Wind turbine arrangement optimization method based on sparrow optimization algorithm

Abstract

The invention is suitable for the technical field of wind power generation, and provides a sparrow optimization algorithm-based wind turbine arrangement optimization method, which comprises the following steps of: respectively determining initial position matrixes of wind turbines in a flush arrangement mode and a staggered arrangement mode in a wind power plant area; optimizing the arrangement position of each wind turbine in the wind power plant area by using a sparrow optimization algorithm to obtain the optimal arrangement position of each wind turbine in the wind power plant area; an updating formula of a discoverer in the sparrow optimization algorithm is determined based on a dynamic weight strategy, the fitness of individuals in the sparrow optimization algorithm is obtained through calculation based on a wind turbine wake flow model, and individuals of an initial population of the sparrow optimization algorithm are obtained based on an initial position matrix and a heuristic initialization mode. Each individual of the initial population comprises coordinate values and yaw angles of all wind turbines in the wind power plant area in a preset coordinate system. The optimization efficiency of the arrangement position of the wind turbine can be improved.

Description

Technical Field

[0001] The present application belongs to the field of wind power generation technology, and in particular to a wind turbine layout optimization method based on a sparrow optimization algorithm. Background Art

[0002] With the depletion of fossil energy and increasing environmental pressure, many scholars and engineers have begun to develop and utilize new energy represented by wind energy. Improving the quality and efficiency of wind farms has been an important factor in promoting the advancement of wind power generation technology in recent years. For wind farms, the layout of each wind turbine is particularly important, which directly affects the power generation efficiency. At present, the modeling solution method is generally used to optimize the layout of wind turbines. Due to the complexity of the modeling solution method, the optimization efficiency is low. Summary of the invention

[0003] The embodiment of the present application provides a wind turbine layout optimization method based on the sparrow optimization algorithm, which can solve the problem of low optimization efficiency of wind turbine layout positions.

[0004] The embodiment of the present application provides a wind turbine layout optimization method based on the sparrow optimization algorithm, including:

[0005] Determine the initial position matrix of each wind turbine in the wind farm area in the even arrangement mode and the staggered arrangement mode respectively; the initial position matrix of each wind turbine in the even arrangement mode includes the position coordinates of all wind turbines in the wind farm area in the even arrangement mode, and the initial position matrix of each wind turbine in the staggered arrangement mode includes the position coordinates of all wind turbines in the wind farm area in the staggered arrangement mode;

[0006] The sparrow optimization algorithm is used to optimize the layout of each wind turbine in the wind farm area to obtain the optimal layout of each wind turbine in the wind farm area; the update formula of the finder in the sparrow optimization algorithm is determined based on the dynamic weight strategy, the fitness of the individual in the sparrow optimization algorithm is calculated based on the wind turbine wake model, and the individuals of the initial population of the sparrow optimization algorithm are obtained based on the initial position matrix and the heuristic initialization method. Each individual in the initial population includes the coordinate values ​​and yaw angles of all wind turbines in the wind farm area in the preset coordinate system.

[0007] Optionally, before the step of optimizing the positions of each wind turbine in the wind farm area by using the sparrow optimization algorithm to obtain the optimal layout positions of each wind turbine in the wind farm area, the wind turbine layout optimization method further includes:

[0008] Convert the initial position matrix of each wind turbine in the parallel arrangement into a one-dimensional vector A j , and convert the initial position matrix of each wind turbine in the staggered arrangement into a one-dimensional vector A m ;

[0009] By formula Generate and vector A j Sobol sequences of the same length n ; a is Sobol sequence n The is-th point in, φ(is) is the is-th Gray code, m is the sequence length, b j is the j-th binary number, v d,j for b j The corresponding direction number;

[0010] By the formula k = k cr D.s n Generate range -k cr D to k cr The difference sequence k of D; k cr is the given variation range, and D is the wind turbine diameter.

[0011] Optionally, the position coordinates of the wind turbine are coordinate values ​​of the wind turbine in a preset coordinate system, and the coordinate values ​​include coordinate values ​​of the wind turbine on the horizontal axis and the vertical axis of the preset coordinate system.

[0012] Optionally, the initialization process of the initial population is:

[0013] Init by formula ic =A jm +k ic Get the icth initial vector Init ic ; A jm is the vector A j With vector A m Any one of jm A j The probability of A jm A m The probability of k ic is the element at the icth position in the difference sequence k, ic = 1, ..., I, I is the number of elements in the difference sequence k;

[0014] Randomly generate a vertical position matrix and a yaw angle matrix of each wind turbine in the wind farm area, and convert the vertical position matrix and the yaw angle matrix into one-dimensional vectors respectively; the vertical position matrix includes the coordinate values ​​of all wind turbines in the wind farm area on the vertical axis of a preset coordinate system, and the yaw angle matrix includes the yaw angles of all wind turbines in the wind farm area;

[0015] For the icth initial vector Init ic The icth individual in the initial population is obtained by concatenating the one-dimensional vector corresponding to the vertical position matrix and the concatenated vector with the one-dimensional vector corresponding to the yaw angle matrix.

[0016] Optionally, the update formula of the discoverer is:

[0017]

[0018] Among them, X new is the new position of the individual, X old is the old position of the individual, ω is the dynamic weight, rand 1×dim is a random one-dimensional vector of dimension dim, R2 is the warning value, ST is the safety value, r1 is a random number between (0, 1), M is the maximum number of iterations, and iy is the iy-th individual.

[0019] Optionally, the adjustment formula of the dynamic weight ω is:

[0020]

[0021] Wherein, ω(t) is the dynamic weight when updating the position of the discoverer in the t-th iteration, ω(t-1) is the dynamic weight when updating the position of the discoverer in the t-1th iteration, ω(0)=1, ε is a preset constant term, E t =β1E t-1 +(1-β1)U t , β1=0.9, E0 is the zero vector, V t =β2V t-1 +(1-β2)U t 2 , β2=0.95, V0 is the zero vector, N is the total number of wind turbines, u M (x i ,y i ,z i ) is the wind speed at the i-th wind turbine rotor.

[0022] Optionally, the wind turbine wake model is used to describe the wind speed deficit when a grid point in the wind farm area is affected by the wakes of multiple wind turbines. The grid point is a single grid area after the wind farm area is gridded. The expression of the wind turbine wake model is:

[0023]

[0024] Among them, u M (x,y,z) is the wind speed value corresponding to the grid point (x,y,z), u0 is the fan inflow velocity, z h is the reference height, a is the axial induction factor of the wind turbine, C T is the wind turbine thrust coefficient, N is the total number of wind turbines, FQ iis the wind speed loss of the i-th wind turbine;

[0025]

[0026] Where p is the p coordinate of the grid point, y d is the center offset of the wake impact area, o is the o coordinate of the grid point, β is the yaw angle, k w is the wake development coefficient, r q is the wake influence radius in the q-axis direction, n is the power index, u * is the mass flux downstream of the wind turbine, u * =u0(1-2a / (1+ko / r0) 2 ), k is the growth rate of Jensen wake radius, h0 is the height of wind turbine, r0 is the radius of wind turbine, Δu is the wind speed difference considering shear inflow,

[0027] Optionally, the calculation formula of the fitness DAEP of an individual in the sparrow optimization algorithm is:

[0028]

[0029] in, K is the total number of wind direction angles, A(θ) is the probability of incoming flow with wind direction angle θ, and u a i is the average inflow wind speed of the i-th wind turbine, w(·) is the power calculation function of a single wind turbine, v m is the wind speed when the maximum power is reached, σ is the standard deviation of the wind turbine power curve, and v is u a i ; S r0 is the rotor area, ds represents the area integral, (x i ,y i ) and (x j ,y j ) is the horizontal and vertical coordinates of any two wind turbines in the wind farm area, and 2.5D is the minimum spacing between wind turbines.

[0030] Optionally, in the process of optimizing the layout positions of each wind turbine in the wind farm area using the sparrow optimization algorithm, the best individual in the population obtained when the iteration termination condition is met is used as the optimal layout position of each wind turbine in the wind farm area.

[0031] Optionally, the best individual is the one with the largest fitness.

[0032] The above solution of the present application has the following beneficial effects:

[0033] In an embodiment of the present application, in the process of optimizing the layout positions of each wind turbine using the improved sparrow optimization algorithm, the discoverer position in the sparrow optimization algorithm is updated based on the dynamic weight strategy. When the fitness of the current generation changes little from the fitness of the previous generation, the weight is dynamically adjusted so that the discoverer has a larger update step size to help the population jump out of the local optimal solution and achieve the effect of improving the optimization efficiency of the wind turbine layout position.

[0034] Other beneficial effects of the present application will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0036] Figure 1 A flow chart of a wind turbine layout optimization method based on a sparrow optimization algorithm provided in one embodiment of the present application;

[0037] Figure 2 A schematic diagram of a local coordinate system in an embodiment of the present application;

[0038] Figure 3 It is a probability map of the joint distribution of wind speed and direction in an example;

[0039] Figure 4 A wind turbine layout point map in an example;

[0040] Figure 5 FIG. 4 is a graph showing the overall power improvement curve of a wind farm in an example. DETAILED DESCRIPTION

[0041] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0042] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0043] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0044] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0045] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0046] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0047] In response to the current problem of low optimization efficiency of wind turbine layout positions, an embodiment of the present application provides a wind turbine layout optimization method based on a sparrow optimization algorithm. The method utilizes an improved sparrow optimization algorithm to optimize the layout positions of each wind turbine in a wind farm area. During the optimization process, the discoverer position in the sparrow optimization algorithm is updated based on a dynamic weight strategy. When the fitness of the current generation changes little from the fitness of the previous generation, the weight is dynamically adjusted so that the discoverer has a larger update step, thereby helping the population to jump out of the local optimal solution and achieving the effect of improving the optimization efficiency of the wind turbine layout position.

[0048] The wind turbine layout optimization method based on the sparrow optimization algorithm provided in the embodiment of the present application is exemplarily described below in conjunction with specific embodiments.

[0049] like Figure 1 As shown, the wind turbine layout optimization method based on the sparrow optimization algorithm provided in the embodiment of the present application includes the following steps:

[0050] Step 11, respectively determining the initial position matrix of each wind turbine in the wind farm area in the even arrangement mode and the staggered arrangement mode.

[0051] The initial position matrix of the above-mentioned wind turbines in the even-positioned arrangement (also called "matrix" type) includes the position coordinates of all wind turbines in the wind farm area in the even-positioned arrangement, and the initial position matrix of the above-mentioned wind turbines in the staggered arrangement (also called "plum blossom" type) includes the position coordinates of all wind turbines in the wind farm area in the staggered arrangement.

[0052] The position coordinates of the above-mentioned wind turbine are the coordinate values ​​of the wind turbine in the preset coordinate system, and the coordinate values ​​include the coordinate values ​​of the wind turbine on the horizontal axis and the vertical axis of the preset coordinate system. The preset coordinate system is an overall coordinate system, which is unique. Its origin can be selected as any point. Once determined, it will not be moved. The horizontal axis (x-axis) and the vertical axis (y-axis) can be any set of orthogonal basis vectors on the plane (the plane is perpendicular to the height direction of the wind turbine). After determination, it will not be moved. The vertical axis (z-axis) is the coordinate axis in the height direction of the wind turbine. Specifically, within the target site size range (i.e., the wind farm area), appropriate spacing can be selected to determine the positions of each wind turbine in a "matrix" or "plum blossom" manner.

[0053] Exemplarily, the number of rows of the initial position matrix is ​​the total number of wind turbines, the first column in the initial position matrix is ​​the coordinate value of each wind turbine on the horizontal axis of the preset coordinate system, the second column is the coordinate value of each wind turbine on the vertical axis of the preset coordinate system, and each row of the initial position matrix is ​​the coordinate value of the same wind turbine on the horizontal axis and vertical axis of the preset coordinate system.

[0054] After determining the initial position matrices under the two arrangement modes, the above wind turbine layout optimization method further includes the following preprocessing steps to subsequently optimize the layout positions of the wind turbines:

[0055] Convert the initial position matrix of each wind turbine in the parallel arrangement into a one-dimensional vector A j , and convert the initial position matrix of each wind turbine in the staggered arrangement into a one-dimensional vector A m ; One-dimensional vector A j With the one-dimensional vector A m are the same length;

[0056] By formula Generate and vector A j Sobol sequences of the same length n ; a is Sobol sequences n is the is-th point in, φ(is) is the is-th Gray code, m is the sequence length (i.e. the length of the Sobol sequence), b jis the j-th binary number, v d,j for b j The corresponding direction number;

[0057] By the formula k = k cr D.s n Generate range -k cr D to k cr The difference sequence k of D; k cr is the given variation range, and D is the wind wheel diameter.

[0058] It should be noted that since the ordinal number i is used in many places, but the meaning of i in different places is not necessarily the same, is is used as the ordinal number in the sobol sequence, ic is used as the ordinal number in the initialization process, iy is used as the ordinal number in the optimization algorithm, and i is used as the ordinal number of the wind turbine.

[0059] Step 12, using the sparrow optimization algorithm to optimize the layout of each wind turbine in the wind farm area to obtain the optimal layout of each wind turbine in the wind farm area; the update formula of the finder in the sparrow optimization algorithm is determined based on the dynamic weight strategy, the fitness of the individual in the sparrow optimization algorithm is calculated based on the wind turbine wake model, and the individuals of the initial population of the sparrow optimization algorithm are obtained based on the initial position matrix and the heuristic initialization method. Each individual of the initial population includes the coordinate values ​​and yaw angles of all wind turbines in the wind farm area in the preset coordinate system.

[0060] The above wind turbine wake model is used to describe the wind speed loss when the grid point in the wind farm area is affected by the wake of multiple wind turbines. The grid point is a single grid area after the wind farm area is gridded. The expression of the wind turbine wake model is:

[0061]

[0062] Among them, u M (x,y,z) is the wind speed value corresponding to the grid point (x,y,z), u0 is the fan inflow velocity, z h is the reference height, α is the axial induction factor of the wind turbine, C T is the wind turbine thrust coefficient, N is the total number of wind turbines, FQ i is the wind speed loss of the ith wind turbine, that is, the wind speed loss generated by the ith wind turbine at the position (x, y, z), x is the horizontal coordinate of the wind turbine in the overall coordinate system, y is the vertical coordinate of the wind turbine in the overall coordinate system, and z is the vertical coordinate of the wind turbine in the overall coordinate system.

[0063] The derivation process of the wind turbine wake model is exemplified below.

[0064] Since the wind turbine wake model is calculated based on the wind turbine, the coordinate system involved in the derivation of the wind turbine wake model is a local coordinate system, that is, the coordinate system of each wind turbine is different. Figure 2 As shown, for any wind turbine 200, its corresponding local coordinate system is composed of an origin, an o-axis, a p-axis and a q-axis which are perpendicular to each other, the origin is the center of the wind turbine's rotor, the q-axis is the coordinate axis in the height direction of the wind turbine, the o-axis is perpendicular to the wind turbine hub plane, and the p-axis is perpendicular to the oq plane.

[0065] First, the linear expansion rate of the longitudinal and vertical wake radii downstream of any wind turbine is calculated using the anisotropic wake radius expansion rate calculation formula. The calculation formula is as follows:

[0066]

[0067] In the above formula: k p is the linear expansion rate of the wake radius in the longitudinal direction (i.e., p-axis direction), k q is the linear expansion rate of the wake radius in the vertical direction (i.e., q-axis direction), C T is the thrust coefficient of the wind turbine, and I0 is the turbulence of the incoming wind at the hub height.

[0068] Then, the wake influence radius is calculated based on the linear assumption:

[0069]

[0070] In the above formula: r p is the longitudinal (i.e., p-axis) wake impact radius, o is the coordinate value of the wind turbine on the o-axis, r0 is the wind turbine radius, r q is the vertical (i.e. q-axis direction) wake influence radius.

[0071] After the linear expansion rate and wake influence radius are determined, the Jensen wake model is used to calculate the mass flux downstream of any wind turbine. The formula is as follows:

[0072] u * =u0(1-2a / (1+ko / r0) 2 )

[0073] In the above formula: u * is the mass flux downstream of the wind turbine, k is the Jensen wake radius growth rate, u0 is the wind turbine inflow velocity, a is the wind turbine axial induction factor, and k is the wake radius growth rate.

[0074] Then, based on the Jiménez wake center offset assumption, the calculation formula for calculating the center offset of the wake influence area is obtained:

[0075]

[0076] In the above formula: y d is the center offset of the wake influence area, β is the yaw angle, k w is the wake development coefficient, and o is the o-coordinate of the grid point (i.e., the coordinate value of the grid point on the o-axis of the local coordinate system).

[0077] Combined with the assumption of power function wind deficit distribution, the average wake velocity downstream of any wind turbine is calculated as:

[0078]

[0079] In the above formula: u(o,p,q) is the velocity calculation result under the influence of a single fan at point (o,p,q) (i.e., at the grid point), r q is the wake influence radius in the q direction (i.e., q-axis direction), n is the power index, h0 is the height of the wind turbine, Δu is the wind speed difference considering the shear inflow,

[0080] In the above formula, the wind speed loss expression caused by a single wind turbine is defined as:

[0081]

[0082] In the above formula: FQ is the wind speed loss of a single wind turbine, p is the p-coordinate of the grid point (i.e. the coordinate value of the grid point on the p-axis of the local coordinate system), and q is the q-coordinate of the grid point (i.e. the coordinate value of the grid point on the q-axis of the local coordinate system). It should be noted that when calculating the wind speed loss FQ of the i-th wind turbine, i When the relevant parameters of the i-th wind turbine are substituted into the calculation formula of FQ, FQ can be obtained. i .

[0083] Based on the square sum law of the superposition of the wakes of multiple wind turbines, when a grid point is affected by the wakes of multiple wind turbines, its expression is the wind turbine wake model:

[0084] In some embodiments of the present application, the initialization process of the initial population includes the following steps:

[0085] Step 12.1, through the formula Init ic =A jm +k ic Get the icth initial vector Init ic ; A jm is the vector A j With vector A m Any one of jm A j The probability of A jm A m The probability of kic is the element at the icth position in the difference sequence k (i.e., the value at the icth position), ic = 1,…,I, I is the number of elements in the difference sequence k.

[0086] Step 12.2, randomly generate the vertical position matrix and yaw angle matrix of each wind turbine in the wind farm area, and convert the vertical position matrix and yaw angle matrix into one-dimensional vectors respectively; the vertical position matrix includes the coordinate values ​​of all wind turbines in the wind farm area on the vertical coordinate axis of the preset coordinate system, and the yaw angle matrix includes the yaw angles of all wind turbines in the wind farm area.

[0087] Exemplarily, the vertical position matrix and the yaw angle matrix may be randomly generated by a Latin hypercube sampling method, so as to optimize the height and yaw angle of the wind turbine.

[0088] Step 12.3, for the icth initial vector Init ic The icth individual in the initial population is obtained by concatenating the one-dimensional vector corresponding to the vertical position matrix and the concatenated vector with the one-dimensional vector corresponding to the yaw angle matrix.

[0089] It can be seen that each individual obtained through the above initialization process contains the coordinate values ​​and yaw angles of all wind turbines in the preset coordinate system, so that the optimal individual can be obtained by iteratively updating the initial population. It can be understood that in the process of optimizing the layout of each wind turbine in the wind farm area using the sparrow optimization algorithm, the optimal individual in the population obtained when the iteration termination condition is met can be used as the optimal layout of each wind turbine in the wind farm area. Among them, the optimal individual is the individual with the highest fitness.

[0090] As an optional example, the above-mentioned iteration termination condition may be: the number of iteration updates reaches a preset maximum number of iterations, there is an individual with a fitness greater than or equal to a preset fitness, or the change in fitness (i.e., the change in the maximum fitness of the current generation relative to the maximum fitness of the previous generation) is less than a preset threshold. It should be noted that when the iteration termination condition is met, the update of the population will be stopped, and the optimal individual in the population obtained at this time can be used as the optimal layout position of each wind turbine in the wind farm area. It should be noted that when the initial population individual is obtained, the coordinate position matrix is ​​fully connected to obtain a vector. Therefore, at this time, each coordinate position and yaw angle can be obtained by reshaping through the inverse process of full connection.

[0091] In some embodiments of the present application, during the optimization process, the dimensionless annual energy production (DAEP) can be used as the fitness of the individual. Accordingly, the calculation formula of the fitness DAEP of the individual in the sparrow optimization algorithm is:

[0092]

[0093] Where W is the average total power of the wind farm, K is the total number of wind direction angles, A(θ) is the probability of incoming flow with a wind direction angle of θ, and u a i is the average inflow wind speed of the i-th wind turbine, w(·) is the power calculation function of a single wind turbine, v m is the wind speed when the maximum power is reached (i.e. the wind speed when W reaches the maximum), σ is the standard deviation of the wind turbine power curve, and v is u a i .

[0094] u a is the average inflow wind speed, S r0 is the rotor area, ds represents the area integral, (x i ,y i ) and (x j ,y j ) is the horizontal and vertical coordinates of any two wind turbines in the wind farm area, 2.5D is the minimum spacing between wind turbines, and WN means that the spacing between any two wind turbines less than the minimum spacing is unacceptable. It should be noted that the relevant parameters of the i-th wind turbine are substituted into u a The calculation formula for u can be obtained a i .u M (x, y, z) is the calculation result of the wind turbine wake model, and its double integral is obtained by dividing the grid and performing numerical calculations. a As an important parameter for calculating the power generation of a wind turbine, the power generation of a single wind turbine is calculated from the average incident wind speed on the rotor plane using the power curve. The power curve can be changed according to different types of wind turbines.

[0095] In some embodiments of the present application, during the optimization process, the update formula of the finder is:

[0096]

[0097] Among them, X new is the new position of the individual, X old is the old position of the individual, ω is the dynamic weight, rand 1×dim is a random one-dimensional vector of dimension dim, R2 is the warning value, its range is [0,1], ST is the safety value, its range is [0.5,1], r1 is a random number between (0,1), M is the maximum number of iterations, and iy is the iy-th individual.

[0098] In the sparrow optimization algorithm, after the positions of all individuals in the population are updated and the fitness is calculated, it is determined whether the dynamic weight ω in the next iteration needs to be updated according to the change in fitness. Generally, when the fitness of the current generation changes little from the fitness of the previous generation (that is, the change of the current maximum fitness relative to the maximum fitness of the previous generation is less than a given value), the weight ω can be dynamically adjusted so that the discoverer has a larger update step size to help the population jump out of the local optimal solution and achieve the effect of improving the optimization efficiency of the wind turbine layout position.

[0099] The adjustment formula of the above dynamic weight ω is:

[0100]

[0101] Wherein, ω(t) is the dynamic weight when updating the position of the discoverer in the t-th iteration, ω(t-1) is the dynamic weight when updating the position of the discoverer in the t-1-th iteration, ω(0) is the initialization dynamic weight, ω(0)=1, ε is a preset constant term, ε is a sufficiently small amount to prevent zero division, E t =β1E t-1 +(1-β1)U t , E t is the first-order moment vector, For E t The result after correction, U t is the modulus of the wind speed drop gradient at all wind turbine rotors, β1=0.9, E0 is the zero vector, V t =β2V t-1 +(1-β2)U t 2 , β2 = 0.95 to ensure a longer historical gradient information, V0 is the zero vector, V t is the second-order moment vector, For V t The result after correction, N is the total number of wind turbines, u M (x i ,y i ,z i ) is the wind speed at the i-th wind turbine rotor, i.e., the wind turbine wake model u M (x, y, z) calculation results. It should be noted that the wind turbine wake model u M The calculation method of (x, y, z) will be explained in detail later. It can be understood that u M (x i ,y i ,z i ) and u M (x,y,z) is calculated in the same way, except that uM (x i ,y i ,z i ) uses the relevant parameters of the i-th wind turbine, that is, substitutes the relevant parameters of the i-th wind turbine into u M The calculation formula of (x,y,z) can get u M (x i ,y i ,z i ).

[0102] In some embodiments of the present application, during the optimization process, the update formula of the follower in the sparrow search algorithm is improved by combining the Levy flight strategy:

[0103]

[0104] In the above formula, X new is the new position of the individual, rand(0,1) is a random number between 0 and 1, and X worst is the position of the individual with the worst fitness in the population, X old is the old position of the individual, iy is the iyth individual, X best is the position of the individual with the best fitness in the population, step is the random step length generated according to the levy flight strategy, step = θ / (|ω| 1 / γ ),θ~N(0,σ μ 2 ),ω~N(0,σ υ 2 ),σ υ =1, γ=1.5.

[0105] In some embodiments of the present application, during the optimization process, the alarm source position is updated according to the sparrow search algorithm framework, and the update method is shown in the following formula:

[0106]

[0107] In the above formula, X new is the new position of the individual, X old is the old position of the individual, rand(0,1) is a random number between 0 and 1, and X worst is the position of the individual with the worst fitness in the population, f iy is the fitness of the iyth individual, f g With f w are the best and worst fitness of the current population, X best is the position of the individual with the best fitness in the population, and β is a random number that obeys a normal distribution with a mean of 0 and a variance of 1.

[0108] The above wind turbine layout optimization method is exemplified by an experiment below.

[0109] In this experiment, it is assumed that 16 wind turbines are arranged in an area of ​​1000m long and 1000m wide. Two hub heights of 60m and 80m are planned. The wind turbine diameter is 60m. The wind shear index of the local annual average wind profile is 0.17. The reference height is set to 80m. The local wind speed and wind direction joint distribution probability is as follows: Figure 3 The number of individuals in the optimization algorithm is set to 50, and the maximum number of iterations is 1000. The optimized wind turbine layout points are shown in Figure 4 As shown in Figure 2, the overall power improvement curve of the wind farm is as follows: Figure 5 As shown. It can be seen that when the maximum number of iterations is reached, the overall power of the wind farm tends to be stable and reaches the optimal state. Figure 3 N, NNE, NE, ENE, E, ESE, SE, SSE, S, SSW, SW, WSW, W, WNW, NW, and NNW represent the 16 directions of wind direction; Figure 4 The XYZ in the figure are the coordinate values ​​in the global coordinate system.

[0110] To sum up, the wind turbine layout optimization method of the present application optimizes the layout position of each wind turbine in the wind farm area by utilizing the improved sparrow optimization algorithm. During the optimization process, the discoverer position in the sparrow optimization algorithm is updated based on the dynamic weight strategy. When the fitness of the current generation changes little from the fitness of the previous generation, the weight is dynamically adjusted so that the discoverer has a larger update step size to help the population jump out of the local optimal solution and achieve the effect of improving the optimization efficiency of the wind turbine layout position.

[0111] The above is a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles described in the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A wind turbine layout optimization method based on the sparrow optimization algorithm, characterized in that: include: Determine the initial position matrix of each wind turbine in the wind farm area in the parallel arrangement mode and the staggered arrangement mode respectively; The initial position matrix of each wind turbine in the even arrangement includes the position coordinates of all wind turbines in the wind farm area in the even arrangement, and the initial position matrix of each wind turbine in the staggered arrangement includes the position coordinates of all wind turbines in the wind farm area in the staggered arrangement; Optimizing the layout of each wind turbine in the wind farm area by using the sparrow optimization algorithm to obtain the optimal layout of each wind turbine in the wind farm area; The update formula of the discoverer in the sparrow optimization algorithm is determined based on a dynamic weight strategy, the fitness of the individuals in the sparrow optimization algorithm is calculated based on a wind turbine wake model, the individuals of the initial population of the sparrow optimization algorithm are obtained based on the initial position matrix and a heuristic initialization method, and each individual of the initial population includes the coordinate values ​​and yaw angles of all wind turbines in the wind farm area in a preset coordinate system.

2. The wind turbine layout optimization method according to claim 1, characterized in that: Before the step of optimizing the positions of the wind turbines in the wind farm area by using the sparrow optimization algorithm to obtain the optimal layout positions of the wind turbines in the wind farm area, the wind turbine layout optimization method further includes: Convert the initial position matrix of each wind turbine in the parallel arrangement into a one-dimensional vector A j , and convert the initial position matrix of each wind turbine in the staggered arrangement into a one-dimensional vector A m ; By formula Generate and vector A j Sobol sequences of the same length n ; a is Sobol sequence n The is-th point in, φ(is) is the is-th Gray code, m is the sequence length, b j is the j-th binary number, v d,j for b j The corresponding direction number; By the formula k = k cr D.s n Generate range -k cr D to k cr The difference sequence k of D; k cr is the given variation range, and D is the wind turbine diameter.

3. The wind turbine layout optimization method according to claim 2, characterized in that: The position coordinates of the wind turbine are coordinate values ​​of the wind turbine in a preset coordinate system, and the coordinate values ​​include coordinate values ​​of the wind turbine on the horizontal axis and the vertical axis of the preset coordinate system.

4. The wind turbine layout optimization method according to claim 3, characterized in that: The initialization process of the initial population is: Init by formula ic =A jm +k ic Get the icth initial vector Init ic ; A jm is the vector A j With vector A m Any one of jm A j The probability of A jm A m The probability of k ic is the element at the icth position in the difference sequence k, ic = 1, ..., I, I is the number of elements in the difference sequence k; Randomly generate a vertical position matrix and a yaw angle matrix of each wind turbine in the wind farm area, and convert the vertical position matrix and the yaw angle matrix into one-dimensional vectors respectively; the vertical position matrix includes the coordinate values ​​of all wind turbines in the wind farm area on the vertical coordinate axis of the preset coordinate system, and the yaw angle matrix includes the yaw angles of all wind turbines in the wind farm area; For the icth initial vector Init ic The icth individual in the initial population is obtained by splicing the one-dimensional vector corresponding to the vertical position matrix and splicing the spliced ​​vector with the one-dimensional vector corresponding to the yaw angle matrix.

5. The wind turbine layout optimization method according to claim 4, characterized in that: The update formula of the discoverer is: Among them, X new is the new position of the individual, X old is the old position of the individual, ω is the dynamic weight, rand 1×dim is a random one-dimensional vector of dimension dim, R2 is the warning value, ST is the safety value, r1 is a random number between (0, 1), M is the maximum number of iterations, and iy is the iy-th individual.

6. The wind turbine layout optimization method according to claim 5, characterized in that: The adjustment formula of dynamic weight ω is: Wherein, ω(t) is the dynamic weight when updating the position of the discoverer in the t-th iteration, ω(t-1) is the dynamic weight when updating the position of the discoverer in the t-1-th iteration, ω(0)=1, ε is a preset constant term, E t =β1E t-1 +(1-β1)U t , β1=0.9, E0 is the zero vector, β2=0.95, V0 is the zero vector, N is the total number of wind turbines, u M (x i ,y i ,z i ) is the wind speed at the i-th wind turbine rotor.

7. The method for optimizing the layout of wind turbines according to claim 6, characterized in that: The wind turbine wake model is used to describe the wind speed deficit when a grid point in the wind farm area is affected by the wakes of multiple wind turbines. The grid point is a single grid area after the wind farm area is gridded. The expression of the wind turbine wake model is: Among them, u M (x,y,z) is the wind speed value corresponding to the grid point (x,y,z), u0 is the fan inflow velocity, z h is the reference height, a is the axial induction factor of the wind turbine, C T is the wind turbine thrust coefficient, N is the total number of wind turbines, FQ i is the wind speed loss of the i-th wind turbine; Where p is the p coordinate of the grid point, y d is the center offset of the wake impact area, o is the o coordinate of the grid point, β is the yaw angle, k w is the wake development coefficient, r q is the wake influence radius in the q-axis direction, n is the power index, u * is the mass flux downstream of the wind turbine, u * =u0(1-2a / (1+ko / r0) 2 ), k is the growth rate of Jensen wake radius, h0 is the height of wind turbine, r0 is the radius of wind turbine, Δu is the wind speed difference considering shear inflow, 8. The method for optimizing the layout of wind turbines according to claim 7, characterized in that: The calculation formula of the individual fitness DAEP in the sparrow optimization algorithm is: in, K is the total number of wind direction angles, A(θ) is the probability of incoming flow with a wind direction angle of θ, is the average inflow wind speed of the i-th wind turbine, w(·) is the power calculation function of a single wind turbine, v m is the wind speed when the maximum power is reached, σ is the standard deviation of the wind turbine power curve, and v is S r0 is the rotor area, ds represents the area integral, (x i ,y i ) and (x j ,y j ) are the horizontal and vertical coordinates of any two wind turbines in the wind farm area, and 2.5D is the minimum spacing between wind turbines.

9. The wind turbine layout optimization method according to claim 1, characterized in that: In the process of optimizing the layout positions of each wind turbine in the wind farm area by using the sparrow optimization algorithm, the best individual in the population obtained when the iteration termination condition is met is used as the optimal layout position of each wind turbine in the wind farm area.

10. The wind turbine layout optimization method according to claim 9, characterized in that: The optimal individual is the individual with the largest fitness.

Images