Offshore wind power plant layout evaluation and optimization method based on sea space resource constraint

Abstract

The invention belongs to the technical field of wind power generation, and particularly discloses an offshore wind power plant layout evaluation and optimization method based on sea space resource constraint, which comprises the following steps of: constructing a dynamic coupling functional block diagram and a hierarchical constraint space model by fusing multi-source sea data, and evaluating the space compatibility, constructability and ecological accumulation influence of a layout scheme according to the dynamic coupling functional block diagram and the hierarchical constraint space model; and extracting a comprehensive efficiency feature vector to drive a multi-objective optimization algorithm, optimizing a fan position and a cable path through main and auxiliary group interactive iteration, and finally generating an optimized conflict-free wind power plant layout scheme in combination with dynamic constraint boundary adjustment and a conflict verification relaxation strategy. Dynamic collaborative optimization of wind power plant layout and matched resources can be realized, and improvement of comprehensive feasibility, economy and environmental compatibility of a project scheme is facilitated.

Description

Technical Field

[0001] This invention belongs to the field of wind power generation technology and relates to a method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints. Background Technology

[0002] Offshore wind farms, as an important form of renewable energy utilization, essentially involve deploying multiple wind turbines within a specific sea area to capture wind energy and convert it into electricity. The layout of the wind turbines—the specific locations of the turbines in the sea—is a crucial aspect of offshore wind farm planning and design, directly determining the farm's power generation efficiency, construction and operation costs, and impact on the marine environment. Therefore, scientifically and rationally optimizing the layout of offshore wind farms is essential for improving project economics, ensuring safe operation, and promoting sustainable marine development.

[0003] In existing offshore wind farm layout designs, the common approach is to first identify various restrictive areas in the sea, such as waterways, military restricted zones, marine protected areas, and submarine pipelines, and mark these areas as no-go zones for wind turbine installation in a geographic information system. Then, within the remaining available sea areas, optimization algorithms are used to search for the optimal combination of wind turbine locations, with the primary objective of maximizing annual power generation or minimizing wake losses. Some advanced methods also incorporate factors such as cable length and cost into the optimization objective, performing multi-objective optimization.

[0004] However, existing technical solutions have significant shortcomings in handling complex marine spatial resource constraints. They generally treat various constraints as simple superpositions of independent static geographical boundaries, ignoring the dynamic coupling relationships and time-varying characteristics between these marine functional activities, such as the tidal variations in shipping traffic and the seasonality of fishing activities. Furthermore, the layout assessment and optimization processes are often separated, and the assessment results fail to provide real-time feedback to the optimization algorithm's search direction, making the optimization process prone to getting trapped in local optima. This simplified constraint handling and fragmented optimization process can easily lead to unforeseen conflict risks or economic losses in the final layout scheme. Summary of the Invention

[0005] In view of this, in order to solve the problems mentioned in the background technology, a method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints is proposed.

[0006] The objective of this invention can be achieved through the following technical solution: This invention provides a method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints, including: acquiring multi-source spatiotemporal marine resource data, integrating channel trajectory data, fishery activity cycles, ecological monitoring data, sea condition forecast information, seabed geological data and regional planning information, and constructing a dynamically coupled marine functional block map.

[0007] Based on the dynamically coupled marine functional block map, the constraint correlation is quantified to generate a hierarchical constraint space model.

[0008] The initial wind turbine location parameters are set, and the spatial compatibility index, constructability score, and ecological cumulative impact are calculated based on the hierarchical constraint space model to generate multi-dimensional performance evaluation results.

[0009] Key performance indicators are extracted from the multidimensional performance evaluation results, and a comprehensive performance feature vector is generated through feature aggregation processing.

[0010] The comprehensive performance feature vector is input into a multi-objective optimization algorithm to drive the main wind turbine location group and the auxiliary resource scheduling group to interact and iterate, and output an optimized wind turbine coordinate set and supporting cable path scheme.

[0011] The buffer boundary thresholds in the hierarchical constraint space model are dynamically adjusted based on the optimized wind turbine coordinate set to generate an updated constraint space model.

[0012] Based on the updated constraint space model, conflict verification is performed on the optimized wind turbine coordinate set. When a spatial conflict is detected, a constraint relaxation strategy based on penalty cost is triggered to generate an optimized conflict-free layout scheme.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0014] (1) This invention constructs a dynamically coupled marine functional block map and generates a dynamically updatable hierarchical constraint space model on this basis, so that the wind farm layout decision can respond in real time to changes in the time and space such as waterway flow and fishery activities, overcoming the problem that traditional static planning leads to rigid schemes or frequent conflicts with actual operation, and improving the adaptability of the layout scheme to changes in the marine environment.

[0015] (2) This invention establishes a multi-dimensional performance evaluation system that includes spatial compatibility, constructability, ecological cumulative impact, power generation benefits and investment costs. Through feature extraction and aggregation processing, the optimization process can systematically balance multiple objectives such as energy development, engineering economy, environmental protection and marine space sharing, thereby obtaining a layout scheme with better comprehensive benefits and avoiding the one-sidedness and potential risks caused by single-objective optimization.

[0016] (3) The present invention adopts a multi-objective optimization algorithm to drive the parallel interactive iteration of auxiliary resource scheduling such as wind turbine location layout and cable path scheme, realizing the deep coupling and synchronous optimization of layout planning and resource matching. The resulting system-level gain effect surpasses the simple superposition of traditional step-by-step and serial decision-making, reducing the overall construction cost and technical risk of the project.

[0017] (4) This invention introduces a constraint relaxation strategy based on conflict verification and a dynamic model adjustment mechanism, which has the ability to make flexible decisions and adaptive adjustments under complex constraints. It can intelligently resolve spatial conflicts in non-critical areas while ensuring that core interests are not infringed, thereby enhancing the feasibility and high-efficiency implementation of the final output solution in actual engineering. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the implementation steps of the method of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Please see Figure 1 As shown, this invention provides a method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints. The specific steps are as follows: S1. Obtain multi-source spatiotemporal marine resource data, integrate channel trajectory data, fishery activity records, ecological monitoring data, sea condition forecast information, seabed geological data and regional planning information, and construct a dynamically coupled marine functional block map.

[0022] Specifically, a dynamically coupled marine functional block map that reflects the spatiotemporal changes and intrinsic relationships of marine resources is constructed to provide a highly simulated decision-making environment for subsequent layout assessment and optimization. First, using Geographic Information System (GIS) tools, the marine planning electronic map and nautical chart from the regional planning information, as well as the ecological protection red line scope provided in the ecological monitoring data, need to be loaded and parsed to identify and vectorize the geographical boundaries of three types of areas in the target marine space. These three types of areas include: no-navigation zones where the construction of any wind power facilities is explicitly prohibited, such as the centerline of the main shipping channel; buffer zones surrounding the no-navigation zones that allow limited activities but require strict control of their impact; and negotiation zones where wind power development can be coordinated with other marine utilization activities through technical or management means, such as seasonal fishing grounds or secondary shipping channels.

[0023] Subsequently, a channel-anchorage coupling impact coefficient is introduced to assess the impact of channel traffic on the safety and availability of adjacent anchorages. Establishing this coefficient first requires processing channel trajectory data, extracting vessel traffic flow, speed, and vessel type within a specific channel in the target sea area, and identifying vessel dwell time and density within the anchorage area to create an anchorage usage record. The calculation formula is then used... Obtain the channel-anchorage coupling influence coefficient of point m within the anchorage at time t. This coefficient is a dimensionless parameter. The normalized vessel traffic flow within the channel at time t; is the shortest distance from point m to the channel boundary L; k is an empirical weighting constant set according to maritime safety regulations, used to transform static geographical proximity into dynamic influence intensity.

[0024] Next, a spatiotemporal conflict probability matrix is ​​constructed to reveal the potential conflicts between wind farm activities and time-varying marine activities. Based on the sea state safety threshold for operation of maintenance vessels defined in the wind farm design and operation and maintenance scheme, and combined with the sea state forecast information, time window prediction is performed. Specifically, the prediction involves determining whether the sea state conditions of each time unit within the planned operation and maintenance cycle meet the safety threshold for operation of maintenance vessels, based on the time-series wave, wind speed, and ocean current data of the target marine area in the sea state forecast information.

[0025] The time periods that continuously meet the safety threshold conditions are identified and extracted as candidate maintenance time windows.

[0026] Based on long-term simulations of historical meteorological statistics or sea state forecast information, the probability of occurrence and average duration of each candidate maintenance time window within the planned operation and maintenance cycle are calculated. This generates a structured list of predicted maintenance time windows for the planned operation and maintenance cycle. In this list, each predicted maintenance time window is associated with an expected time range, an estimated probability of occurrence, and an estimated available duration, thereby completing the quantitative prediction of wind farm maintenance time windows.

[0027] For each grid cell in the target sea area, the following calculations are performed within a specific time period, such as a quarter or a year:

[0028] Retrieve fishery activity records, calculate the historical frequency of fishery activity in each basic time unit (e.g., daily) within the grid cell, and quantify this frequency as the probability of fishery activity in that cell. , where i is the grid cell index and t is the basic time cell index.

[0029] Based on the predicted wind farm maintenance time windows, the set of all time windows for planned maintenance activities within the cycle is determined.

[0030] For each basic time unit within a specific time period, determine whether it falls within any planned maintenance window. If so, define a maintenance activity occurrence flag for that time unit. If the value is 1, then the value is 0; otherwise, the value is 0.

[0031] Through formula The probability of a spatiotemporal conflict occurring in a grid cell within a specific time period is calculated. This formula calculates the average probability of fishery activities occurring concurrently in the grid cell across all planned wind farm maintenance time cells. This represents the likelihood of two types of activities occurring simultaneously in the grid cell, thus triggering a potential conflict. This generates a two-dimensional spatiotemporal conflict probability matrix. The value of the matrix represents the probability of a functional conflict occurring at that point in time.

[0032] Finally, multi-layered data fusion is performed. The generated geographical boundaries, waterway-anchorage coupling influence coefficients, and spatiotemporal conflict probability matrices are superimposed, and spatiotemporal weighting factors are introduced. These spatiotemporal weighting factors are determined based on three dimensions: historical navigation density, fishery activity frequency, and ecological sensitivity, through a combination of analytic hierarchy process (AHP) and expert scoring, and are dynamically adjusted according to season and time period.

[0033] This fusion process ultimately generates a dynamically coupled marine functional zone map. This map is not a traditional static partition map, but a multi-dimensional data structure in which each data point contains the initial constraint strength at that location at a specific time, thereby achieving accurate modeling of the dynamic and coupled characteristics of marine spatial resources.

[0034] This invention constructs a dynamically coupled marine functional block map and generates a dynamically updatable hierarchical constraint space model based on it. This enables wind farm layout decisions to respond in real time to spatiotemporal changes such as channel traffic and fishing activities, overcoming the problems of rigid schemes or frequent conflicts with actual operation caused by traditional static planning, and improving the adaptability of the layout scheme to changes in the marine environment.

[0035] S2. Based on the dynamically coupled marine functional block map, quantify the constraint correlation to generate a hierarchical constraint space model.

[0036] The generation process of the hierarchical constraint space model includes: based on the dynamically coupled marine functional block map, extracting the initial constraint intensity of each grid cell as the basic constraint field as it changes over time.

[0037] Based on the multi-source data and intermediate calculation results used to construct the dynamically coupled marine functional zoning map, a reverse mapping is performed to generate an independent quantitative constraint layer. Specifically, based on the regional planning information input into the dynamically coupled marine functional zoning map, the original geographic boundaries and attributes of navigation restricted areas, buffer zones, and negotiation zones are directly extracted. Each zone is assigned a fixed constraint strength benchmark value, for example, 1.0 for restricted areas, 0.6 for buffer zones, and 0.3 for negotiation zones, thereby generating a spatial geographic constraint layer.

[0038] Based on the calculation results of the channel-anchorage coupling influence coefficient of the input dynamic coupled sea area functional block map, it is used as a dynamic influence intensity field, directly mapped and resampled onto a unified spatial grid to form a functional activity coupling constraint layer.

[0039] Based on the spatiotemporal conflict probability matrix of the input dynamic coupled marine functional block map, it is directly used as the probability field representing the potential spatiotemporal conflict between fishery and wind power maintenance activities, thus forming a conflict risk constraint layer.

[0040] The quantitative constraint layers representing different physical meanings are mapped to the same numerical range using the min-max normalization method to make them comparable. Each standardized constraint layer is assigned a weight coefficient, which reflects the relative importance of that type of constraint in the overall evaluation. This coefficient can be pre-set using a combination of expert scoring and analytic hierarchy process. Specifically, the spatial geographic constraint layer has the highest weight coefficient, followed by the conflict risk constraint layer, and the functional activity coupling constraint layer has the lowest weight coefficient, with the sum of the three weight coefficients being 1.

[0041] The comprehensive constraint strength of each spatial grid cell at a specific time step is calculated using a weighted summation method.

[0042] A pre-defined set of ordered threshold intervals is used to classify the calculated comprehensive constraint strength values ​​across the entire domain into different levels. The threshold intervals are set based on statistical analysis results of the comprehensive constraint strength values ​​and marine management policy requirements. The specific method is as follows:

[0043] Before applying the model to the target sea area, a large number of random points are generated using historical data or simulations to calculate their comprehensive constraint intensity values ​​and generate a distribution histogram. Combined with the management regulations of sea area functional zoning, such as the target area ratio for priority development zones and the clearly defined scope of prohibited development zones, the boundary points on the distribution histogram that can classify the sea area into the expected management intensity are selected as thresholds. For example, if the policy requires that approximately 30% of the sea area be designated as priority development zones, the 30th percentile of the cumulative distribution function of the comprehensive constraint intensity value can be used as the first threshold; the 85th percentile of the cumulative distribution function can be used as the second threshold, thereby dividing the sea area into low, medium, and high constraint levels.

[0044] Ultimately, a hierarchical constraint spatial model with consistent attributes across the spatiotemporal dimensions is generated. This model, in the form of grid data or hierarchical zoning maps, intuitively identifies the development restriction level of each location in the target sea area at any planning time, providing direct and quantifiable spatial access rules for the evaluation and selection of subsequent layout schemes.

[0045] S3. Set the initial wind turbine location parameters, calculate the spatial compatibility index, constructability score and ecological cumulative impact based on the hierarchical constraint space model, and generate multi-dimensional performance evaluation results.

[0046] The generation of multi-dimensional performance evaluation results includes: inputting the initial wind turbine location coordinates, the wind turbine standard performance curve, and the annual wind resource data of the target sea area obtained through historical data statistics; and calculating the annual equivalent full-load hours of the wind farm in the target sea area based on the wind turbine wake model. It should be noted that the wind turbine wake model is a mature technology in the industry, and can adopt models such as the Jensen model or more complex computational fluid dynamics (CFD) models. The calculation logic for the annual equivalent full-load hours of the wind farm in the target sea area is as follows: by calculating the wind speed loss of each wind turbine caused by the wake effect of the upstream wind turbine, the actual power generation of each turbine under all wind conditions is determined. Finally, the total power generation is obtained by integrating the total power generation for the whole year and divided by the total installed capacity to obtain the annual equivalent full-load hours.

[0047] Based on the input wind turbine location coordinates and combiner station location, an optimized submarine cable connection topology with the goal of minimizing total construction cost is determined using a graph theory algorithm. The graph theory algorithm can be, for example, a minimum spanning tree or Steiner tree algorithm. Combining submarine topographic data and cable laying costs, the total cable length and total construction cost are calculated. Dividing these by the total installed capacity of the wind farm yields the investment cost per kilowatt.

[0048] A three-dimensional performance evaluation coordinate system is constructed based on the spatial compatibility index, constructability score, and cumulative ecological impact.

[0049] For each wind turbine in a given wind turbine layout scheme, its spatial compatibility index is calculated according to the following steps: taking the coordinates of the wind turbine location as the center, a circular buffer is constructed by taking a radius of 3-5 times the diameter of the wind turbine impeller.

[0050] From the hierarchical constraint space model, the comprehensive constraint strength values ​​of all grid cells within the buffer zone are extracted, and the maximum and average values ​​are identified. Simultaneously, from the spatiotemporal conflict probability matrix, the average conflict probability value of the grid cell containing the wind turbine during its primary maintenance period is extracted.

[0051] The difference between 1 and the maximum comprehensive constraint strength, the difference between 1 and the average comprehensive constraint strength, and the difference between 1 and the conflict probability value are accumulated. The accumulated result is the spatial compatibility index of each wind turbine. The spatial compatibility index of the wind turbine layout scheme is then obtained by averaging the results.

[0052] The constructability score calculation process is as follows: For each wind turbine location, the static prohibited zone and dynamic restriction window during the construction period are determined based on the constraint layer data. The value of the channel-anchorage coupling influence coefficient in the dynamic restriction window is extracted. The difference between 1 and this value is defined as the remaining time index after avoiding time loss due to dynamic conflicts. The proportion of time windows that meet the sea state conditions for the operation of key construction equipment is statistically calculated based on the sea state forecast information. Based on the seabed geological data, the key geological parameters affecting the construction of wind turbine foundations at the evaluation location are extracted. The key geological parameters include at least water depth, seabed surface sediment type, underlying soil and rock mechanical properties, and seabed slope.

[0053] Based on the preset engineering geological classification standards, the measured or inferred values ​​of each key geological parameter are mapped to a standardized sub-difficulty coefficient. The sediment type is assigned a baseline coefficient from easy to difficult according to its engineering properties, such as silt, sand, gravel, and bedrock.

[0054] Geomechanical parameters and water depth are divided into threshold ranges, corresponding to different difficulty adjustment coefficients.

[0055] When the seabed slope exceeds the safe construction threshold, an additional difficulty penalty coefficient will be triggered.

[0056] By weighted and integrated sub-difficulty coefficients of all key geological parameters, the original score of foundation treatment difficulty at the location is calculated. The original score is then normalized to a preset numerical range to generate a foundation treatment difficulty score for comprehensive evaluation of constructability. The higher the value, the greater the difficulty of foundation treatment during construction.

[0057] Multiply the sum of the remaining time index and the time window ratio by 1 and the difference between the foundation treatment difficulty score to obtain the constructability score of each wind turbine location. The constructability score of the entire layout scheme is the arithmetic mean of the constructability scores of all wind turbine locations.

[0058] The calculation process for cumulative ecological impact is as follows: Data from the constraint layers related to ecological protection in the hierarchical constraint spatial model are read. This is combined with information on the distribution, extent, and protection level of sensitive ecological areas from the ecological monitoring data.

[0059] Centered on each wind turbine foundation, cable path corridor, and the pre-defined operation and maintenance activity space, and based on the radius set according to the species activity range or pollutant diffusion model, a pre-defined ecological impact buffer zone is extended outward. The buffer zone is then spatially overlaid with various sensitive ecological areas delineated in the ecological monitoring data. Various sensitive ecological areas, such as spawning grounds, feeding grounds, migration channels, and the core area of ​​protected areas, are identified, and all areas where spatial intersection occurs are identified as affected ecologically sensitive areas.

[0060] Based on the constraint layer data in the hierarchical constraint space model, the average or maximum ecological constraint intensity value of each affected ecologically sensitive area is extracted.

[0061] Based on the type, protection level, and biodiversity index of each sensitive ecological area in the ecological monitoring data, and according to the preset rating rules, the ecological sensitivity level value is determined. The cumulative results of the normalized area, ecological constraint intensity value, and ecological sensitivity index of each affected area are aggregated to obtain the cumulative ecological impact of the wind turbine layout scheme.

[0062] It should be noted that the above-mentioned preset rating rules can be specifically based on the regulations for nature reserves, with a sensitivity level of 1 for the navigation restricted area, 0.7 for the buffer zone, and 0.5 for the negotiation zone, etc.

[0063] The annual equivalent full-load hours and unit kilowatt investment cost are mapped to the three-dimensional performance evaluation coordinate system to generate multi-dimensional performance evaluation results.

[0064] The embodiments of this invention establish a multi-dimensional performance evaluation system that includes spatial compatibility, constructability, cumulative ecological impact, power generation benefits, and investment costs. Through feature extraction and aggregation processing, the optimization process can systematically balance multiple objectives such as energy development, engineering economics, environmental protection, and marine space sharing, thereby obtaining a layout scheme with better comprehensive benefits and avoiding the one-sidedness and potential risks caused by single-objective optimization.

[0065] S4. Extract the key performance indicators from the multidimensional performance evaluation results and generate a comprehensive performance feature vector through feature aggregation processing.

[0066] The generated comprehensive performance feature vector includes three sets of key features logically separated from the multi-dimensional performance evaluation results based on their physical meaning and business attributes. The first set is the conflict risk level feature, directly derived from the previously calculated spatial compatibility index, quantifying the potential conflict between the layout scheme and the functional activities of the surrounding sea area. The second set is the economic feature, including two indicators directly reflecting the project's profitability and investment efficiency: annual equivalent full-load hours and unit kilowatt investment cost. These can be normalized and added together to serve as the quantitative indicator of the economic feature. The third set is the ecological sensitivity feature, derived from the weighted fusion result of ecological cumulative impact and constructability score, comprehensively reflecting the pressure of the layout on the ecological environment and the engineering feasibility under environmental constraints.

[0067] The conflict risk level characteristics are converted into a spatial compatibility scalar using a linear weighting method. This method assigns different risk weights to different types of spatial conflicts based on expert experience or historical data. Then, the scores of the layout schemes on each conflict item are weighted and summed to obtain a comprehensive spatial compatibility scalar. The value range of this scalar is standardized to [0,1], with higher values ​​indicating better spatial compatibility.

[0068] The economic and ecological sensitivity characteristics are mapped to a unified numerical range through normalization.

[0069] By splicing together the spatial compatibility scalar, normalized economic characteristics, and normalized ecological sensitivity characteristics, a comprehensive performance feature vector is generated.

[0070] S5. Input the comprehensive performance feature vector into the multi-objective optimization algorithm to drive the main wind turbine location group and the auxiliary resource scheduling group to interact and iterate, and output the optimized wind turbine coordinate set and the supporting cable path scheme.

[0071] The interactive iteration between the main wind turbine location group and the auxiliary resource scheduling group includes: decoding the comprehensive efficiency feature vector into an evolutionary direction instruction for the main group. This decoding process is implemented through a preset instruction mapping rule table or a lightweight feedback controller. Specifically:

[0072] Decoding based on a rule table: A pre-defined mapping table associates the numerical range of each component of the comprehensive performance feature vector with a set of predefined evolutionary strategy parameter adjustments. For example, if the spatial compatibility scalar is below a threshold, the mapping rule is: in the next generation of evolution, increase the weight of the spatial compatibility-related terms in the fitness function and trigger a mutation preference flag that moves to regions with lower constraint strength.

[0073] Controller-based decoding: A proportional-integral controller is employed. The differences between each component of the feature vector and a preset target value are used as input errors, with the preset target value being 1.0 for each component. The controller's output serves as the dynamic adjustment amount for the target weights in the main population's fitness function, thereby achieving real-time guidance of the evolutionary direction.

[0074] The auxiliary group's search space boundary is generated based on the hierarchical constraint space model. This boundary is not a simple rectangular region, but rather a complex geometric boundary consisting of navigation exclusion zones, buffer zones, and negotiation zones defined in the hierarchical constraint space model. This ensures that when the auxiliary group explores resource scheduling schemes such as cable routes or construction sequences, all its activities are naturally conducted within a compliant and risk-assessed space, avoiding ineffective searches.

[0075] The main wind turbine location group performs global position mutation based on the main group evolution direction command, generating the main group mutation result. This group consists of a large number of candidate wind turbine location layout schemes, each scheme containing a set of wind turbine coordinates. In each generation of evolution, the schemes in the main group undergo mutation and crossover operations, but these operations are not completely random, but are strictly guided by the aforementioned decoded evolution direction command. The guided global position mutation is achieved by introducing a bias term or modifying the probability distribution in the standard mutation operator. Specifically:

[0076] Taking the offset mutation method as an example: For the wind turbine coordinates to be mutated, the new candidate positions are generated by the following formula: ;

[0077] in, For standard Gaussian random mutation, The strength coefficient is determined by the strength of the decoded instruction. This serves as the guiding direction vector. When the instruction is to avoid high-collision areas, It is set as a direction vector pointing from the current position to the center of the nearest low constraint strength level region, which can be obtained by querying the hierarchical constraint space model.

[0078] Simultaneously, the auxiliary resource scheduling group performs resource scheduling optimization for each wind turbine layout scheme generated by the main group within its strictly defined search space boundaries. Its main task is to find the cable path with the lowest cost or lowest risk for a given set of wind turbine coordinates. This is achieved by running a path search algorithm on the cost map represented by the hierarchical constraint space model, generating a set of candidate cable paths.

[0079] The feasible solution fragments from the main population's mutation results are exchanged with obstacle avoidance strategies from the candidate cable paths to complete collaborative knowledge transfer. This collaborative knowledge transfer is specifically implemented through a shared fitness map mechanism, as follows: When the main population discovers that a certain region's wind turbine layout has an extremely high spatial compatibility index, it marks that region's coordinates as a search hotspot. When planning cable paths, the auxiliary population reduces the search cost of traversing this hotspot region, thereby guiding cable nodes towards that region. Conversely, when the auxiliary population discovers that a cable path traversing a certain region is extremely costly, such as due to complex seabed geology, it maps that path segment as a layout repulsion field. In the next generation of mutations, the main population increases the penalty probability of placing wind turbines in that region, thereby guiding wind turbine locations away from this high-cost path area. Through this bidirectional information mapping and feedback, implicit knowledge sharing between two independent evolutionary populations is achieved.

[0080] Specifically, after each generation of evolution, elite layout schemes with high spatial compatibility index are selected from the main population, and the stable geometric patterns presented in their wind turbine clusters, such as specific wind turbine spacing and arrangement, are extracted and encoded into a layout template. The layout template can be represented by a set of relative coordinates or a graph structure.

[0081] Meanwhile, the lowest-cost elite cable route scheme is selected from the auxiliary group, and the key path segments that successfully cross complex constraint areas are extracted and encoded into a path template, which can be represented by a series of ordered waypoints.

[0082] Store layout templates and path templates in a shared collaborative knowledge base.

[0083] At the start of the next generation of evolution, when the main group generates a new individual, it has a certain probability of randomly selecting a layout template from the knowledge base and replacing the corresponding random coordinates in the new individual. Similarly, when the auxiliary group calculates a path for a given layout, it will prioritize trying to adapt and connect the path templates in the knowledge base with the start and end points of the current layout as a high-quality initial solution for path search. In this way, locally superior structures discovered between the two groups are preserved and propagated.

[0084] The embodiments of this invention employ a multi-objective optimization algorithm to drive the parallel interactive iteration of auxiliary resource scheduling, such as wind turbine location layout and cable path scheme. This achieves deep coupling and synchronous optimization of layout planning and resource allocation, and the resulting system-level gain effect surpasses the simple superposition of traditional step-by-step and serial decision-making, reducing the overall construction cost and technical risk of the project.

[0085] S6. Dynamically adjust the buffer boundary threshold in the hierarchical constraint space model according to the optimized wind turbine coordinate set, and generate an updated constraint space model.

[0086] The dynamic adjustment of the buffer boundary threshold includes: monitoring the change trend of the spatial compatibility index within the iteration cycle to form a compatibility change trend.

[0087] When the compatibility trend indicates that the spatial compatibility index is continuously higher than the compatibility threshold, the expansion distance is calculated proportionally based on the extent of the excess of the spatial compatibility index, and the boundary translation is performed to generate the expanded buffer boundary range. Here, "continuously higher" refers to the case where the index is higher than the threshold for multiple consecutive iterations.

[0088] When the cumulative ecological impact is detected to exceed the preset warning value, the shrinkage distance is calculated proportionally based on the extent to which the cumulative ecological impact exceeds the preset warning value, and the boundary is shifted to generate the usable area of ​​the shrunken negotiated area.

[0089] It should be noted that the specific execution process for calculating the expansion distance proportionally is as follows:

[0090] Subtract the preset compatibility threshold from the spatial compatibility index of the current optimization scheme, and then divide by the preset compatibility threshold to obtain the extent of the spatial compatibility index exceeding the threshold.

[0091] A basic adjustment step size and a maximum allowable adjustment range are preset. The product of the spatial compatibility index exceeding the limit and the basic adjustment step size is used as the specific expansion distance. If the specific expansion distance is less than the maximum allowable adjustment range, the specific expansion distance is used as the final expansion distance. Otherwise, the maximum allowable adjustment range is directly set as the final expansion distance.

[0092] The logical principle for calculating the shrinkage distance and expansion distance proportionally is the same, and will not be elaborated here.

[0093] When expanding, the outer boundary of the buffer is translated outward along its normal direction. When shrinking, all boundaries of the negotiation area are translated inward along their respective inner normal directions.

[0094] The boundary translation is implemented using a polygon buffer analysis algorithm in geographic information processing. Specifically:

[0095] For buffer expansion: Call the positive buffer generation function in the GIS library, take the original buffer polygon as input, specify the buffer distance as the expansion distance, and generate a new, larger polygon as the expanded boundary range.

[0096] For negotiation area shrinkage: The negative buffer generation function in the GIS library is called, taking the original negotiation area polygon as input and specifying the buffer distance as the shrinkage distance, to generate a new, smaller polygon as the usable area after shrinkage. If the shrinkage causes the polygon to disappear or the area to be less than the minimum effective threshold, the negotiation area is marked as temporarily unavailable.

[0097] The expanded buffer boundary range and the available area of ​​the shrunken negotiation zone are written into the hierarchical constraint space model to generate the updated constraint space model. At the data processing level, writing into the updated constraint space model means replacing the original geometric features in the corresponding constraint layer of the hierarchical constraint space model with the newly generated expanded buffer polygons and shrunken negotiation zone polygons, and recalculating the constraint strength and level of the relevant mesh cells.

[0098] It should be noted that the above compatibility thresholds are determined in any of the following ways:

[0099] (a) The default constraint strength values ​​of the core restricted area and buffer zone in the hierarchical constraint space model are calculated and determined according to a preset ratio;

[0100] (b) Based on the statistical analysis of the spatial compatibility index of historical application layout schemes, the high quantile value of its distribution (such as the top 70 percentile) is used as the setting.

[0101] (c) Directly set according to the quantitative requirements for coordination and compatibility of different functional zones in the regulations on marine space management.

[0102] The aforementioned preset warning values ​​are directly referenced from the project-level ecological impact limits specified in regional ecological planning or environmental impact assessment guidelines.

[0103] Furthermore, the process of dynamically adjusting the buffer boundary threshold involves each round of adjustment in the outer negotiation loop based on the wind turbine coordinate set after convergence of the inner multi-objective optimization algorithm, with a maximum number of adjustment rounds and a change threshold set as convergence control conditions. Specifically, if the calculated expansion or contraction distance is less than the preset minimum effective adjustment distance in multiple consecutive iteration cycles (e.g., 5), or if the spatial compatibility index and the cumulative ecological impact oscillate narrowly within a small neighborhood below their respective thresholds (e.g., fluctuation amplitude < 5%), then dynamic adjustment is stopped, the current constrained spatial model is locked, and the final optimization stage is entered.

[0104] If the boundary is detected to oscillate repeatedly between expansion and contraction, the termination determination mechanism is activated, and the buffer boundary is fixed based on the current optimal layout scheme, without further adjustment.

[0105] S7. Based on the updated constraint space model, perform conflict verification on the optimized wind turbine coordinate set. When a spatial conflict is detected, trigger a constraint relaxation strategy based on penalty cost to generate an optimized conflict-free layout scheme.

[0106] The conflict verification process checks whether each wind turbine foundation and its safety range geometrically overlap with any navigation restricted areas, buffer zones, or high-conflict-risk negotiation areas in the model. When no spatial conflict is detected, the optimized wind turbine coordinate set is directly output.

[0107] When a spatial conflict is detected, the trigger constraint relaxation strategy includes: performing a qualitative analysis of the detected spatial conflict. When the conflict verification step identifies that the location of a wind turbine or cable has encroached on a constraint area, the system immediately queries the constraint type corresponding to the conflict area, such as a waterway, anchorage, fishing activity area, or ecologically sensitive area, and retrieves the preset negotiation priority associated with that constraint type. This negotiation priority is a discrete level, for example, from 1 (highest, non-negotiable) to 5 (lowest, flexibly adjustable), which determines the degree of compromise of the constraint in conflict resolution.

[0108] For constraints marked as negotiable attributes, an additional economic penalty cost factor is introduced to replace the absolute prohibition constraint. Specifically, instead of forcibly prohibiting access to the area, a pre-set negotiated compensation cost (such as simulated fishery compensation or ecological restoration costs) is added to the objective function for wind turbines falling into the area. This makes the location mathematically feasible, but significantly reduces its economic score. The algorithm will only retain the location if the power generation revenue from that location is sufficient to cover this additional cost, thus generating a temporary relaxed constraint state.

[0109] For constraints marked as non-negotiable attributes, their strength should be maintained unchanged, such as the centerline of the main shipping channel, military restricted areas, or ecological red lines, ensuring that they exist as insurmountable rigid barriers under any circumstances.

[0110] Under the relaxed constraint state, the wind turbine position offset is recalculated. This calculation aims to minimize the offset distance and completely move the turbine out of all non-relaxed constraint regions, searching for a new acceptable location within the relaxed constraint regions. Once a location without hard conflicts is found, the system confirms the offset and corrects the original layout scheme to generate the final conflict-free layout scheme.

[0111] The recalculation of the wind turbine position offset is achieved through a constrained local optimization process, specifically: the decision variable is defined as the new position to be determined for the conflicting wind turbine, and the optimization objective is defined as minimizing the Euclidean distance offset between the new position and the original position.

[0112] The constraint condition is defined as follows: the new location must be within the area allowed by the temporary relaxed constraint state. Specifically, it cannot be located within any constraint area with a negotiation priority of non-negotiable; for constraint areas with a negotiation priority, the penalty cost of its intrusion must be lower than the preset relaxation tolerance limit.

[0113] The solution is obtained using local optimization algorithms such as pattern search or sequential quadratic programming. Starting from the original location, the algorithm iteratively searches for a feasible solution that minimizes the Euclidean distance offset while satisfying the constraints mentioned above. If no feasible solution is found within the preset search radius, the wind turbine is deemed unfit for deployment under the current relaxation conditions and must be removed from the layout scheme.

[0114] The method further includes: based on the final conflict-free layout scheme, simulating different sea conditions including parameters such as wave height, wind speed and current speed, and using path planning algorithms, such as the A* algorithm, to simulate the feasibility of operation and maintenance vessels traveling from the home port to various wind turbine locations in a dynamic marine environment, and generating a set of reachable paths for operation and maintenance vessels covering multiple typical meteorological scenarios.

[0115] Combining the constructability score calculated for each wind turbine location in the previous steps with the generated set of reachable paths for the maintenance vessel, a traveling salesman problem model with weighted nodes and dynamic edge costs is constructed. In this model, each wind turbine requiring maintenance is defined as a node. The node weight consists of two parts: first, a maintenance urgency score determined based on wind turbine operating status monitoring data (such as fault warning levels and regular maintenance cycles); and second, the reciprocal of the constructability score for that wind turbine location, reflecting the potential difficulty or time cost of performing maintenance work at that point. A higher node weight means that the wind turbine should be prioritized for access in the maintenance sequence.

[0116] An edge is defined between two wind turbine nodes, or between a wind turbine node and the maintenance home port. The cost of this edge is not a fixed distance, but is determined by the set of reachable paths for the maintenance vessel generated in the aforementioned simulation. For a given pair of nodes, the optimal or feasible navigation path under the corresponding sea state conditions is extracted from the set, and the estimated navigation time or fuel consumption of this path is used as the dynamic cost of the edge.

[0117] In this model, the goal is to find a path that starts from the home port, visits all wind turbine nodes at least once according to the maintenance task requirements, and finally returns to the home port, so that the sum of the total cost of all edges and the delay penalty implied by the node weights is minimized under the premise of satisfying actual constraints such as maintenance time window and ship capacity. Specifically, the model can be solved by heuristic or metaheuristic algorithms to generate an optimized wind turbine maintenance sequence.

[0118] It should be noted that the Traveling Salesman Problem model mentioned in this embodiment is a classic mathematical abstraction and application of operation and maintenance path planning, which belongs to the prior art and will not be described in detail here.

[0119] Based on the reachable path of the maintenance vessel and the optimized wind turbine maintenance sequence, a wind farm full lifecycle operation and maintenance strategy with time-series attributes is output.

[0120] The constraint relaxation strategy is triggered only after the multi-objective optimization algorithm converges, during the final conflict check of the output optimized wind turbine coordinate set. In the main optimization loop, the basic principle remains to fully adhere to the hierarchical constraint space model.

[0121] The method further includes: in a geographic information system platform, spatially registering and overlaying the dynamically coupled marine functional block map, which serves as the background layer, with the final conflict-free layout scheme, which serves as the foreground element, to form a comprehensive situation map that intuitively displays the relationship between the wind farm layout and the environmental constraints of the marine area where it is located.

[0122] To proactively identify and warn of potential coordination risks, the system traverses each wind turbine or cable path segment in the final conflict-free layout scheme, reads the spatial compatibility index corresponding to its location, and when the index is lower than a preset warning value, such as 0.5, the area is determined to be an area with high coordination difficulty, and is marked on the comprehensive situation map with a highlight or special symbol, and defined as a high-sensitivity coordination area.

[0123] Extract the boundary coordinates of the highly sensitive coordination zone, attach the specific constraint types involved in the zone, the characteristics of the conflict risk level, and the avoidance or relaxation strategies adopted by this method in the optimization process, and output the coordination zone data signal.

[0124] The embodiments of this invention introduce a constraint relaxation strategy based on conflict verification and a dynamic model adjustment mechanism, which has the ability to make flexible decisions and adaptive adjustments under complex constraints. It can intelligently resolve spatial conflicts in non-critical areas while ensuring that core interests are not infringed, thereby enhancing the feasibility and high-efficiency implementation of the final output solution in actual engineering.

[0125] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. A method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints, characterized in that, include: Acquire multi-source spatiotemporal marine resource data, integrate channel trajectory data, fishery activity cycles, ecological monitoring data, sea state forecast information, seabed geological data and regional planning information, and construct a dynamically coupled marine functional block map; Based on the dynamically coupled marine functional block map, the constraint correlation is quantified to generate a hierarchical constraint space model; Set initial wind turbine location parameters, calculate spatial compatibility index, constructability score and ecological cumulative impact based on the hierarchical constraint space model, and generate multi-dimensional performance evaluation results; Key performance indicators are extracted from the multidimensional performance evaluation results, and a comprehensive performance feature vector is generated through feature aggregation processing. The comprehensive performance feature vector is input into a multi-objective optimization algorithm to drive the main wind turbine location group and the auxiliary resource scheduling group to interact and iterate, and output an optimized wind turbine coordinate set and supporting cable path scheme. Based on the optimized wind turbine coordinate set, the buffer boundary threshold in the hierarchical constraint space model is dynamically adjusted to generate an updated constraint space model; Based on the updated constraint space model, conflict verification is performed on the optimized wind turbine coordinate set. When a spatial conflict is detected, a constraint relaxation strategy based on penalty cost is triggered to generate an optimized conflict-free layout scheme.

2. The method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints according to claim 1, characterized in that, The construction of the dynamically coupled marine functional block map includes: Identify the geographical boundaries of restricted navigation areas, buffer zones, and negotiation zones within the target sea area; By linking channel flow data with anchorage usage records, a channel-anchorage coupling influence coefficient is established. Match the fishery activity cycle with the wind farm maintenance time window, divide the target sea area into several grid units, calculate the probability that fishery activities and wind farm maintenance activities will occur simultaneously in each grid unit within a specific time period, and generate a spatiotemporal conflict probability matrix. By introducing a spatiotemporal weighting factor and integrating the geographic boundary, the channel-anchorage coupling influence coefficient, and the spatiotemporal conflict probability matrix, the initial constraint strength of each grid cell as it changes over time is obtained, thereby generating a dynamic coupled marine functional block map.

3. The method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints according to claim 1, characterized in that, The generation process of the hierarchical constraint space model includes: Based on the aforementioned dynamically coupled marine functional block map, the initial constraint strength of each grid cell as a function of time is extracted as the basic constraint field. The constraint correlation in the basic constraint field is analyzed and mapped into multiple sub-item quantitative constraint layers according to the constraint attribute dimension. The layers include at least a spatial geographic constraint layer, a functional activity coupling constraint layer, and a conflict risk constraint layer. The multiple sub-item quantification constraint layers are hierarchically aggregated according to preset weight coefficients under a unified spatiotemporal coordinate framework, and the target sea area is divided into multiple level regions with different development restriction intensities to form a hierarchical constraint space model. The hierarchical constraint space model includes an initial buffer boundary threshold, which is used to define the geographical range and constraint strength of the buffer.

4. The method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints according to claim 1, characterized in that, The generated multidimensional performance evaluation results include: Input the initial wind turbine location coordinates, the wind turbine standard performance curve, and the annual wind resource data of the target sea area obtained through historical data statistics. Calculate the annual equivalent full-load hours of the wind farm in the target sea area based on the wind turbine wake model. Based on the input wind turbine location coordinates and combiner station location, the optimal submarine cable connection topology with the goal of minimizing the total construction cost is determined by graph theory algorithm. Combined with seabed topography data and cable laying cost, the total cable length and total construction cost are calculated. Divided by the total installed capacity of the wind farm, the investment cost per kilowatt is obtained. A three-dimensional performance evaluation coordinate system is constructed based on the spatial compatibility index, constructability score, and cumulative ecological impact. This involves mapping the annual equivalent full-load hours and unit kilowatt investment cost to the three-dimensional performance evaluation coordinate system, and merging them to generate a multi-dimensional performance evaluation result that includes the spatial compatibility index, constructability score, cumulative ecological impact, annual equivalent full-load hours, and unit kilowatt investment cost.

5. The method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints according to claim 4, characterized in that, The generated comprehensive performance feature vector includes: Separate conflict risk level characteristics, economic characteristics, and ecological sensitivity characteristics from the multidimensional effectiveness assessment results; The conflict risk level characteristics are converted into spatial compatibility scalars using a linear weighting method. The economic and ecological sensitivity characteristics are mapped to a unified numerical range through normalization. The spatial compatibility scalar, normalized economic features, and normalized ecological sensitivity features are spliced ​​together according to the preset feature structure to generate a comprehensive performance feature vector.

6. The method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints according to claim 1, characterized in that, The interactive iteration between the main wind turbine location group and the auxiliary resource scheduling group includes: Decode the comprehensive performance feature vector into the main population evolution direction instruction; Generate the auxiliary group search space boundary based on the hierarchical constraint space model; The main wind turbine position group performs global position mutation based on the main group evolution direction command to generate the main group mutation result; The auxiliary resource scheduling group generates a cable path candidate set based on the mutation results of the main group within the boundary of the auxiliary group search space. By exchanging feasible solution fragments from the main population mutation results with obstacle avoidance strategies from the cable path candidate set, collaborative knowledge transfer is completed.

7. The method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints according to claim 1, characterized in that, The dynamically adjusted buffer boundary threshold includes: Monitor the changing trend of the spatial compatibility index within the iteration cycle to form a compatibility change trend; When the compatibility trend indicates that the spatial compatibility index is continuously higher than the compatibility threshold, the buffer boundary range in the hierarchical constraint spatial model is expanded to generate an expanded buffer boundary range. When the cumulative ecological impact is detected to exceed the preset warning value, the available area of ​​the negotiation zone in the hierarchical constraint space model is reduced to generate a reduced available area of ​​the negotiation zone. The expanded buffer boundary range and the shrunken negotiation area are written into the hierarchical constraint space model to generate the updated constraint space model.

8. The method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints according to claim 1, characterized in that, The constraint relaxation strategies triggered based on penalty costs include: Identify the constraint type corresponding to the spatial conflict and obtain the negotiation priority associated with the constraint type, wherein the negotiation priority includes negotiable attributes and non-negotiable attributes; For constraints marked as negotiable attributes, an additional economic penalty cost factor is introduced to replace the absolute prohibition constraint, generating a relaxed constraint state. For constraints marked as non-negotiable attributes, maintain their strength unchanged; The wind turbine position offset is recalculated under the relaxed constraint state to generate the optimized conflict-free layout scheme.

9. The method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints according to claim 1, characterized in that, The method further includes: Based on the aforementioned final conflict-free layout scheme, the reachable paths of maintenance vessels under different sea state conditions are simulated; The wind turbine maintenance sequence is optimized by combining the constructability score, and an optimized wind turbine maintenance sequence is generated. Based on the reachable path of the maintenance vessel and the optimized wind turbine maintenance sequence, a wind farm full lifecycle operation and maintenance strategy with time-series attributes is output.

10. The method for evaluating and optimizing the layout of offshore wind farms based on marine spatial resource constraints according to claim 1, characterized in that, The method further includes: The dynamically coupled sea area functional block map is overlaid with the final conflict-free layout scheme for display; Areas in the final conflict-free layout scheme with a spatial compatibility index lower than a preset warning value are marked as highly sensitive coordination areas. Extract the boundary coordinates of the highly sensitive coordination zone and output the coordination zone data signal.

Images