WIND TURBINE LAYOUT OPTIMIZATION METHOD COMBINING WITH DISPATCHING STRATEGY FOR WIND FARM

Abstract

Disclosed is a wind turbine layout optimization method combining with a dispatching strategy for the wind farm. In the wind farm micro-siting stage, the installed wind turbines number and the arrangement positions are optimized. In this method, the dispatching strategy of wind turbines is considered during the layout optimization of wind turbines, and the axial induction factor of each wind turbine is introduced into the layout optimization variables. The dispatching strategy of maximizing the wind farm power generation is combined with the layout optimization of wind turbines in the construction stage of the wind farm, so that the wake effect is effectively reduced and the capacity cost is reduced, which meet the requirement of actual wind farm. A hybrid optimization algorithm is proposed in this method, with a greedy algorithm to optimize the turbine number and a particle swarm optimization (PSO) algorithm to refine the turbine layout scheme.

Claims

1. A wind turbine layout optimization method combining with a dispatching strategy for a wind farm, comprising the following steps: step 1): obtaining terrain data, wind speed and direction measurement data and meteorological parameters of a wind farm, and performing wind resource analysis; step 2): dividing the wind farm into grids, according to a result of the wind resource analysis and a requirement of a safe distance between wind turbines, wherein a total number of the grids is a maximum possible number of installed wind turbines; taking centers of the grids as optional positions of the wind turbines to obtain a set of feasible positions for installation of the wind turbines; step 3): taking the number of the wind turbines to be installed in the area of the wind farm, an arrangement position of each wind turbine, and an axial induction factor of each wind turbine in the dispatching strategy for the wind farm collectively as optimization variables of a layout of wind turbines; applying a greedy algorithm to optimize the number of the installed wind turbines in a feasible region of the optimization variables; taking an installation number n of wind turbines with a lowest cost of energy CoE.sub.n as an optimized installation number n.sup.opt of wind turbines; then obtaining a preliminary wind turbine arrangement position optimization solution corresponding to n.sup.opt, wherein in the process of optimization, a corresponding individual fitness is composed of the cost of energy CoE.sub.n, and a calculation formula of the individual fitness fitness1 is: $\mathrm{fitness}1={\mathrm{CoE}}_n=\frac{{\mathrm{cost}}_n}{T_{\mathrm{life}}.Math.{\mathrm{AEP}}_n}$ where T.sub.life is an effective life of wind turbines, cost.sub.n is a cost of the turbine location layout optimization solution corresponding to the installation number n, and AEP.sub.a is an annual average energy production of the wind farm corresponding to the optimization solution; step 4): optimizing the arrangement positions of the n.sup.opt wind turbines obtained in step 3), removing restriction of the grids, introducing a penalty function to ensure the safe distance between the wind turbines, and optimizing the positions of the n.sup.opt wind turbines and axial induction factors thereof by using particle swarm optimization, so as to reduce the cost of energy CoE.sub.n opt and obtain an arrangement position optimization solution of the wind turbines in the continuous spatial positions within the wind farm; wherein in the process of optimization, the individual fitness consists of two parts, one of which is the cost of energy and the other of which is a distance function between the wind turbines, and the calculation formula of the individual fitness fitness2 is:
fitness2=CoE+J.Math.Dis where CoE is the cost of energy corresponding to a wind turbine layout solution combining with the dispatching strategy for the wind farm; the wind turbine layout solution includes two parts: the installation number of the wind turbines and arrangement positions of the wind turbines; J is a set penalty function coefficient, and Dis is a function that is set to ensure the safe distance between wind turbines, an expression of which is: $\mathrm{Dis}=\underset{i=1}{\overset{n^{\mathrm{opt}}}{.Math.}}\underset{\begin{array}{c}j=1\\ j\ne i\end{array}}{\overset{n^{\mathrm{opt}}}{.Math.}}\max \left\{0,D_s^2-d_{i,j}^2\right\}$ where d.sub.i,j is the straight-line distance between a wind turbine i and a wind turbine j; wherein in the layout optimization of the wind turbines according to the method, taking the dispatching strategy for the wind farm into account, the axial induction factor of each wind turbine is introduced into the layout optimization variables of wind turbines, and the dispatching strategy of maximizing a wind farm production capacity is combined with the layout optimization of wind turbines in a construction stage of the wind farm; a greedy algorithm is applied to ensure the quality of the layout solution, and then a particle swarm optimization algorithm is applied to optimizing the wind turbine layout solution for continuous spatial positions.

2. The wind turbine layout optimization method combining with a dispatching strategy for a wind farm according to claim 1, wherein the dispatching strategy for the wind farm is considered in the process of the layout optimization, and the installation number, the arrangement positions and the axial induction factors of the wind turbines are optimized, and a step-by-step layout optimization method is used; in a first step, the greedy algorithm is used to solve an optimization problem of optimizing the installation number of wind turbines; the installation number of wind turbines in the area of the wind farm, the arrangement position of each wind turbine and the axial induction factor thereof are collectively taken as optimization variables; a search domain is a feasible set of the wind farm composed of the maximum number of the wind turbines that can be installed, grid-based discrete arrangement positions of the wind turbine and an artificially discretized set of the axial induction factors; an optimization objective function is a minimum of the cost of energy; in a second step, on the basis of the preliminary wind turbine arrangement position optimization solution obtained by optimizing the number of the wind turbines, the particle swarm optimization algorithm is used for further optimization, the restriction of the grids is removed, and the penalty function is introduced into a fitness function to ensure the safe distance of the wind turbines; the positions of the n.sup.opt wind turbine and the axial induction factors are optimized, and the search domain of the arrangement positions of the wind turbine is continuous.

Description

BRIEF DESCRIPTION OF DRAWINGS

[0024] FIG. 1 is a flow chart of the method of the present disclosure;

[0025] FIG. 2 is a result of layout optimization based on grid division according to an embodiment of the present disclosure;

[0026] FIG. 3 is a result of layout optimization based on continuous arrangement positions according to an embodiment of the present disclosure;

[0027] FIG. 4 is a layout result of a wind farm based on grid division without considering dispatching strategy according to an embodiment.

DESCRIPTION OF EMBODIMENTS

[0028] The implementation of the present disclosure will be described in detail with reference to the following drawings:

EMBODIMENTS

[0029] In this embodiment, the turbine layout optimization in the wind farm construction stage is carried out for a wind farm. Wind turbines with a rated power of 1.5 MW and a diameter D of the impeller face of 77 meters are assembled in a square wind farm area with a side length of 1,232 meters. It is assumed that the abscissa of the wind farm area is 0-1232 m and the ordinate is 0-1232 m. The wind turbine layout optimization research is carried out with a wind speed of 8 m/s in a westerly wind direction, and the feasible domain for the layout is the whole wind farm. In the wind turbine layout optimization framework combining with the dispatching strategy for the wind farm, the first optimization goal is to minimize the cost of energy of the wind farm on the premise of ensuring that the safe distance is 4 D based on grid division; the second optimization goal is to minimize the cost of energy of the wind farm with a penalty function is introduced to ensure the safety distance is 4 D. The flowchart of the whole process is shown in FIG. 1. The implementation steps are as follows:

[0030] 1) Obtaining terrain data, wind speed and direction measurement data and meteorological parameters of a wind farm, and performing wind resource analysis.

[0031] 2) According to a result of the wind resource analysis and a requirement of a safe distance between wind turbines, dividing the wind farm into grids. A total number of the grids is a maximum number of the wind turbines that can be installed, and taking centers of the grids as optional positions of the wind turbines to obtain a set of feasible positions for turbine installation.

[0032] 3) Taking the number of the wind turbines to be installed in the wind farm, an arrangement position of each wind turbine, and an axial induction factor of each wind turbine in the dispatching strategy for the wind farm collectively as optimization variables of the turbines layout problem. Applying a greedy algorithm to optimize the number of the wind turbine in a feasible region of the optimization variables. Taking the installation number n of wind turbines with a lowest cost of energy CoE.sub.n as an optimized installation number n.sup.opt of the wind turbines, and obtaining a preliminary wind turbine arrangement position optimization solution corresponding to n.sup.opt.

[0033] 4) Further optimizing the arrangement positions of the n.sup.opt wind turbines obtained in step 3), removing restriction of the grids and introducing a penalty function to ensure the safe distance between the wind turbines. Optimizing the positions of the n.sup.opt wind turbines and the corresponding axial induction factors by using particle swarm optimization, so as to further reduce the cost of energy CoE.sub.n.sub.opt and obtain a turbine arrangement position optimization solution in the continuous spatial positions within the wind farm.

[0034] The installation number, the arrangement positions and the axial induction factors of wind turbines are optimized, and a two-step layout optimization method is used.

[0035] In a first step, the greedy algorithm is used to solve an optimization problem of optimizing the installation number of the wind turbines. The installation number of the wind turbines in the area of the wind farm, the arrangement position of each wind turbine and the axial induction factor thereof are collectively taken as optimization variables. A search domain is a feasible set of the wind farm composed of the maximum number of wind turbines that can be installed, grid-based discrete arrangement positions of the wind turbine and an artificially discretized set of the axial induction factors. An optimization objective function is a minimum of the cost of energy CoE. A corresponding individual fitness is composed of the cost of energy CoE.sub.n, and the smaller the value of the individual fitness, the better the individual fitness. A calculation formula of the fitness fitness1 is:

[00003]$\mathrm{fitness}1=\mathrm{CoE}=\frac{{\mathrm{cost}}_n}{T_{\mathrm{life}}.Math.{\mathrm{AEP}}_n}$

[0036] where T.sub.life is an effective life of wind turbines; cost.sub.a is a cost of the turbine location layout optimization solution corresponding to the installation number n, and AEP is an annual average energy production of the wind farm corresponding to the optimization solution.

[0037] In a second step, on the basis of the preliminary wind turbine arrangement position optimization solution obtained by optimizing the number of wind turbines, the particle swarm optimization algorithm is used for further turbine position optimization. The restriction of the grids is removed, and a penalty function is introduced into a fitness function to ensure the safe distance between wind turbines. The positions of the n.sup.opt wind turbine and the axial induction factors are optimized, and the search domain of turbine arrangement positions is continuous. The individual fitness consists of two parts, one is the cost of energy and the other is a function on distances between the wind turbines, the smaller the value of the individual fitness, the better the individual fitness. The calculation formula of the individual fitness fitness2 is:

fitness2=CoE+J.Math.Dis

[0038] where CoE is the cost of energy corresponding to a wind turbine layout solution combining with the dispatching strategy for the wind farm. The wind turbine layout solution includes two parts: the installation number of the wind turbines and arrangement positions of the wind turbines. J is a set penalty function coefficient, and Dis is a function that is set to ensure the safe distance between wind turbines, an expression of which is:

[00004]$Dis=\underset{i=1}{\overset{n^{\mathrm{opt}}}{.Math.}}\underset{\begin{array}{c}j=1\\ j\ne i\end{array}}{\overset{n^{\mathrm{opt}}}{.Math.}}\max \left\{0,D_s^2-d_{i,j}^2\right\}$

[0039] where d.sub.i,j is the straight-line distance between a wind turbine i and a wind turbine j; n.sup.opt is the installation number of the wind turbines after optimization in the first step; D.sub.s is the safe distance between the wind turbines, which is set to 4 times the rotor diameter of the wind turbine in this embodiment, i. e 4 D.

[0040] The above wind turbine layout optimization method combining with a dispatching strategy for a wind farm mainly includes grid division of the wind farm, application of the greedy algorithm to optimize the installation number of wind turbines in grid-based wind farm, removal of restriction of the grids, introduction of a penalty function, and application of the particle swarm optimization to further optimize the arrangement positions of wind turbines. In the embodiment, the wind turbine layout optimization calculation is carried out according to the flowchart shown in FIG. 1. FIG. 2 is the layout result of the wind turbine layout optimization method based on grid division combining with the dispatching strategy according to the present disclosure, namely, the layout result of the first step optimization. FIG. 3 is the layout result obtained by the wind turbine layout optimization method based on the continuous arrangement positions of wind turbines combining with the dispatching strategy according to the present disclosure, namely, the layout result of the second step optimization.

[0041] FIG. 4 is the layout result of the wind turbine layout optimization method based on grid division without considering dispatching strategy. In this embodiment, the safe distance of the wind turbines is 4 D, and there are 16 grid centers where wind turbines can be placed. Table 1 shows the comparison of the results between the wind turbine layout method based on grid division without considering the dispatching strategy and the wind turbine layout optimization method combining with the dispatching strategy for the wind farm. It can be seen that the wind turbine layout optimization method combining with the dispatching strategy for the wind farm of the present disclosure is much better than the wind turbine layout method based on grid division without considering the dispatching strategy, in terms of annual energy production and cost of energy. The method of the present disclosure combines the dispatching strategy of maximizing the wind farm capacity with the layout optimization of wind turbine, effectively reduces the wake effect and the wind power capacity cost, which is more in line with the actual demands of the wind farm.

TABLE-US-00001 TABLE 1 Annual energy Wind motor layout Optimization production Cost of energy Installation method (MWh/year) (RMB/kWh) number Wind turbine layout Optimization 3.927 * 10.sup.3 1.931 9 method based on grid division without considering dispatching strategy Wind turbine layout optimization 4.498 * 10.sup.3 1.087 12 method based on grid division combining with dispatching strategy for wind farm (first step of optimization) Wind turbine layout optimization 4.608 * 10.sup.3 1.061 12 method based on continuous arrangement positions of wind turbines combining with dispatching strategy for the wind farm (second step of optimization)

[0042] The above embodiment is only a preferred embodiment of one or more embodiments of this description, and it is not intended to limit one or more embodiments of this description. Any modification, equivalent substitution, improvement and the like made within the spirit and principle of one or more embodiments of this description should be included in the scope of protection of one or more embodiments of this description.