An Optimization Method and System for Wind Farm Layout Based on Adaptive Differential Evolution

Abstract

The present invention discloses a method and system for optimizing the layout of a wind farm based on adaptive differential evolution. The method includes the steps of: S1, obtaining wind energy data of the wind farm, wind farm and fan parameters, and establishing an optimization model for the layout of the wind farm; S2, randomly generating wind farm layout schemes to form an initial population; S3, randomly dividing the initial population into several sub-populations and performing sub-population search to obtain mutation vectors; S4, obtaining candidate layout solutions by performing local differences on the mutation vectors, and then performing an adaptive replacement operation on the target vectors to obtain trial vectors; S5, performing a repair and replacement operation on the trial variables, and selecting between the target vectors and the repaired trial vectors to obtain suitable individuals; S6, continuing to execute the process of steps S3-S5 until the set maximum evaluation times are reached, and finally obtaining the optimal wind farm layout solution. The present invention has the advantages of reducing the wake effect and maximizing the expected power generation, etc.

Description

Technical Field

[0001] The present invention mainly relates to the field of wind farm layout optimization, and particularly relates to a method and system for wind farm layout optimization based on adaptive differential evolution. Background Art

[0002] With the world's increasing attention to new energy and sustainable energy, wind power generation has gradually become the main force of renewable energy power generation equipment. Wind turbines adopt flexible structures and are extremely sensitive to wind loads. During the process of converting wind energy into electrical energy by wind turbines, the wake of the upstream wind turbines will cause a loss in the wind speed of the downstream wind turbines, resulting in a reduction in the energy obtained by the wind turbine units. Traditional layout methods require a large amount of manual comparison, with heavy workloads, and usually cannot obtain an optimal layout scheme. Summary of the Invention

[0003] In view of the technical problems existing in the prior art, the present invention provides a method and system for wind farm layout optimization based on adaptive differential evolution that reduces the wake effect and maximizes the expected power generation.

[0004] To solve the above technical problems, the technical solutions proposed by the present invention are as follows:

[0005] A method for wind farm layout optimization based on adaptive differential evolution, comprising the steps of:

[0006] S1. Obtain the wind energy data of the wind farm, the parameters of the wind farm and the wind turbines, and establish a wind farm layout optimization model;

[0007] S2. Randomly generate a preset number of wind farm layout schemes that meet the constraint conditions to form an initial population;

[0008] S3. Randomly divide the initial population into several subpopulations, and then perform subpopulation searches to obtain mutant vectors;

[0009] S4. Perform local differences on the mutant vectors to obtain candidate layout solutions, and based on the candidate layout solutions, perform an adaptive replacement operation on the target vectors to obtain trial vectors;

[0010] S5. Perform a repair and replacement operation on the trial variables to obtain a repaired trial vector, and then select between the target vector and the repaired trial vector to obtain a suitable individual;

[0011] S6. Continue to perform the processes of steps S3 - S5 on the population until the set maximum number of evaluations is reached, and finally obtain the optimal wind farm layout solution.

[0012] Preferably, the specific process of step S1 is as follows:

[0013] S11. Obtain the wind energy data of the wind farm, the parameters of the wind farm and the wind turbines, and perform preprocessing to establish a wind speed probability distribution model, a wake model, and a wind turbine power curve model;

[0014] S12. Based on the wind speed probability distribution model, the wake model, and the wind turbine power curve model, establish the power output expectation function of each wind turbine, and then obtain the wind farm optimization model.

[0015] Preferably, in step S11, the two-parameter Weibull distribution is used to represent the wind speed probability distribution model; the Jensen model is used to establish the wake model; the Gaussian function is used to represent the wind turbine power curve model, and the formulas are as follows:

[0016]

[0017] where is the power of the wind turbine i ; is the rated power of the wind turbine; is the wind speed, , and are the cut-in wind speed, the rated wind speed, and the cut-out wind speed of the wind turbine respectively, a , b and c are the parameters obtained by fitting the wind turbine power curve data with the Gaussian function respectively.

[0018] Preferably, in step S12, the formula of the power output expectation function is as follows:

[0019]

[0020] where is the power output expectation of the wind turbine i ; is the rated power of the wind turbine, is the scaling parameter considering the wake effect, is the shape parameter, is the probability of each interval of the wind direction, is the median of each interval of the divided wind direction, n = 1, …, g ; is the median of each interval of the divided wind speed; divide , into s intervals on average, and each interval is , , j = 1, …, s ; g is the number of intervals obtained by dividing the wind direction angle on average;

[0021] Among them, the variables of the wind farm layout optimization model are the two-dimensional coordinates of each wind turbine, expressed as x i1 , x i2 , i = 1, …, N ; N is the number of wind turbines in the wind farm;

[0022] The objective function of the wind farm layout optimization model is expressed as the following formula:

[0023]

[0024] The boundary constraint condition of the wind farm layout optimization model is: the wind turbines are within the preset range;

[0025] The distance constraint condition of the wind farm layout optimization model is: the safety distance between two wind turbines is m times the rotor diameter of the wind turbine.

[0026] Preferably, in step S4, the specific process of obtaining the candidate layout solution by performing local difference on the mutation vector is as follows: re-encode the mutation vector, take a wind farm layout plan as a population Q, where an individual is the coordinate of a wind turbine, perform local difference on the population Q, and then re-encode the differentiated offspring population once to obtain the candidate layout solution.

[0027] Preferably, in step S4, the specific process of obtaining the trial vector by performing an adaptive replacement operation on the target vector based on the candidate layout solution is as follows:

[0028] When the number of evaluations is less than the learning period of the adaptive replacement operation, use the coordinates of the first wind turbine in the candidate layout solution to randomly replace one coordinate in the target vector, and the index of the replaced coordinate is r , to obtain the trial vector;

[0029] When the number of evaluations reaches or is greater than the learning period of the adaptive replacement operation, perform adaptive replacement to obtain the trial vector.

[0030] Preferably, the process of adaptive replacement is as follows:

[0031] Calculate the success rate of each wind turbine coordinate index, and select the index of the coordinate to be replaced in the individual by using stochastic universal sampling according to the success rate information k , and use the coordinates of the first wind turbine in the candidate layout solution to replace the coordinate corresponding to the index k in the target vector to obtain the trial vector.

[0032] Preferably, the formula for calculating the success rate of each wind turbine coordinate index is as follows:

[0033]

[0034] wherein represents the current number of evaluations, is the learning period of the adaptive replacement operation, Nsc i,g and Nfc i,g represents the state of whether the trial vector after the replacement operation satisfies the constraint conditions, Nse i,g and Nfe i,g represents the state of whether the trial vector after selection is better than the target vector, ; is a constant.

[0035] Preferably, in step S5, the specific process of performing a repair and replacement operation on the trial variables to obtain the repaired trial vector is as follows:

[0036] Find the coordinates in the trial vector that violate the constraint conditions, and replace that coordinate with the next coordinate in the candidate layout solution. Repeat this process until the p th coordinate is replaced to obtain the repaired trial vector; p = N / 2, N is the number of wind turbines in the wind farm.

[0037] The present invention also discloses a wind farm layout optimization system based on adaptive differential evolution, including a memory and a processor connected to each other. A computer program is stored on the memory, and the computer program executes the steps of the above method when being run by the processor.

[0038] Compared with the prior art, the advantages of the present invention are as follows:

[0039] The wind farm layout optimization method based on adaptive differential evolution of the present invention first obtains the wind energy data of the wind farm, the parameters of the wind farm and the wind turbines and performs preprocessing, establishes a wind farm layout optimization model for reducing the wake effect and improving the output power, then randomly generates a certain number of wind farm layout schemes that meet the constraint conditions to form an initial population, evaluates the initial population based on the objective function, then randomly divides the initial population into several subpopulations, performs subpopulation search to obtain mutant vectors, obtains candidate layout solutions by performing local differences on the mutant vectors, performs an adaptive replacement operation on the candidate layout solutions to obtain trial variables, then performs a repair operation on the trial variables, and finally performs selection to obtain suitable individuals, and continues the above iterative process for the population until the set maximum evaluation number is reached, so as to obtain the optimal layout scheme. The present invention can optimize the layout of onshore or offshore wind farms within an acceptable time range, find the best layout of the wind turbines to minimize the wake effect to the greatest extent, and thus maximize the expected power generation.

[0040] The present invention uses a Gaussian function with a high degree of fitting to the actual wind turbine power curve data to represent the power curve, and establishes a wind farm layout optimization model that conforms to the actual situation of the wind farm; performs wind farm layout optimization based on the differential evolution method with simple structure, easy implementation, fast convergence speed, and strong robustness. Combining the characteristics of the "curse of dimensionality" that is likely to occur due to the increase in the number of wind turbines during wind farm layout optimization, a local adaptive replacement strategy is adopted. While maintaining the global search ability, under the premise of the same number of evaluations, it is easier to find the optimal solution using the learning experience in the early stage, thereby improving the convergence speed and stability of the algorithm. In addition, considering that the wind farm layout problem has strict constraint conditions, the calculation speed of the algorithm can be accelerated through the repair and replacement operations, and the quantity and quality of the feasible solutions can be improved. The present invention is applicable to offshore or onshore wind farms of various shapes, has high efficiency in layout optimization, and can provide layout schemes with high optimization quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a flowchart of the wind farm layout optimization method of the present invention in an embodiment.

[0042] Figure 2 It is a schematic diagram of the adaptive replacement operation and the repair and replacement operation in the embodiment of the present invention.

[0043] Figure 3 It is a schematic diagram of the structure of "success and failure memory" in the embodiment of the present invention.

[0044] Figure 4 It is an update schematic diagram of "success and failure memory" in the embodiment of the present invention.

[0045] Figure 5It is a comparison diagram of the traditional layout scheme and the optimized layout scheme of the present invention; among them, (a) is the traditional layout scheme; (b) is the optimized layout scheme. Detailed implementation manners

[0046] The present invention will be further described below in conjunction with the accompanying drawings of the specification and specific embodiments.

[0047] As Figure 1 shown, the wind farm layout optimization method based on adaptive differential evolution provided by the embodiment of the present invention specifically includes the steps:

[0048] S1. Obtain the wind energy data of the wind farm, the parameters of the wind farm and the wind turbines, and perform preprocessing to establish a wind farm layout optimization model for reducing the wake effect and increasing the output power; wherein the wind energy data includes wind speed, wind direction, wind energy density, wind speed probability distribution, air temperature, air pressure, humidity, etc.; the wind farm parameters include the location, scale, area and layout of the wind farm, etc.; the wind turbine parameters include the rated power of the wind turbine, cut-in wind speed, cut-out wind speed, wind turbine height and rotor diameter, etc.

[0049] The specific process of step S1 is as follows:

[0050] S11. Obtain the parameters of the wind farm and the wind turbines, as well as the wind direction and wind speed data measured by the wind measurement towers of the wind farm for 12 consecutive months. There are N wind turbines in the wind farm, and the indexes of each wind turbine are i , i = 1, …, N ;

[0051] Preprocess the wind direction data of the wind energy data: divide the wind direction into g (such as 16) integer intervals, and calculate the probability of each interval, n = 1, …, g ;

[0052] Preprocess the wind speed data of the wind energy data: use the two-parameter Weibull distribution to represent the wind speed probability distribution model, and divide the wind speed into s (such as 34) integer intervals;

[0053] Establish a wake model using the Jensen model;

[0054] Use the Gaussian function to represent the wind turbine power curve model, and the formula is as follows:

[0055]

[0056] Where is the power of the wind turbine i , = 3000kW, is the rated power of the wind turbine; is the wind speed, , and are the cut-in wind speed, rated wind speed and cut-out wind speed of the wind turbine respectively, , and have values of 3 m / s, 11.5 m / s and 25 m / s respectively, a , b and c are the parameters obtained by fitting the wind turbine power curve data with the Gaussian function respectively, a , b and c are specifically 2959, 12.85 and 4.934;

[0057] S12. Combine the two-parameter Weibull wind speed probability distribution model, Jensen wake model and wind turbine power curve model to establish the expected power output function of each wind turbine. The formula is as follows:

[0058]

[0059] where is the expected power output of wind turbine i , is the rated power of the wind turbine, is the scaling parameter considering the wake effect, is the shape parameter, is the probability of each interval of the wind direction, is the median of each interval of the divided wind direction, n = 1, …, g , is the median of each interval of the divided wind speed, , is each interval of the wind speed, j = 1, …, s ;

[0060] Finally, the variables of the wind farm layout optimization model are the two-dimensional coordinates of each wind turbine, expressed as x i1 , x i2 , i = 1, …, N ; For example N = 16;

[0061] The objective function of the wind farm optimization model represents the maximum total power output expectation of each wind turbine i , which is as follows:

[0062]

[0063] The boundary constraint conditions of the wind farm layout optimization model are as follows: The wind turbines are within the shape range of the wind farm, such as within a rectangular wind farm with a horizontal length of 2500 m and a vertical length of 4000 m.

[0064] The distance constraint conditions of the wind farm layout optimization model are as follows: The safety distance between two wind turbines is m times the rotor diameter of the wind turbine; where m ranges from 3 to 5; in this embodiment, it is taken as 4 times.

[0065] S2. Under the condition that the number of wind turbines in the wind farm is N (such as 16), randomly generate N p (such as 12) wind farm layout schemes that meet the constraint conditions to form an initial population P; where each individual is a layout scheme, and each individual is represented as: { x 11 , x 12 , x 21 , x 22 , …, x N1 , x N2};

[0066] S3. Randomly divide the initial population P into N f groups (such as 4 groups) of sub - populations, and then perform sub - population search to obtain a mutation vector. The specific process is as follows:

[0067] Set the condition for neighborhood exploration as: The number of evaluation times reaches a certain number and the objective function value remains unchanged for a certain number of times;

[0068] First, judge whether the set condition for neighborhood exploration is reached. If it is reached, perform the mutation operation of "NSDE / best / 1" for neighborhood search (which is a conventional mutation strategy in differential evolution algorithm); if it is not reached, perform the traditional "DE / best / 1" mutation operation (which is also a classic mutation strategy in differential evolution algorithm) to obtain a mutation vector.

[0069] S4. Based on the mutation vector, implement a local adaptive replacement strategy for the target vector to obtain a trial vector:

[0070] First, perform local difference to obtain a candidate layout solution. Specifically:

[0071] Re-encode the mutation vector, and regard a wind farm layout plan as a population Q; an individual among them is the coordinate of a wind turbine. Perform local difference on the population Q, and then re-encode the differential offspring population once to obtain a candidate layout solution (corresponding to Figure 2 the candidate solution in u 11 , u 12 , u 21 , u 22 , u 31 , u 32 ,… , u N1 , u N2 ).

[0072] Secondly, implement the adaptive replacement strategy, as Figure 2 shown, and the specific process is as follows:

[0073] Set the learning period of the adaptive replacement operation to L p, such as 150 times;

[0074] (1) When the number of evaluation times is less than 150, perform random replacement and record the random replacement information. Specifically:

[0075] Use the coordinate of the first wind turbine in the candidate layout solution to randomly replace a coordinate in the target vector, and the index of the replaced coordinate is r , obtain the trial vector, and record the random replacement information in the "Success and failure memories (SFM)", as Figure 3 shown. The information to be recorded in the "Success and failure memories (SFM)" includes: Nsc i,G , Nfc i,G , Nse i,G and Nfe i,G , where i = 1, …, N , G represents the current number of evaluation times, Nsc i,G and Nfc i,G represent the state of whether the trial vector after the replacement operation satisfies the constraint conditions, Nse i,G and Nfe i,GIndicates the status of whether the trial vector after selection is better than the target vector;

[0076] When the trial vector after replacement operation satisfies the constraint conditions: Nsc r,G = 1 ， And Nfc i,G = 0, i ≠ r ; Otherwise Nfc r,G = 1 ， And Nsc i,G = 0, i ≠ r ;

[0077] When the trial vector after selection is better than the target vector: Nse r,G = 1 ， And Nfe i,G = 0, i ≠ r ; Otherwise Nfe r,G = 1 ， And Nse i,G = 0, i ≠ r ; Where r Is the index of the coordinate to be replaced for the adaptive replacement operation.

[0078] (2) When the number of evaluations reaches or is greater than 150, calculate the success rate corresponding to each coordinate index i According to the success rate information, formulate a replacement strategy, perform adaptive replacement and update the adaptive replacement information. The process is as follows:

[0079] Calculate the success rate of each fan coordinate index, and randomly traverse and sample according to the success rate information to select the index of the coordinate to be replaced in the individual k , and use the coordinates of the first fan of the candidate layout solution u 11 , u 12 to replace the target vector x l = ( x 11 , x 12 , …, x k1 , x k2 , …, x N1 ,x N2 the index in k the corresponding coordinates x k1 , x k2 , to obtain the test vector u l = ( x 11 , x 12 , …, u 21 , u 22 , …, x i1 , x i2 , …, x N1 , x N2 ), and update the adaptive replacement information in SFM, as Figure 4 shown. Due to SFM overflow, the earliest information includes Nsc 1,G-LP , Nfc 1,G-LP , Nse 1,G-LP and Nfe 1,G-LP will be removed so that the current replacement information can be stored in SFM.

[0080] The success rate of the coordinate index of each fan is calculated as follows:

[0081]

[0082] where G represents the current number of evaluations, Nsc i,g and Nfc i,g represent the state of whether the test vector after the replacement operation satisfies the constraint conditions, Nse i,g and Nfe i,g represent the state of whether the test vector after the selection is better than the target vector, ; is a constant and can take 0.01.

[0083] S5. As Figure 2 shown, perform a repair and replacement operation on the test vector, and finally select between the target vector and the test vector to obtain a suitable individual, specifically:

[0084] Find the test vector u l = ([[]] x 11 , x 12 , …, u 21 , u 22 , …, x i1 , x i2 ,… , x N1 , x N2 ), the coordinates in violation of the constraints x i1 , x i2 , and use the next coordinate in the candidate layout solution ( u 11 , u 12 , u 21 , u 22 , u 31 , u 32 ,… , u N1 , u N2 ) to replace the coordinates in violation of the constraints in the test vector u 21 , u 22 ; x i1 , x i2 ;

[0085] Repeat the above process until the p th coordinate is replaced to obtain the repaired test vector u l = ([[]] x 11 , x 12 , …, u 21 , u 22 , …, u p1 , u p2 ,… , x N1 , x N2); p The value of N / 2 = 8;

[0086] Finally, a selection operation is performed on the target vector and the repaired trial vector to obtain a suitable individual.

[0087] S6. Continue the S3 - S5 process for the population until the set maximum evaluation times of 150,000 are reached, thereby obtaining the optimal layout solution.

[0088] As Figure 5 shown in (a) therein, the expected power output of the traditional layout scheme is 14,712.69 kW; as Figure 5 shown in (b) therein, the expected power output of the optimized layout scheme is 18,327.56 kW; compared with the traditional layout scheme, the expected power output of the optimized layout scheme of the present invention is increased by 23.04%, and the optimization time is 87.18 seconds, which is relatively short and within an acceptable range.

[0089] The wind farm layout optimization method based on adaptive differential evolution of the present invention first obtains the wind energy data of the wind farm, the wind farm and fan parameters and performs pre - processing, establishes a wind farm layout optimization model to reduce the wake effect and improve the output power, then randomly generates a certain number of wind farm layout schemes that meet the constraint conditions to form an initial population, and evaluates the initial population based on the objective function. Then, the initial population is randomly divided into several sub - populations, and the sub - population search is performed to obtain the mutation vector, the local difference is performed on the mutation vector to obtain the candidate layout solution, the adaptive replacement operation is performed on the candidate layout solution to obtain the trial variable, the repair operation is performed on the trial variable, and finally the selection is performed to obtain a suitable individual. Continue the above - mentioned iterative process for the population until the set maximum evaluation times are reached, thereby obtaining the optimal layout scheme. The present invention can optimize the layout of onshore or offshore wind farms within an acceptable time range, find the best layout of the fans to minimize the wake effect to the greatest extent, and thus maximize the expected power generation.

[0090] The present invention uses a Gaussian function with a high fitting degree to the actual fan power curve data to represent the power curve, and establishes a wind farm layout optimization model that conforms to the actual situation of the wind farm. Based on the differential evolution method with simple structure, easy implementation, fast convergence speed, and strong robustness, the wind farm layout is optimized. Considering the characteristic that the "curse of dimensionality" is likely to occur due to the increase in the number of fans during the wind farm layout optimization, a local adaptive replacement operation strategy is adopted. While maintaining the global search ability, under the premise of the same number of evaluations, it is easier to find the optimal solution using the learning experience in the early stage, thereby improving the convergence speed and stability of the algorithm. Additionally, considering that the wind farm layout problem has strict constraints, the repair and replacement operations can accelerate the calculation speed of the algorithm and improve the quantity and quality of feasible solutions. The present invention is applicable to offshore or onshore wind farms of various shapes, has high efficiency in layout optimization, and can provide a layout scheme with high optimization quality.

[0091] An embodiment of the present invention further provides a wind farm layout optimization system based on adaptive differential evolution, including a memory and a processor connected to each other. A computer program is stored on the memory, and when the computer program is run by the processor, it executes the steps of the method described above. The present invention can implement all or part of the processes in the above embodiment methods, and can also be completed by hardware related to computer program instructions. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable storage medium includes: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal, and software distribution medium, etc. The memory is used to store computer programs and / or modules. The processor realizes various functions by running or executing the computer programs and / or modules stored in the memory, and by calling the data stored in the memory. The memory can include high-speed random access memory, and can also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card (Flash Card), at least one magnetic disk storage device, flash device, or other volatile solid-state storage devices, etc.

[0092] The above are only the preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the concept of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art, several improvements and refinements made without departing from the principle of the present invention should be regarded as within the protection scope of the present invention.

Claims

1. An optimization method for wind farm layout based on adaptive differential evolution, characterized in that Including the steps: S1. Obtain the wind energy data of the wind farm, the parameters of the wind farm and the wind turbines, and establish an optimization model for the wind farm layout; S2. Randomly generate a preset number of wind farm layout schemes that meet the constraint conditions to form an initial population; S3. Randomly divide the initial population into several sub-populations, and then perform sub-population search to obtain a mutation vector; S4. Perform local difference on the mutation vector to obtain a candidate layout solution. Based on the candidate layout solution, perform an adaptive replacement operation on the target vector to obtain a trial vector; S5. Perform a repair replacement operation on the trial variable to obtain a repaired trial vector, and then select between the target vector and the repaired trial vector to obtain a suitable individual; S6. Continue to execute the process of steps S3 - S5 on the population until the set maximum evaluation times are reached, and finally obtain the optimal wind farm layout solution; In step S4, the specific process of performing local difference on the mutation vector to obtain a candidate layout solution is as follows: Re-encode the mutation vector, take a wind farm layout scheme as a population Q, where an individual is the coordinate of a wind turbine, perform local difference on the population Q, and then re-encode the differentiated offspring population once to obtain a candidate layout solution; In step S4, the specific process of performing an adaptive replacement operation on the target vector based on the candidate layout solution to obtain a trial vector is as follows: When the number of evaluations is less than the learning period of the adaptive replacement operation, the coordinates of the first wind turbine in the candidate layout solution are used to randomly replace one coordinate in the target vector, and the index of the replaced coordinate is r , obtaining a trial vector; When the evaluation times reach or are greater than the learning period of the adaptive replacement operation, use adaptive replacement to obtain a trial vector; The process of adaptive replacement is as follows: Calculate the success rate of the coordinate indices of each fan, and randomly traverse and sample according to the success rate information to select the index of the coordinate to be replaced in the individual k , and use the coordinates of the first fan in the candidate layout solution to replace the coordinates corresponding to the index k in the target vector to obtain a trial vector.

2. The method for optimizing the layout of a wind farm based on adaptive differential evolution according to claim 1, wherein The specific process of step S1 is as follows: S11. Obtain the wind energy data of the wind farm, the parameters of the wind farm and the wind turbines and perform preprocessing, and establish a wind speed probability distribution model, a wake model, and a wind turbine power curve model; S12. Based on the wind speed probability distribution model, the wake model, and the wind turbine power curve model, establish the power output expectation function of each wind turbine, and then obtain the wind farm optimization model.

3. The wind farm layout optimization method based on adaptive differential evolution according to claim 2, wherein In step S11, use a two-parameter Weibull distribution to represent the wind speed probability distribution model; use the Jensen model to establish the wake model; use a Gaussian function to represent the wind turbine power curve model, and the formulas are as follows: wherein is the power of the fan i , is the rated power of the fan; is the wind speed , and are the cut-in wind speed, rated wind speed and cut-out wind speed of the fan respectively a , b and c are the parameters obtained by fitting the fan power curve data with the Gaussian function respectively.

4. The method for optimizing the layout of a wind farm based on adaptive differential evolution according to claim 3, wherein, In step S12, the formula of the power output expectation function is as follows: where is the expected power output of the wind turbine i , is the rated power of the wind turbine is the scaling parameter considering the wake effect is the shape parameter is the probability of each wind direction interval is the median of each divided wind direction interval n = 1, …, g ; is the median of each divided wind speed interval; , is evenly divided into s intervals, and each interval is , , j = 1, …, s ; g is the number of intervals for evenly dividing the wind direction angle; The variables of the wind farm layout optimization model are the two-dimensional coordinates of each wind turbine, expressed as x i1 , x i2 , i = 1, …, N ; N is the number of wind turbines in the wind farm; The objective function of the wind farm layout optimization model is expressed as the following formula: The boundary constraint conditions of the wind farm layout optimization model are: The wind turbines are within the preset range; The distance constraint conditions of the wind farm layout optimization model are: The safety distance between two wind turbines is m times the rotor diameter of the wind turbine.

5. The wind farm layout optimization method based on adaptive differential evolution according to claim 1, characterized in that Calculate the success rate of the coordinate index of each fan The formula is as follows: Among them represents the current number of evaluations, is the learning period of the adaptive replacement operation, Nsc i,g and Nfc i,g represent the status of whether the trial vector after the replacement operation meets the constraint conditions, Nse i,g and Nfe i,g represent the status of whether the trial vector after selection is better than the target vector, ; is a constant; N is the number of wind turbines in the wind farm.

6. The method for optimizing the layout of a wind farm based on adaptive differential evolution according to claim 5, wherein In step S5, the specific process of performing a repair replacement operation on the trial variable to obtain a repaired trial vector is as follows: Find the coordinates in the test vector that violate the constraint conditions, and replace those coordinates with the next coordinates in the candidate layout solution. Repeat this process until the p th coordinate is reached to obtain the repaired test vector; p = N / 2.

7. A wind farm layout optimization system based on adaptive differential evolution, comprising a memory and a processor connected to each other, wherein a computer program is stored on the memory, and is characterized in that The computer program, when run by a processor, executes the steps of the method according to any one of claims 1 - 6.

Images