used for the agency evaluation screening and actuarial evaluation confirmation.
7. The method for jointly optimizing the location of offshore wind farm collector circuits and booster station as described in claim 1, characterized in that, During the agent evaluation and selection process, the total agent cost is the sum of the outgoing cable agent cost and the geometric default penalty; the outgoing cable agent cost is estimated based on the straight-line distance from the booster station to the landing point; during the agent evaluation and selection process, only the initial population is evaluated, and genetic operator iteration is not performed; During the actuarial assessment and confirmation process, the total actuarial cost of the entire site is the sum of the precise cost of the outgoing cables and the precise cost increments of the wiring for all wind turbine collection zones and all substations; the precise cost of the outgoing cables is estimated based on the obstacle-bypassing broken line distance from the substation to the landing point; and the actuarial assessment and confirmation process involves the complete multi-generation evolution.
8. A system for joint optimization of offshore wind farm circuit layout and substation location, characterized in that, The method for joint optimization of offshore wind farm collector circuit composition and booster station location as described in any one of claims 1-7 includes: The module for constructing the feasible region and constraints of the booster station is used to construct the feasible region determination function and geometric constraint environment of the booster station based on the site boundary polygon, the set of obstacles, and the preset safety distance constraints of the booster station. The candidate optimal booster station location generation module is used to perform a multi-starting-point adaptive step-size random walk search based on the booster station's two-dimensional coordinates as decision variables, the booster station's feasible region determination function, and the geometric constraint environment. With the cooperation of proxy evaluation screening and actuarial evaluation confirmation, the candidate optimal booster station location is obtained. The wind turbine power collection zoning result generation module is used to take the position coordinates of each wind turbine in the wind turbine position coordinate set as the pole, sort and encode them based on the polar angle of the tangent vector, and obtain the wind turbine power collection zoning result that simultaneously meets the loop size constraint and the overall cost under the cooperation of proxy evaluation screening and actuarial evaluation confirmation. The collector loop topology generation module is used to generate a collector loop topology with each candidate optimal substation as the root node, based on the system cost weighted minimum spanning tree algorithm and in combination with preset electrical and cost parameters, within the collector zone results of each wind turbine. The submarine cable polyline path planning module is used to call the geometric obstacle bypass path planning algorithm on each collector loop segment in the collector loop topology to generate a submarine cable polyline path that meets the set constraints. The joint optimization scheme generation module is used to automatically select the submarine cable specifications and models based on the downstream capacity of each submarine cable segment and summarize relevant cost information to obtain the joint optimization scheme. The joint optimization scheme includes the candidate optimal booster station location, the power collection zone results of each wind turbine and its corresponding power collection loop topology, the submarine cable broken line path, the submarine cable specifications and models, and a detailed breakdown of the total construction cost.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the method for joint optimization of offshore wind farm collection circuits with booster station locations as described in any one of claims 1-7.
10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the method for joint optimization of offshore wind farm collection circuits and booster station locations as described in any one of claims 1-7.
