Offshore wind turbine site selection optimization method with wind power evaluation

Optimize offshore fan site selection through clustering and bat algorithms, and local searches are carried out in combination with KNN algorithms, which solves the problem of large amount of data calculation and low degree of optimization in offshore fan site selection, and achieves more efficient energy utilization and construction costs.

CN118982103AInactive Publication Date: 2024-11-19UNIV OF ELECTRONICS SCI & TECH OF CHINA ZHONGSHAN INST
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411012984.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-19
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

In the prior art, the data calculation amount in the site selection optimization process of offshore fan is large and the construction of different types of offshore fans is not considered, resulting in a low degree of site selection optimization.

Method used

The clustering algorithm determines the constructable sub-region, constructs the fitness function of the bat algorithm to optimize fan site selection, and performs local optimization searches in combination with the KNN algorithm to reduce the amount of data calculation and improve the degree of site selection optimization.

Benefits of technology

It improves energy utilization efficiency, reduces construction costs, enhances the accuracy and stability of site selection optimization, and is suitable for the construction of different types of offshore fans.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118982103A_ABST
    Figure CN118982103A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of fan site selection, in particular to an offshore fan site selection optimization method with wind power evaluation, which comprises the following steps: acquiring daily offshore wind speed data in a future target time period; constructing a constructable point data set, and determining a plurality of constructable sub-regions; determining the generation power and construction cost of each offshore wind turbine according to the input wind speed so as to construct a fitness function of a target bat algorithm, and performing site selection optimization on each constructable sub-region; boundary point random local optimization search is carried out on boundary points in the boundary point data set, and a boundary point random local optimization search result is determined according to the fitness of the constructable point data set and the number of times of boundary point random local optimization search; and performing KNN algorithm local optimization search on neighborhood points in the neighborhood point data set, and outputting a final offshore wind turbine construction site selection optimization result. According to the method, the construction conditions of different types of fans are considered, the energy utilization efficiency can be improved, the construction cost is reduced, and the site selection optimization degree is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of wind turbine site selection, and in particular to an offshore wind turbine site selection optimization method with wind force assessment. Background Art

[0002] In recent years, as the development of offshore wind energy resources has become saturated, offshore wind power has gradually developed towards the deep sea and offshore. The deep and offshore wind energy resources are rich, and the wind speed is higher and more stable than that of the offshore and onshore. But at the same time, deep and offshore wind farms are also more susceptible to the impact of marine climate, and deep and offshore wind farms have higher development costs and difficulties. Therefore, reducing the development investment cost of the deep sea and increasing the revenue of wind farms are issues that need to be urgently addressed. The micro-site selection of deep and offshore wind farms is of great significance to promoting the development of deep and offshore wind farms by arranging wind turbines within the planning range of the wind farm, reducing the impact of wake between wind turbines, improving the efficiency of the wind farm, and reducing the investment and construction costs of deep and offshore wind farms.

[0003] The key to the site selection of offshore wind turbines is how to arrange the wind turbines considering the influence of the wake effect to reduce the wake loss. The theoretical calculation of the wind speed loss after the wind passes through the blades is a non-convex nonlinear model. The commonly used model is the Jensen model. The calculation of the wind speed loss considering the influence of the wake between multiple wind turbines also needs to determine the degree of mutual influence between wind turbines under different wind speeds and wind directions. There are two types of wind turbine optimization arrangements: regular arrangement and discrete arrangement. Regular arrangement generally refers to the arrangement of wind turbines in rows and columns, and discrete arrangement means that there is no certain regularity in the arrangement of wind turbines in the construction area. The current regular arrangement scheme of wind turbines generally follows a regular arrangement scheme with the same distance between rows and columns. In the current wind farm site selection technology, the regular arrangement of positions and the technology of using wake to increase energy utilization are often not taken into account.

[0004] Chinese patent publication number CN117422168A discloses a deep-sea wind farm micro-site selection method and system based on deep reinforcement learning, including: based on the wind speed, wind direction and planning range data of the deep-sea wind farm, a wind turbine wake model, a wind turbine power model and a wind farm multi-wake superposition model are established to estimate the benefits of the deep-sea wind farm; deep reinforcement learning is introduced, PPO truncation is used to restrict policy updates, a proximal policy optimization algorithm is constructed and a training environment and a reward function are set; appropriate training parameters are set to train the layout of wind turbines in the deep-sea wind farm to obtain the optimized result of the wind turbine layout; the particle swarm algorithm is combined with the fuzzy C-means clustering algorithm and the Prim algorithm to jointly optimize the topological structure of the offshore substation and the collection system to obtain the optimized result of the topological structure of the offshore substation and the collection system.

[0005] It can be seen that the above invention has the following problems: the amount of data calculation is relatively large, and the construction conditions of different types of offshore wind turbines are not taken into consideration, and the degree of site optimization is not high. Summary of the invention

[0006] To this end, the present invention provides an offshore wind turbine site selection optimization method with wind force assessment to overcome the problems in the prior art of large data calculation amount, failure to consider the construction conditions of different types of offshore wind turbines, and low degree of site selection optimization.

[0007] To achieve the above object, the present invention provides an offshore wind turbine site selection optimization method with wind force assessment, comprising:

[0008] Step S1, obtaining daily offshore wind speed data within a future target time period based on historical offshore wind speed data;

[0009] Step S2, acquiring the constructible point data of the offshore wind turbine to construct a constructible point data set, and clustering the constructible point data set according to a preset clustering algorithm to determine a number of constructible sub-areas;

[0010] Step S3, calculating the input wind speed of the offshore wind turbines at each construction point, determining the power generation and construction cost of each offshore wind turbine according to the input wind speed to construct the fitness function of the target bat algorithm, and optimizing the site selection for each buildable sub-area to obtain the preliminary construction result of each offshore wind turbine buildable point;

[0011] Step S4, calculating the boundary points of each buildable sub-area according to a preset boundary point calculation method to obtain a boundary point data set;

[0012] Step S5, performing a random local optimization search for boundary points in the boundary point data set, and determining a random local optimization search result for boundary points according to the fitness of the buildable point data set and the number of random local optimization searches for boundary points;

[0013] Step S6, obtaining a neighborhood point data set of a random buildable point in the buildable point data set, performing a KNN algorithm local optimization search on the neighborhood points in the neighborhood point data set, and determining a KNN algorithm local optimization search result according to the fitness of the buildable point data set and the random local optimization search result of the boundary point;

[0014] Step S7: outputting the final optimization result of the offshore wind turbine construction site selection according to the search results of the KNN random local optimization search.

[0015] Furthermore, the step S1 comprises:

[0016] Step S11, constructing a target training data set and a target prediction data set based on historical offshore wind speed data;

[0017] Step S12, training the initial neural network model according to the target training data set to obtain a target neural network model;

[0018] Step S13, inputting the target prediction data set into the target neural network model to obtain daily offshore wind speed data within the future target time period.

[0019] Furthermore, the step S2 comprises:

[0020] Step S21, acquiring constructable point data of offshore wind turbines to construct a constructable point data set;

[0021] Step S22, clustering the buildable point data set according to a preset clustering algorithm to obtain a plurality of cluster sets; each cluster set includes at least one buildable point;

[0022] Step S23, determining a number of corresponding buildable sub-areas according to the positions of the number of buildable points in the number of cluster sets; each cluster set corresponds to a buildable sub-area.

[0023] Furthermore, the step S3 comprises:

[0024] Step S31, initializing the optimization parameters of the target bat algorithm, and constructing the particle update formula of the target bat algorithm and the particle coding of the target bat algorithm;

[0025] Step S32, determining the input wind speed of the offshore wind turbine at each construction point according to a preset wake model;

[0026] Step S33, determining the power generation and construction cost of each offshore wind turbine according to the input wind speed, and constructing a fitness function of the target bat algorithm;

[0027] Step S34, optimizing the site selection of each buildable sub-area according to the fitness function of the target bat algorithm, so as to obtain a preliminary construction result of each buildable point of the offshore wind turbine.

[0028] Furthermore, the step S5 comprises:

[0029] Step S51, calculating the fitness corresponding to the buildable point data set according to a preset fitness calculation method to obtain an initial fitness;

[0030] Step S52, randomly selecting any two boundary points in the boundary point data set, exchanging the offshore wind turbine types corresponding to the two boundary points, and recalculating the fitness of the buildable point data set according to a preset fitness calculation method to obtain a first candidate fitness.

[0031] Step S53, if the first candidate fitness is greater than the initial fitness, retain the exchanged result, otherwise, discard the exchanged result;

[0032] Step S54, randomly selecting any boundary point in the boundary point data set, randomly changing the offshore wind turbine type corresponding to the boundary point, and recalculating the fitness of the buildable point data set according to a preset fitness calculation method to obtain a second candidate fitness;

[0033] Step S55, if the second candidate fitness is greater than the initial fitness, retain the exchanged result, otherwise, discard the exchanged result;

[0034] Step S56, determining whether the number of local optimization searches for boundary points is greater than or equal to a first preset number, if it is greater than or equal to the first preset number, executing S6, otherwise, executing step S52.

[0035] Furthermore, the step S6 comprises:

[0036] Step S61, randomly selecting a point on the buildable point data set as a random buildable point, and performing a KNN algorithm local optimization search on the random buildable point to obtain a neighborhood point data set; the neighborhood point data set includes K neighborhood points corresponding to the random buildable point, and each neighborhood point has a corresponding buildable point;

[0037] Step S62, randomly selecting any two neighboring points in the neighborhood point data set, and if the offshore wind turbine type corresponding to the neighborhood points is not the first preset wind turbine type, modifying it to the first preset wind turbine type, and recalculating the fitness of the buildable point data set according to a preset fitness calculation method to obtain a third candidate fitness;

[0038] Step S63, if the third candidate fitness is greater than the initial fitness, retain the exchanged result, otherwise, discard the exchanged result;

[0039] Step S64, randomly selecting any neighborhood point in the neighborhood point data set, randomly changing the offshore wind turbine type corresponding to the neighborhood point, and recalculating the fitness of the buildable point data set according to a preset fitness calculation method to obtain a fourth candidate fitness;

[0040] Step S65: if the fourth candidate fitness is greater than the initial fitness, the exchanged result is retained; otherwise, the exchanged result is discarded.

[0041] Step S66, calculating the innermost point and the outermost point of the neighborhood point data set; the innermost point is the point closest to the origin, and the outermost point is the point farthest from the origin;

[0042] Step S67, taking the innermost point as the center, executing the KNN algorithm local optimization search, obtaining its nearest K-1 neighboring points to obtain the first neighborhood set, and executing the steps S51 to S55;

[0043] Step S68, taking the outermost point as the center, performing a local optimization search using the KNN algorithm to obtain its nearest K-1 neighboring points to obtain a second neighborhood set, and performing steps S62 to S65;

[0044] Step S69, repeat the above steps until K-1=1, and execute step S7.

[0045] Furthermore, in step S31, the particle update formula of the target bat algorithm is:

[0046] f i =f min +(f max -f min )×β;

[0047]

[0048] Among them, β is a random value in [0,1], f i is the sound wave frequency of the ith particle, f i Between the interval [f min ,f max ]middle, is the particle velocity of the ith particle at time t, is the particle velocity of the ith particle at time t+1, is the position of the ith particle at time t, is the historical optimal position of all particles at time t, is the position of the ith particle at time t+1.

[0049] Furthermore, in step S32, the expression formula of the preset wake model is:

[0050]

[0051] Where r is the radius of the offshore wind turbine; α is the attenuation constant; W0 is the offshore wind speed; γ is the inference coefficient; W is the input wind speed at a distance D downstream of the offshore wind turbine in the direction of the upstream wind speed.

[0052] Furthermore, the step S33 includes:

[0053] Step S331, determining the power generation power of each offshore wind turbine according to a preset input wind speed and power generation comparison table;

[0054] Step S332, determining the construction cost of each offshore wind turbine according to a preset comparison table of input wind speed and construction cost;

[0055] Step S333, constructing a fitness function of the target bat algorithm according to the power generation and construction cost of each offshore wind turbine.

[0056] Furthermore, the preset fitness calculation method is to calculate the fitness according to the following formula:

[0057]

[0058] Where N is the total number of buildable points, j = 1, 2, ..., N; P j is the power generation capacity of the offshore wind turbine at the jth constructible point; M j is the construction cost of the offshore wind turbine at the jth buildable point.

[0059] Compared with the prior art, the beneficial effect of the present invention is that the present invention can improve the efficiency of subsequent data processing by clustering the buildable point data set to determine several buildable sub-areas. The power generation power and construction cost of each offshore wind turbine are determined according to the input wind speed to construct the fitness function of the target bat algorithm, and the site selection is optimized for each buildable sub-area, which can improve the efficiency of energy utilization while reducing the construction cost to optimize the site selection. By performing a random local optimization search of the boundary points based on the boundary points of the buildable sub-area, the site selection results are further optimized, and the degree of site selection optimization is improved. The KNN algorithm is performed on the neighborhood points of a random buildable point in the buildable point data set to optimize the site selection results again.

[0060] Furthermore, the present invention trains a target neural network model based on historical offshore wind speed data to predict daily offshore wind speeds within a future preset time period, which can improve the accuracy of the prediction and thus improve the accuracy of subsequent calculations of input wind speeds for offshore wind turbines.

[0061] Furthermore, the present invention determines the input wind speed of the offshore wind turbine at each construction point according to the preset wake model, introduces wind force evaluation, can reduce the influence of the wake effect, and improve stability. The site selection of each buildable sub-area is optimized through the fitness function of the target bat algorithm, which can reduce the amount of data calculation and improve the optimization ability of site selection.

[0062] Furthermore, the present invention changes the offshore wind turbine type corresponding to the boundary point and recalculates the fitness to perform local optimization search of the boundary point, which can provide site selection optimization for the construction of different types of offshore wind turbines and improve the degree of optimization of site selection.

[0063] Furthermore, the present invention can realize site selection optimization for the construction of different types of offshore wind turbines by performing KNN algorithm local optimization search on randomly selected buildable points on the buildable point data set, and recalculating the fitness by changing the offshore wind turbine type corresponding to the buildable point. And by calculating the innermost point and outermost point of the neighborhood point data set, performing KNN algorithm local optimization search, the optimization effect and degree of site selection optimization can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 A flow chart of the offshore wind turbine site selection optimization method with wind force assessment according to the present invention;

[0065] Figure 2 Schematic diagram of the process of step S1 of the embodiment of the present invention;

[0066] Figure 3 Schematic diagram of the process of step S2 of the embodiment of the present invention;

[0067] Figure 4 Schematic diagram of the process of step S3 of the embodiment of the present invention;

[0068] Figure 5 Schematic diagram of the process of step S5 of the embodiment of the present invention;

[0069] Figure 6 Schematic diagram of the process of step S6 of the embodiment of the present invention;

[0070] Figure 7 Schematic diagram of the process of step S33 of the embodiment of the present invention. DETAILED DESCRIPTION

[0071] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0072] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0073] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0074] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0075] See also Figure 1 As shown, it is a flow chart of the offshore wind turbine site selection optimization method with wind force assessment of the present invention.

[0076] The present invention provides an offshore wind turbine site selection optimization method with wind force assessment, comprising:

[0077] Step S1, obtaining daily offshore wind speed data within a future target time period based on historical offshore wind speed data;

[0078] In actual application scenarios, actual implementers can set the target time period according to actual conditions. Preferably, the target time period is set to 1 year.

[0079] See also Figure 2 As shown, it is a flow chart of step S1 of an embodiment of the present invention. Specifically, step S1 includes:

[0080] Step S11, constructing a target training data set and a target prediction data set based on historical offshore wind speed data;

[0081] Step S12, training the initial neural network model according to the target training data set to obtain a target neural network model;

[0082] It should be noted that those skilled in the art are aware that any prediction model in the prior art falls within the protection scope of the present invention and will not be described in detail herein.

[0083] In a specific embodiment, the initial neural network model can be a long short-term memory neural network (LSTM) model, and the set LSTM neural network hyperparameters are: neural network sequence length (Sequence): 15, number of neural network hidden layers (Layer): 3, neural network learning rate (LearningRate): 0.5.

[0084] Step S13, inputting the target prediction data set into the target neural network model to obtain daily offshore wind speed data within the future target time period.

[0085] The present invention trains a target neural network model based on historical offshore wind speed data to predict daily offshore wind speeds within a future preset time period, which can improve the accuracy of the prediction and thus improve the accuracy of subsequent calculations of input wind speeds for offshore wind turbines.

[0086] Step S2, acquiring the constructible point data of the offshore wind turbine to construct a constructible point data set, and clustering the constructible point data set according to a preset clustering algorithm to determine a number of constructible sub-areas;

[0087] In implementation, the constructible point data set U may be a two-dimensional point set, for example: U = {(x0, y0), (x1, y1), (x2, y2), ..., (x j ,y j ),…,(x N ,y N )}, where N is the total number of buildable points, j = 1, 2, ..., N; (x j ,y j ) is the coordinate of the jth offshore wind turbine buildable point.

[0088] It should be noted that those skilled in the art know that any clustering algorithm in the prior art that can cluster data with the same or similar characteristics into one category, such as the K-Means clustering algorithm, the K-Medoids clustering algorithm, etc., falls within the protection scope of the present invention and will not be described in detail here.

[0089] See also Figure 3 As shown, it is a flow chart of step S2 of an embodiment of the present invention. Specifically, step S2 includes:

[0090] Step S21, acquiring constructable point data of offshore wind turbines to construct a constructable point data set;

[0091] Step S22, clustering the buildable point data set according to a preset clustering algorithm to obtain a plurality of cluster sets; each cluster set includes at least one buildable point;

[0092] Step S23, determining a number of corresponding buildable sub-areas according to the positions of the number of buildable points in the number of cluster sets; each cluster set corresponds to a buildable sub-area.

[0093] In implementation, the area enclosed by the maximum longitude, minimum longitude, maximum latitude, and minimum latitude in the longitude and latitude coordinates of several buildable points in each cluster set can be used as the buildable sub-area corresponding to each cluster set.

[0094] Step S3, calculating the input wind speed of the offshore wind turbines at each construction point, determining the power generation and construction cost of each offshore wind turbine according to the input wind speed to construct the fitness function of the target bat algorithm, and optimizing the site selection for each buildable sub-area to obtain the preliminary construction result of each offshore wind turbine buildable point;

[0095] See also Figure 4 As shown, it is a flow chart of step S3 of an embodiment of the present invention. Specifically, step S3 includes:

[0096] Step S31, initializing the optimization parameters of the target bat algorithm, and constructing the particle update formula of the target bat algorithm and the particle coding of the target bat algorithm;

[0097] In the implementation, the optimization parameters of the target bat algorithm are the number of particles num, the particle velocity v, the inertia weight w, and the learning factors c1 and c2.

[0098] Specifically, the particle update formula of the target bat algorithm is:

[0099] f i =f min +(f max -f min )×β;

[0100]

[0101] Among them, β is a random value in [0,1], f i is the sound wave frequency of the ith particle, f i Between the interval [f min , f max ]middle, is the particle velocity of the ith particle at time t, is the particle velocity of the ith particle at time t+1, is the position of the ith particle at time t, is the historical optimal position of all particles at time t, is the position of the ith particle at time t+1.

[0102] A particle encoding of the target bat algorithm is:

[0103] particle=((x0,y0,T0), (x1,y1,T1 ) , (x2, y2, T2), ..., (x n ,y n , T n )),

[0104] Where (x0, y0, T0) indicates that an offshore wind turbine of type T0 is installed at the buildable point)x0, y0(. T0 indicates that no offshore wind turbine is installed, T1 indicates an offshore wind turbine of type 1, T2 indicates an offshore wind turbine of type 2, and so on. n is the total number of buildable points in the buildable sub-area. Different types of offshore wind turbines have different electrical parameters and different construction costs.

[0105] Step S32, determining the input wind speed of the offshore wind turbine at each construction point according to a preset wake model;

[0106] Specifically, the expression formula of the preset wake model is:

[0107]

[0108] Where r is the radius of the offshore wind turbine; α is the attenuation constant; W0 is the offshore wind speed; γ is the inference coefficient; W is the input wind speed at a distance D downstream of the offshore wind turbine in the direction of the upstream wind speed.

[0109] Step S33, determining the power generation and construction cost of each offshore wind turbine according to the input wind speed, and constructing a fitness function of the target bat algorithm;

[0110] See also Figure 7 As shown, it is a schematic diagram of the process of step S33 of an embodiment of the present invention. Specifically, step S33 includes:

[0111] Step S331, determining the power generation power of each offshore wind turbine according to a preset input wind speed and power generation comparison table;

[0112] Step S332, determining the construction cost of each offshore wind turbine according to a preset comparison table of input wind speed and construction cost;

[0113] Step S333, constructing a fitness function of the target bat algorithm according to the power generation and construction cost of each offshore wind turbine.

[0114] In actual application scenarios, actual implementers can set up a preset input wind speed and power generation comparison table and a preset input wind speed and construction cost comparison table based on the expert system according to actual conditions, or comprehensively determine the power generation and construction cost corresponding to the different input wind speeds of each offshore wind turbine based on historical data.

[0115] Step S34, optimizing the site selection of each buildable sub-area according to the fitness function of the target bat algorithm, so as to obtain a preliminary construction result of each buildable point of the offshore wind turbine.

[0116] Specifically, the preliminary construction results are expressed as follows:

[0117] ((x0, y0, T0), (x1, y1, T1), (x2, y2, T2),..., (x, n y n , T n )).

[0118] The present invention determines the input wind speed of the offshore wind turbine at each construction point according to the preset wake model, introduces wind force evaluation, can reduce the influence of the wake effect and improve stability. The site selection of each buildable sub-area is optimized through the fitness function of the target bat algorithm, which can reduce the amount of data calculation and improve the optimization ability of site selection.

[0119] Step S4, calculating the boundary points of each buildable sub-area according to a preset boundary point calculation method to obtain a boundary point data set;

[0120] Specifically, the preset boundary point calculation method is determined by the following conditions:

[0121] ||Point-Center m ||-||Point-Center h ||<θ,

[0122] m≠h,m≤k,h≤k,

[0123] Point is any point in the constructible point data set; Center m Represents the buildable sub-area A obtained by clustering m The cluster center of h Represents the buildable sub-area A obtained by clustering h The cluster center of ; k is the number of buildable sub-regions; θ is the boundary region constant, generally, θ is set to 40-60.

[0124] Step S5, performing a random local optimization search for boundary points in the boundary point data set, and determining a random local optimization search result for boundary points according to the fitness of the buildable point data set and the number of random local optimization searches for boundary points;

[0125] See also Figure 5 As shown, it is a flow chart of step S5 of an embodiment of the present invention. Specifically, step S5 includes:

[0126] Step S51, calculating the fitness corresponding to the buildable point data set according to a preset fitness calculation method to obtain an initial fitness;

[0127] Specifically, the preset fitness calculation method is to calculate the fitness according to the following formula:

[0128]

[0129] Among them, P j is the power generation capacity of the offshore wind turbine at the jth constructible point; M j is the construction cost of the offshore wind turbine at the jth buildable point.

[0130] Step S52, randomly selecting any two boundary points in the boundary point data set, exchanging the offshore wind turbine types corresponding to the two boundary points, and recalculating the fitness of the buildable point data set according to a preset fitness calculation method to obtain a first candidate fitness.

[0131] Step S53: if the first candidate fitness is greater than the initial fitness, retain the exchanged result; otherwise, discard the exchanged result;

[0132] In implementation, if the first candidate fitness is greater than the initial fitness, it indicates that the site optimization scheme corresponding to the first candidate fitness is more reasonable and has a better optimization effect, so the result after the exchange is retained; otherwise, if the site optimization scheme before the exchange is more reasonable and has a better optimization effect, the result after the exchange is discarded.

[0133] Step S54, randomly selecting any boundary point in the boundary point data set, randomly changing the offshore wind turbine type corresponding to the boundary point, and recalculating the fitness of the buildable point data set according to a preset fitness calculation method to obtain a second candidate fitness;

[0134] Step S55, if the second candidate fitness is greater than the initial fitness, retain the exchanged result, otherwise, discard the exchanged result;

[0135] In implementation, the second candidate fitness is greater than the initial fitness, indicating that the site optimization plan corresponding to the second candidate fitness is more reasonable and has a better optimization effect. Therefore, the result after the exchange is retained; otherwise, the site optimization plan before the exchange is more reasonable and has a better optimization effect, and the result after the exchange is discarded.

[0136] Step S56, determining whether the number of local optimization searches for boundary points is greater than or equal to a first preset number, if it is greater than or equal to the first preset number, executing S6, otherwise, executing step S52.

[0137] In implementation, the corresponding first preset number may be determined based on the number of buildable points in the buildable sub-area. Preferably, the first preset number is set to 1 / 5 to 1 / 4 of the number of buildable points in the buildable sub-area.

[0138] The present invention changes the offshore wind turbine type corresponding to the boundary point and recalculates the fitness to perform local optimization search of the boundary point, which can provide site selection optimization for the construction of different types of offshore wind turbines and improve the degree of optimization of site selection.

[0139] Step S6, obtaining a neighborhood point data set of a random buildable point in the buildable point data set, performing a KNN algorithm local optimization search on the neighborhood points in the neighborhood point data set, and determining a KNN algorithm local optimization search result according to the fitness of the buildable point data set and the random local optimization search result of the boundary point;

[0140] See also Figure 6 As shown, it is a schematic diagram of the process of step S6 of an embodiment of the present invention. Specifically, step S6 includes:

[0141] Step S61, randomly selecting a point on the buildable point data set as a random buildable point, and performing a KNN algorithm local optimization search on the random buildable point to obtain a neighborhood point data set; the neighborhood point data set includes K neighborhood points corresponding to the random buildable point, and each neighborhood point has a corresponding buildable point;

[0142] Step S62, randomly selecting any two neighboring points in the neighborhood point data set, and if the offshore wind turbine type corresponding to the neighborhood points is not the first preset wind turbine type, modifying it to the first preset wind turbine type, and recalculating the fitness of the buildable point data set according to a preset fitness calculation method to obtain a third candidate fitness;

[0143] In implementation, the first preset wind turbine type is T0 (T0 means no offshore wind turbine is installed), that is, the offshore wind turbines of the constructible points corresponding to any two neighboring points in the neighborhood point data set are randomly removed.

[0144] Step S63, if the third candidate fitness is greater than the initial fitness, retain the exchanged result, otherwise, discard the exchanged result;

[0145] In implementation, the third candidate fitness is greater than the initial fitness, indicating that the site optimization plan corresponding to the third candidate fitness is more reasonable and has a better optimization effect. Therefore, the result after the exchange is retained; otherwise, the site optimization plan before the exchange is more reasonable and has a better optimization effect, and the result after the exchange is discarded.

[0146] Step S64, randomly selecting any neighborhood point in the neighborhood point data set, randomly changing the offshore wind turbine type corresponding to the neighborhood point, and recalculating the fitness of the buildable point data set according to a preset fitness calculation method to obtain a fourth candidate fitness;

[0147] Step S65: if the fourth candidate fitness is greater than the initial fitness, the exchanged result is retained; otherwise, the exchanged result is discarded.

[0148] In implementation, the fourth candidate fitness is greater than the initial fitness, indicating that the site optimization plan corresponding to the fourth candidate fitness is more reasonable and has a better optimization effect. Therefore, the result after the exchange is retained; otherwise, the site optimization plan before the exchange is more reasonable and has a better optimization effect, and the result after the exchange is discarded.

[0149] Step S66, calculating the innermost point and the outermost point of the neighborhood point data set; the innermost point is the point closest to the origin, and the outermost point is the point farthest from the origin;

[0150] The origin is set to the position corresponding to the random buildable point corresponding to the neighborhood point dataset.

[0151] Step S67, taking the innermost point as the center, executing the KNN algorithm local optimization search, obtaining its nearest K-1 neighboring points to obtain the first neighborhood set, and executing the steps S51 to S55;

[0152] Step S68, taking the outermost point as the center, performing a local optimization search using the KNN algorithm to obtain its nearest K-1 neighboring points to obtain a second neighborhood set, and performing steps S62 to S65;

[0153] Step S69, repeat the above steps until K-1=1, and execute step S7.

[0154] The present invention can realize site selection optimization for the construction of different types of offshore wind turbines by performing KNN algorithm local optimization search on randomly selected buildable points on the buildable point data set, and recalculating the fitness by changing the offshore wind turbine type corresponding to the buildable point. And by calculating the innermost point and outermost point of the neighborhood point data set, the KNN algorithm local optimization search can improve the optimization effect and degree of site selection optimization.

[0155] Step S7: outputting the final optimization result of the offshore wind turbine construction site selection according to the search results of the KNN random local optimization search.

[0156] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. An offshore wind turbine site selection optimization method with wind power assessment, It is characterized in that include: Step S1, obtaining daily offshore wind speed data within a future target time period based on historical offshore wind speed data; Step S2, acquiring the constructible point data of the offshore wind turbine to construct a constructible point data set, and clustering the constructible point data set according to a preset clustering algorithm to determine a number of constructible sub-areas; Step S3, calculating the input wind speed of the offshore wind turbines at each construction point, determining the power generation and construction cost of each offshore wind turbine according to the input wind speed to construct the fitness function of the target bat algorithm, and optimizing the site selection for each buildable sub-area to obtain the preliminary construction result of each offshore wind turbine buildable point; Step S4, calculating the boundary points of each buildable sub-area according to a preset boundary point calculation method to obtain a boundary point data set; Step S5, performing a random local optimization search for boundary points in the boundary point data set, and determining a random local optimization search result for boundary points according to the fitness of the buildable point data set and the number of random local optimization searches for boundary points; Step S6, obtaining a neighborhood point data set of a random buildable point in the buildable point data set, performing a KNN algorithm local optimization search on the neighborhood points in the neighborhood point data set, and determining a KNN algorithm local optimization search result according to the fitness of the buildable point data set and the random local optimization search result of the boundary point; Step S7: outputting the final optimization result of the offshore wind turbine construction site selection according to the search results of the KNN random local optimization search.

2. The offshore wind turbine site selection optimization method with wind force assessment according to claim 1, It is characterized in that The step S1 comprises: Step S11, constructing a target training data set and a target prediction data set based on historical offshore wind speed data; Step S12, training the initial neural network model according to the target training data set to obtain a target neural network model; Step S13, inputting the target prediction data set into the target neural network model to obtain daily offshore wind speed data within the future target time period.

3. The offshore wind turbine site selection optimization method with wind force assessment according to claim 2, It is characterized in that The step S2 comprises: Step S21, acquiring constructable point data of offshore wind turbines to construct a constructable point data set; Step S22, clustering the buildable point data set according to a preset clustering algorithm to obtain a plurality of cluster sets; each cluster set includes at least one buildable point; Step S23, determining a number of corresponding buildable sub-areas according to the positions of the number of buildable points in the number of cluster sets; each cluster set corresponds to a buildable sub-area.

4. The offshore wind turbine site selection optimization method with wind force assessment according to claim 3, It is characterized in that The step S3 comprises: Step S31, initializing the optimization parameters of the target bat algorithm, and constructing the particle update formula of the target bat algorithm and the particle coding of the target bat algorithm; Step S32, determining the input wind speed of the offshore wind turbine at each construction point according to a preset wake model; Step S33, determining the power generation and construction cost of each offshore wind turbine according to the input wind speed, and constructing a fitness function of the target bat algorithm; Step S34, optimizing the site selection of each buildable sub-area according to the fitness function of the target bat algorithm, so as to obtain a preliminary construction result of each buildable point of the offshore wind turbine.

5. The offshore wind turbine site selection optimization method with wind force assessment according to claim 4, It is characterized in that The step S5 comprises: Step S51, calculating the fitness corresponding to the buildable point data set according to a preset fitness calculation method to obtain an initial fitness; Step S52, randomly selecting any two boundary points in the boundary point data set, exchanging the offshore wind turbine types corresponding to the two boundary points, and recalculating the fitness of the buildable point data set according to a preset fitness calculation method to obtain a first candidate fitness. Step S53, if the first candidate fitness is greater than the initial fitness, retain the exchanged result, otherwise, discard the exchanged result; Step S54, randomly selecting any boundary point in the boundary point data set, randomly changing the offshore wind turbine type corresponding to the boundary point, and recalculating the fitness of the buildable point data set according to a preset fitness calculation method to obtain a second candidate fitness; Step S55, if the second candidate fitness is greater than the initial fitness, retain the exchanged result, otherwise, discard the exchanged result; Step S56, determining whether the number of local optimization searches for boundary points is greater than or equal to a first preset number, if it is greater than or equal to the first preset number, executing S6, otherwise, executing step S52.

6. The offshore wind turbine site selection optimization method with wind force assessment according to claim 5, It is characterized in that The step S6 comprises: Step S61, randomly selecting a point on the buildable point data set as a random buildable point, and performing a KNN algorithm local optimization search on the random buildable point to obtain a neighborhood point data set; the neighborhood point data set includes K neighborhood points corresponding to the random buildable point, and each neighborhood point has a corresponding buildable point; Step S62, randomly selecting any two neighboring points in the neighborhood point data set, and if the offshore wind turbine type corresponding to the neighborhood points is not the first preset wind turbine type, modifying it to the first preset wind turbine type, and recalculating the fitness of the buildable point data set according to a preset fitness calculation method to obtain a third candidate fitness; Step S63, if the third candidate fitness is greater than the initial fitness, retain the exchanged result, otherwise, discard the exchanged result; Step S64, randomly selecting any neighborhood point in the neighborhood point data set, randomly changing the offshore wind turbine type corresponding to the neighborhood point, and recalculating the fitness of the buildable point data set according to a preset fitness calculation method to obtain a fourth candidate fitness; Step S65: if the fourth candidate fitness is greater than the initial fitness, retain the exchanged result; otherwise, discard the exchanged result. Step S66, calculating the innermost point and the outermost point of the neighborhood point data set; the innermost point is the point closest to the origin, and the outermost point is the point farthest from the origin; Step S67, taking the innermost point as the center, executing the KNN algorithm local optimization search, obtaining its nearest K-1 neighboring points to obtain the first neighborhood set, and executing the steps S51 to S55; Step S68, taking the outermost point as the center, performing a local optimization search using the KNN algorithm to obtain its nearest K-1 neighboring points to obtain a second neighborhood set, and performing steps S62 to S65; Step S69, repeat the above steps until K-1=1, and execute step S7.

7. The offshore wind turbine site selection optimization method with wind force assessment according to claim 6, It is characterized in that In step S31, the particle update formula of the target bat algorithm is: fi=fmin+(fmax-fmin)×β; Among them, β is a random value in [0,1], f i is the sound wave frequency of the ith particle, f i Between the interval [f min ,f max ]middle, is the particle velocity of the ith particle at time t, is the particle velocity of the ith particle at time t+1, is the position of the ith particle at time t, is the historical optimal position of all particles at time t, is the position of the ith particle at time t+1.

8. The offshore wind turbine site selection optimization method with wind force assessment according to claim 7, It is characterized in that In step S32, the expression formula of the preset wake model is: Where r is the radius of the offshore wind turbine; α is the attenuation constant; W 0 is the offshore wind speed; γ is the inference coefficient; W is the input wind speed at a distance D downstream of the offshore wind turbine along the direction of the offshore wind speed.

9. The offshore wind turbine site selection optimization method with wind force assessment according to claim 8, It is characterized in that The step S33 comprises: Step S331, determining the power generation power of each offshore wind turbine according to a preset input wind speed and power generation comparison table; Step S332, determining the construction cost of each offshore wind turbine according to a preset comparison table of input wind speed and construction cost; Step S333, constructing a fitness function of the target bat algorithm according to the power generation and construction cost of each offshore wind turbine.

10. The offshore wind turbine site selection optimization method with wind force assessment according to claim 9, It is characterized in that The preset fitness calculation method is to calculate the fitness according to the following formula: Where N is the total number of buildable points, j = 1, 2, ..., N; P j is the power generation capacity of the offshore wind turbine at the jth constructible point; M j is the construction cost of the offshore wind turbine at the jth buildable point.