Wind power plant yaw control optimization method based on sequential optimization algorithm

Abstract

The invention discloses a wind power plant yaw control optimization method based on a sequential optimization algorithm, and the method comprises the steps: obtaining the coordinates of all fans at a current inflow wind direction angle, queuing all fans after the coordinates are converted in sequence, and calculating the yaw angle range of each cycle according to the actual demands; (n-1) times of outer layer circulation are set for the first (n-1) fans in the preliminary optimization process, in each outer layer circulation, yaw angle traversal value taking circulation is conducted on a certain fan according to the obtained yaw range and the uniform step length, and then the optimal yaw combination and the corresponding power of the wind power plant are obtained; and then optimization is carried out to obtain the optimal yaw angle of each fan in the whole field. According to the method, the wake flow secondary deflection effect in the wind power plant and the influence effect of the inflow turbulence intensity on the wake flow are considered, the improved Gaussian wake flow model is adopted to comprehensively consider calculation of wake flow loss in the three-dimensional direction, the calculation result is more accurate, and the final control optimization effect can also conform to the actual engineering situation.

Description

Technical Field

[0001] The invention relates to wind power generation and wind farm control technology, and in particular to a wind farm yaw control optimization method based on a sequential optimization algorithm. Background Art

[0002] Yaw control is a wind turbine control method. Wind turbines in wind farms use yaw control to adjust the angle of the wind rotor to the wind, thereby deflecting the direction of their wake to reduce the area of ​​the rear-row wind turbines exposed to their wake, reduce the impact of the wake effect, and increase the power generation of the rear-row wind turbines. The above method reduces the power generation of some front-row wind turbines in exchange for the effect of increasing the total power generation of the entire farm.

[0003] The existing technology explores the wake influence relationship of each wind turbine in the wind farm by establishing a wake influence relationship matrix. The specific means are: calculating the relative angle between the front and rear wind turbines, combining the main parameters such as the inflow wind direction, the expansion rate of the wake along the flow direction, and the centerline of the wake, solving and analyzing the change law of the degree of influence of the wake of the front wind turbine on the rear wind turbine with the yaw angle of the front wind turbine; then based on the gradient descent method (sequential least squares programming) and nonlinear and non-convex optimization methods (such as genetic algorithms, simulated annealing methods), with the goal of maximizing the output power of the entire field, the optimal yaw angle of each wind turbine is determined by performing a wide search in the yaw angle space. However, this technology is only applicable to the solution calculation of certain models and specific inflow conditions. It not only lacks the consideration of the influence of turbulence intensity on the wake, but also does not take into account the secondary deflection effect of the wake caused by the superposition of wakes. It has poor generalization and cannot guarantee the calculation accuracy. At the same time, the time cost required for optimization increases exponentially with the number of wind turbines, which is not conducive to the formulation of the optimal yaw control strategy for large wind farms.

[0004] Although the above technical means can improve the power generation of wind farms by optimizing the yaw angle of wind turbines, they have defects such as poor calculation accuracy, poor generalization, and high optimization cost. Especially in multi-dimensional and large-scale wind farms, the calculation time increases exponentially with the number of wind turbines, and unnecessary yaw with low gain or the situation where a smaller yaw angle is not preferred may occur. In addition, the existing two-dimensional wake model can only analytically solve the wind turbine wake in the flow direction and span direction, ignoring the impact of vertical wind shear on the wake loss. Summary of the invention

[0005] Purpose of the invention: The purpose of the present invention is to address the deficiencies in the prior art and to provide a method for optimizing the yaw control of a wind farm based on a sequential optimization algorithm. The present invention helps to determine the yaw angle of each wind turbine in a wind farm quickly and efficiently from a multi-factor collaborative perspective, and reduce the computational cost of existing solutions. At the same time, it avoids low-benefit yaw or unnecessary increase in yaw, thereby maximizing the power generation of the entire farm.

[0006] Technical solution: A wind farm yaw control optimization method based on a sequential optimization algorithm of the present invention comprises the following steps:

[0007] Step 1: Collect the current inflow wind direction angle α and the coordinates of each wind turbine in the true north coordinate system, and use the coordinate transformation formula of the plane coordinate system rotation to solve the counterclockwise rotation around the origin of the true north coordinate system. The corresponding coordinates of each wind turbine in the new coordinate system are: Then, the horizontal coordinates in the new coordinate system are sorted from small to large. The larger the coordinate value, the closer to the upstream. The wind turbines are sorted and named in order from upstream to downstream. The expression is as follows: (WT1, WT2, WT3…WTn-1, WTn);

[0008] n is the total number of wind turbines, and the original coordinates of the i-th wind turbine are (x i ,y i ), the corresponding coordinate of the fan in the new coordinate system is (x' i ,y' i );

[0009] Step 2: Select the upper limit β of the required traversal yaw angle according to actual engineering requirements min , lower limit β max and the number of yaw angles required to be traversed NYaw S , and then get the yaw angle that needs to be traversed in each inner loop:

[0010]

[0011] Where, j is the number of times the yaw angle is traversed;

[0012] Step 3: Preliminary optimization: First, arrange the i-th wind turbine WT in the upstream and downstream order in step 1. i Set as outer cycle, and perform n-1 outer cycles; in each outer cycle, the fan WTi is [β min ,β max ] is the yaw range, with a uniform step length Perform an inner loop to traverse the yaw angle value. In each inner loop, the fan WT i The yaw angle at the jth traversal is β j :

[0013]

[0014] Among them, each round of inner circulation uses the improved Gaussian wake model to calculate the wake loss value of each wind turbine downstream and the wake loss value of the entire field (considering the secondary deflection effect of the wake loss after the wake loss is superimposed), superimposed on the inflow, and the flow field velocity of the entire field is obtained. The WT of the i-th wind turbine at the current yaw angle is calculated.i The inflow wind speed of each downstream wind turbine is used to calculate the total power generation of the entire site according to the actual power curves of all n wind turbines;

[0015] Setting P opt The initial value of is the total wind farm power when all wind turbines in the wind farm are facing the wind. When the i-th wind turbine traverses the yaw angle, if the total wind farm power P corresponding to the current yaw angle is ij >P opt , then let P opt =P ij , and record the current total wind farm power P ij Corresponding to the fan WT in the traversal cycle i The yaw angle β opt i =β j ;

[0016] Here, P opt Indicates the maximum total power generation of the wind farm in each inner cycle; β opt i Indicates the wind turbine yaw angle corresponding to the maximum wind farm total power generation in each inner cycle;

[0017] Otherwise, it will not be recorded and the next round of loop traversal will be continued;

[0018] After completing the i-th inner cycle, the yaw angles of the first i fans should be fixed and the i+1-th inner cycle should be entered until i=n-1, completing the inner and outer cycles;

[0019] After traversing the inner and outer loops, the optimal yaw combination of the wind farm can be obtained under the condition of limited yaw angle selection;

[0020] Step 4: Refine the optimization and set the outer circulation to the fans WTi arranged in the upstream and downstream order and set the inner circulation to WT i The yaw angle of the wind turbine, which is NYaw R And is an odd number, the step length is dβ j , whose inner loop value is β j ;

[0021] In the i-th inner cycle, calculate the i-th fan in β j Total wind farm power under yaw angle P ij , if the total wind power P corresponding to the current yaw angle ij >P opt , and β min ≤β j ≤β max And let P opt =P ij , and record the current total wind farm power Pij Corresponding to the fan WT in the traversal cycle i The yaw angle β opt i =β j ;

[0022] Otherwise, it will not be recorded and the next round of loop traversal will be continued;

[0023] After completing the i-th inner cycle, the yaw angles of the first i fans should be fixed and the i+1-th inner cycle should be entered until i=n-1, completing the inner and outer cycles;

[0024] After traversing the inner and outer layers, the optimal yaw angle of each wind turbine in the entire field can be obtained.

[0025] After n wind turbines have been arranged in order from upstream to downstream, the last (nth) wind turbine is the most downstream wind turbine, and the wake of this wind turbine will not be transmitted to any other wind turbine in the field. Therefore, the present invention only traverses the yaw angles of the first n-1 wind turbines among the n wind turbines. The outer loop and the inner loop refer to the nesting of the inner and outer loops in the process of traversing and optimizing the yaw angles of each wind turbine. The outer loop is a loop for wind turbine numbering, such as WTi (i=1, 2, 3, ..., n-1), which is looped n-1 times to cover the first n-1 wind turbines in the field; the inner loop is in each outer loop (wind turbine numbering layer), for wind turbine WTi, with [βmin, βmax] as the yaw range and a uniform step length. Perform a loop to traverse and obtain the value of the yaw angle.

[0026] Furthermore, in step 1, the corresponding wind turbine coordinates (x' i ,y' i ) is calculated as follows:

[0027] x i ′=x i sinα-y i cosα

[0028] y′ i =x i cosα+y i sinα.

[0029] Furthermore, during the preliminary optimization process in step 3,

[0030] Each inner cycle fan WT i The yaw angle at the jth traversal is β j :

[0031]

[0032] The improved Gaussian wake model used is used to characterize the wake velocity loss, wake deflection, additional turbulence intensity, wake secondary turning effect and wake superposition.

[0033]

[0034] In the above formula, It represents the ratio of the wake velocity loss to the inflow wind speed, Δu is the wind speed loss, u ∞ is the speed of free flow (no fan); C T It represents the thrust coefficient of the wind turbine at the current inflow wind speed, which can be determined by looking up the table; d is the diameter of the wind rotor; γ represents the yaw angle of the wind turbine, that is, the angle (acute angle) between the normal direction of the wind turbine rotor and the inflow wind vector; cosγ is the cosine value of the yaw angle, which is used to calculate the wake loss of the yaw wind turbine; δ is the wake deflection distance; z h Indicates the hub height; σ y and σ z Represent the standard deviation of the wake loss size of the Gaussian distribution in the y and z directions, respectively, and are determined by the following formula;

[0035]

[0036] In the above formula, k y and k z are the expansion coefficients of the wake in the y and z directions, I is the turbulence intensity of the inflow, α * and β * They are the partial coefficients respectively, and their values ​​are usually 2.32 and 0.154.

[0037] To simulate the secondary turning effect of the wake, the vortex in the wake is considered as a linear viscous anti-rotating Lamb-Oxen vortex, and the reverse vortex circulation of the wake is calculated based on the Biot-Saffar theorem:

[0038]

[0039] In this way, the yaw axial induction factor a of the wind turbine and the free flow velocity U ∞ , the rotor diameter D and the tip speed ratio λ to determine the velocity circulation in its wake;

[0040] Since the main direction of the wake secondary turning is the spanwise direction, it is very necessary to determine the specific value of the spanwise velocity component induced by the circulation. According to the Lamb-Ossen vortex circulation induction theorem in the curl model, the spanwise induced velocity can be calculated as follows:

[0041]

[0042] In the above formula, y0 and z hare the spanwise position and height of the hub, respectively, representing the radius of the vortex core, which is related to the viscosity and is usually about 0.3D;

[0043] After obtaining the spanwise velocity component V, its influence on the “secondary steering effect” of the wake is approximated as the “equivalent yaw angle γ eff " is linearly superimposed on the yaw angle of the downstream wind turbine, γ eff The calculation formula is as follows:

[0044]

[0045] Furthermore, in the refinement optimization process of step 4, the step length of the inner loop is dβ j The calculation formula is as follows:

[0046]

[0047] Inner circulation fan WT i The yaw angle value at the jth traversal is:

[0048]

[0049] Beneficial effects: The present invention uses a strategy of yaw each wind turbine in turn, which can greatly reduce the calculation cost, improve the calculation speed and efficiency without weakening the power improvement effect, and can select a smaller yaw angle or no yaw when the yaw improvement effect is extremely unobvious. Compared with the prior art, the present invention has the following advantages:

[0050] (1) The present invention reduces the calculation cost and shortens the calculation time without weakening the yaw optimization effect.

[0051] (2) The present invention can effectively avoid unnecessary yaw, or select a larger yaw angle under the same optimization effect.

[0052] (3) The present invention takes into account the secondary deflection effect of the wake caused by the interaction of wakes within the wind farm, as well as the influence of the inflow turbulence intensity on the wake. An improved Gaussian wake model is used to comprehensively consider the calculation of wake losses in three dimensions. This has stronger applicability in actual engineering applications, and the calculation results are more accurate. The final control optimization effect will also be in line with the actual engineering situation. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a schematic diagram of the overall process of the present invention;

[0054] Figure 2 Schematic diagram of the inflow direction and the fan in the embodiment;

[0055] Figure 3Schematic diagram of the yaw angles of different wind turbines at different stages in the embodiment. DETAILED DESCRIPTION

[0056] The technical solution of the present invention is described in detail below, but the protection scope of the present invention is not limited to the embodiments.

[0057] like Figure 1 As shown, the wind farm yaw control optimization method based on the sequential optimization algorithm of the present invention comprises the following steps:

[0058] Step 1: Collect the current inflow wind direction angle α and the coordinates of each wind turbine in the true north coordinate system, and use the coordinate transformation formula of the plane coordinate system rotation to solve the counterclockwise rotation around the origin of the true north coordinate system. The corresponding coordinates of each wind turbine in the new coordinate system are: Then, the horizontal coordinates in the new coordinate system are sorted from small to large. The larger the coordinate value, the closer to the upstream. The wind turbines are sorted and named in order from upstream to downstream. The expression is as follows: (WT1, WT2, WT3…WTn-1, WTn);

[0059] n is the total number of wind turbines, and the original coordinates of the i-th wind turbine are (x i ,y i ), the corresponding coordinate of the fan in the new coordinate system is (x' i ,y' i );

[0060] Step 2: Select the upper limit β of the required traversal yaw angle according to actual engineering requirements min , lower limit β max and the number of yaw angles required to be traversed NYaw S , and then get the yaw angle that needs to be traversed in each inner loop:

[0061]

[0062] Where, j is the number of times the yaw angle is traversed;

[0063] Step 3: Preliminary optimization: First, arrange the i-th wind turbine WT in the upstream and downstream order in step 1. i Set as outer cycle, and perform n-1 outer cycles; in each outer cycle, the fan WTi is [β min ,β max ] is the yaw range, with a uniform step length Perform an inner loop to traverse the yaw angle value. In each inner loop, the fan WT i The yaw angle at the jth traversal is β j :

[0064]

[0065] In each inner layer circulation, the downstream wake loss value of each wind turbine is calculated by improving the Gaussian wake model, superimposed on the inflow, and the velocity of the whole flow field is obtained. The WT of the i-th wind turbine at the current yaw angle is calculated. i The inflow wind speed of each downstream wind turbine is used to calculate the total power generation of the entire site according to the actual power curves of all n wind turbines;

[0066] Setting P opt The initial value of is the total wind farm power when all wind turbines in the wind farm are facing the wind. When the i-th wind turbine traverses the yaw angle, if the total wind farm power P corresponding to the current yaw angle is ij >P opt , then let P opt =P ij , and record the current total wind farm power P ij Corresponding to the fan WT in the traversal cycle i The yaw angle β opt i =β j ; Here, P opt Indicates the maximum total power generation of the wind farm in each inner cycle; β opt i Indicates the wind turbine yaw angle corresponding to the maximum wind farm total power generation in each inner cycle;

[0067] Otherwise, it will not be recorded and the next round of loop traversal will be continued;

[0068] After completing the i-th inner cycle, the yaw angles of the first i fans should be fixed and the i+1-th inner cycle should be entered until i=n-1, completing the inner and outer cycles;

[0069] After traversing the inner and outer loops, the optimal yaw combination of the wind farm can be obtained under the condition of limited yaw angle selection;

[0070] For example, when all wind turbines in a wind farm are facing the wind (the yaw angles of all wind turbines are 0°), the total wind farm power (P opt The initial value of the wind turbine is 100kW. i , which is the first outer cycle) When the first inner cycle is performed, the yaw angle is changed to 10° according to the step size (the yaw angles of other wind turbines are still 0), and the total power generation of the whole field is calculated to be 150kW, which is greater than its initial value of 100kW. At this time, P opt will be updated to 150kW, β opt iThe value will be updated to 10°, and then enter the second inner loop; if in the second inner loop, the total power generated by changing the yaw angle of WT1 by 20° according to the step size is 130kW, which is less than the 100kW obtained in the previous loop, then P opt and β opt i Keep 150kW and 10° unchanged respectively, without being updated, and then enter the third inner loop.

[0071] When WT1 completes all its inner loops according to the upper and lower limits and step length of the yaw angle, it enters the second outer loop, that is, WT2 will start to traverse its yaw angle and continuously calculate the full-field power to perform P opt and β opt i The assignment is updated.

[0072] Step 4: Refine the optimization and set the outer circulation to the fans WTi arranged in the upstream and downstream order and set the inner circulation to WT i The yaw angle of the wind turbine, which is NYaw R And is an odd number, the step length is dβ j , whose inner loop value is β j ;

[0073] In the i-th inner cycle, calculate the i-th fan in β j Total wind farm power under yaw angle P ij , if the total wind power P corresponding to the current yaw angle ij >P opt , and β min ≤β j ≤β max And let P opt =P ij , and record the current total wind farm power P ij Corresponding to the fan WT in the traversal cycle i The yaw angle β opt i =β j ;

[0074] Otherwise, it will not be recorded and the next round of loop traversal will be continued;

[0075] After completing the i-th inner cycle, the yaw angles of the first i fans should be fixed and the i+1-th inner cycle should be entered until i=n-1, completing the inner and outer cycles;

[0076] After traversing the inner and outer layers, the optimal yaw angle of each wind turbine in the entire field can be obtained.

[0077] In the above step 1, the corresponding wind turbine coordinates (x' i ,y' i ) is calculated as follows:

[0078] x i ′=x i sinα-y i cosα

[0079] y′ i =x i cosα+y i sinα.

[0080] During the preliminary optimization process of step 3 of this embodiment,

[0081] Each inner cycle fan WT i The yaw angle at the jth traversal is β j :

[0082]

[0083] The improved Gaussian wake model expression used is as follows:

[0084]

[0085] In the above formula, Indicates the ratio of the wake velocity loss to the inflow wind speed; C T It represents the thrust coefficient of the wind turbine at the current inflow wind speed, which can be determined by looking up the table; d is the diameter of the wind rotor; γ represents the yaw angle of the wind turbine; δ is the wake deflection distance; z h Indicates the hub height; σ y and σ z Represent the standard deviation of the wake loss size of the Gaussian distribution in the y and z directions, respectively, and are determined by the following formula;

[0086]

[0087] In the above formula, k y and k z are the expansion coefficients of the wake in the y and z directions, I is the turbulence intensity of the inflow, α * and β * They are the sub-item coefficients, usually taking values ​​of 2.32 and 0.154;

[0088] To simulate the secondary turning effect of the wake, the vortex in the wake is considered as a linear viscous anti-rotating Lamb-Oxen vortex, and the reverse vortex circulation of the wake is calculated based on the Biot-Saffar theorem:

[0089]

[0090] In this way, the yaw axial induction factor a of the wind turbine and the free flow velocity U ∞ , the rotor diameter D and the tip speed ratio λ to determine the velocity circulation in its wake;

[0091] Since the main direction of the wake secondary turning is the spanwise direction, it is very necessary to determine the specific value of the spanwise velocity component induced by the circulation. According to the Lamb-Ossen vortex circulation induction theorem in the curl model, the spanwise induced velocity can be calculated as follows:

[0092]

[0093] In the above formula, y0 and z h are the spanwise position and height of the hub, respectively, representing the radius of the vortex core, which is related to the viscosity and is usually about 0.3D;

[0094] After obtaining the spanwise velocity component V, its influence on the “secondary steering effect” of the wake is approximated as the “equivalent yaw angle γ eff " is linearly superimposed on the yaw angle of the downstream wind turbine, γ eff The calculation formula is as follows:

[0095]

[0096] In the refinement optimization process of step 4 of this embodiment, the step length of the inner loop is dβ j The calculation formula is as follows:

[0097]

[0098] Inner circulation fan WT i The yaw angle value at the jth traversal is:

[0099]

[0100] Example 1

[0101] In order to have a more specific and in-depth understanding of the yaw optimization method of the technical solution of the present invention, the following is a pseudo code of the optimization algorithm.

[0102]

[0103] In the above pseudo code, the initial yaw angle of all wind turbines in the wind farm is first set to 0°, that is, all wind turbines are facing the wind, and the total power output power in the initial state is calculated. int .

[0104] After the initial optimization phase begins, set up the outer loop (the loop for fan numbering) to traverse the first N fans after upstream and downstream sorting. turbine-1 fan; assign yaw to yaw test , which is used for subsequent traversal to find the optimal yaw angle; and setting an inner loop (loop for the yaw angle). In this embodiment, the yaw range [βmin, βmax] of the yaw angle in the inner loop is [-20°, 20°], the step size d is 10°, and the yaw angle traversed in the inner loop is assigned to WT i (i-th wind turbine), and then calculate the total power of the i-th wind turbine after yaw test , and compare power test The power before yaw (before yaw here refers to the current inner cycle yaw, not the initial state where all wind turbines have a yaw angle of 0°) is larger than the power before yaw. test The power is greater than 1.01 times, that is, the total power is increased by 1% due to the yaw of the i-th wind turbine compared with the power before the yaw. Then the increased power power test Assign a value to power for full-field power comparison in the next inner cycle, and record the yaw angle of the i-th wind turbine in the inner cycle of the wheel until the end of the preliminary optimization stage;

[0105] After the refinement optimization phase begins, the range and step length of the yaw angle are refined based on the preliminary optimization results. For example, the step length of the yaw angle traversed in the preliminary optimization phase is 10°, and it is refined to 2.5° in the refinement optimization phase; the range of the yaw angle is also the result of the preliminary optimization phase plus or minus half the step length (half of 10°), that is, 5°, until the refinement optimization phase ends. Then the optimal yaw angle combination of the entire wind farm is obtained, as well as the relative increase rate of the entire field power under the optimal yaw combination compared with the initial state of the entire field power.

[0106] This embodiment shows that the technical solution of the present invention can traverse and optimize the optimal yaw combination of all wind turbines in the field, and ensure that the power of the entire field after optimization is at least 1% higher than that of the non-yaw situation. If both the preliminary optimization and the detailed optimization are effectively optimized, the power of the entire field will be increased by at least 2.01% (the value of 1.01*1.01-1). Such an optimization algorithm can select a smaller yaw angle or no yaw when the yaw improvement effect is extremely insignificant, thereby avoiding the inefficient yaw situation.

[0107] Example 2

[0108] This embodiment applies the technical solution of the present invention to a specific wind farm, and the content is as follows:

[0109] According to the current inflow wind direction, determine the upstream and downstream order of the fans and sort them in order from small to large;

[0110] According to the specific conditions and needs of the wind farm, select the yaw angle limit of each wind turbine and the number of yaw angles to be used in each stage (NYaw in the initial stage). S ; The optimization stage is NYaw R );

[0111] Preliminary optimization: All wind turbines in the field (assuming there are n wind turbines) are numbered from 1 to n in the above order. The wind turbines that need to be optimized are the first (n-1) wind turbines. Each of the above (n-1) wind turbines traverses NYaw S Therefore, in this optimization method, there are two nested loops. Each inner loop includes a certain wind turbine in NYaw S One of the yaw angles in .

[0112] In each inner cycle, the wind speed (including wake) value of the entire field is calculated based on the improved Gaussian wake model, and then the total power generation of the entire field is calculated in combination with the actual power curves of the n wind turbines mentioned above, and the yaw angle of the above wind turbine when the power of the entire field is the maximum in the inner cycle is recorded as the initial optimized yaw angle of the wind turbine. Keep the yaw angle of the above wind turbine unchanged, enter the next inner cycle of the downstream wind turbine, and find the initial optimized yaw angles of the (n-1) wind turbines in the entire field one by one. End the initial optimization.

[0113] Refined optimization: Based on the preliminary optimized wind turbine yaw combination, the NYaw is offset with the above preliminary optimized yaw angle as the center. R According to the above preliminary optimization process, the yaw angles of the (n-1) wind turbines after detailed optimization are determined as the final optimization result. The detailed optimization is ended.

[0114] Take five fans arranged in series as an example, n = 5, Figure 2 shown.

[0115] If the yaw angle limit of the above five wind turbines is initially optimized to [-30°, 30°], and NYaw S =5, then this optimization phase will make each wind turbine traverse each of the five yaw angles [-30°, -15°, 0°, 15°, 30°]. Then, the refinement optimization phase will consider the optimal angle found in the above preliminary optimization phase. If NYaw R =5, then taking the above optimal angle as the center and adding the offset range [-7.5°, -3.75°, 0°, 3.75°, 7.5°], the yaw angle explored by each wind turbine in this embodiment is as follows Figure 3 shown.

[0116] The present invention optimizes the yaw angle of each wind turbine in the entire field under different inflow conditions to maximize the power generation of the entire field. When NYaw=5 or NYaw=7, the total power of the wind farm under different wind conditions is slightly improved compared to the traditional sequential least squares planning method, but the calculation time is shortened to about 10% of the latter. This optimization algorithm is particularly effective in multi-dimensional large-scale wind farms. Therefore, the present application greatly shortens the calculation time without weakening the yaw effect, and takes into account the influence of the secondary deflection of the wake and the turbulence intensity on the optimization effect.

Claims

1. A wind farm yaw control optimization method based on a sequential optimization algorithm, characterized in that: The following steps are involved: Step 1: Collect the current inflow wind direction angle α and the coordinates of each wind turbine in the true north coordinate system, and then use the plane coordinate system rotation change coordinate transformation formula to solve the counterclockwise rotation around the origin of the true north coordinate system. The corresponding coordinates of each wind turbine in the new coordinate system: Then, sort the horizontal coordinates in the new coordinate system from small to large, and sort and name the wind turbines in order from upstream to downstream. The expression is as follows: (WT1, WT2, WT3…WTn-1, WTn); n is the total number of wind turbines, and the original coordinates of the i-th wind turbine are (x i ,y i ), the corresponding coordinate of the fan in the new coordinate system is (x' i ,y' i ); Step 2: Select the upper limit β of the required traversal yaw angle according to actual engineering requirements min , lower limit β max and the number of yaw angles required to be traversed NYaw S , and then get the yaw angle that needs to be traversed in each inner loop: Where, j is the number of times the yaw angle is traversed; Step 3: Preliminary optimization: First, arrange the i-th wind turbine WT in the upstream and downstream order in step 1. i Set as outer cycle, and perform n-1 outer cycles; in each outer cycle, the fan WTi is [β min ,β max ] is the yaw range, with a uniform step length Perform an inner loop to traverse the yaw angle value. In each inner loop, the fan WT i The yaw angle at the jth traversal is β j : i = [1, 2, ... n-1]; 1≤j≤NYaw S ,j∈N * ; Among them, each round of inner circulation uses the improved Gaussian wake model to calculate the wake loss value of each wind turbine downstream and the wake loss value of the entire field, superimposed on the inflow, to obtain the flow field velocity of the entire field, and calculate the WT of the i-th wind turbine at the current yaw angle. i The inflow wind speed of each downstream wind turbine is used to calculate the total power generation of the entire site according to the actual power curves of all n wind turbines; Setting P opt The initial value of is the total wind farm power when all wind turbines in the wind farm are facing the wind. When the i-th wind turbine traverses the yaw angle, if the total wind farm power P corresponding to the current yaw angle is ij >P opt , then let P opt =P ij , and record the current total wind farm power P ij Corresponding to the fan WT in the traversal cycle i The yaw angle β opti =β j ; Here, P opt Indicates the maximum total power generation of the wind farm in each inner cycle; β opti Indicates the wind turbine yaw angle corresponding to the maximum wind farm total power generation in each inner cycle; Otherwise, it will not be recorded and the next round of loop traversal will be continued; After completing the i-th inner cycle, the yaw angles of the first i fans should be fixed and the i+1-th inner cycle should be entered until i=n-1, completing the inner and outer cycles; After the inner and outer loops are traversed, the optimal yaw combination of the wind farm can be obtained under the condition of limited yaw angle selection; Step 4: Refine the optimization and set the outer circulation to the fans WTi arranged in the upstream and downstream order and set the inner circulation to WT i The yaw angle of the wind turbine, which is NYaw R And is an odd number, the step length is dβ j , whose inner loop value is β j ; In the i-th inner cycle, calculate the i-th fan in β j Total wind farm power under yaw angle P ij , if the total wind power P corresponding to the current yaw angle ij >P opt , and β min ≤β j ≤β max And let P opt =P ij , and record the current total wind farm power P ij Corresponding to the fan WT in the traversal cycle i The yaw angle β opti =β j ; Otherwise, it will not be recorded and the next round of loop traversal will be continued; After completing the i-th inner cycle, the yaw angles of the first i fans should be fixed and the i+1-th inner cycle should be entered until i=n-1, completing the inner and outer cycles; After traversing the inner and outer layers, the optimal yaw angle of each wind turbine in the entire field can be obtained.

2. The wind farm yaw control optimization method based on sequential optimization algorithm according to claim 1 is characterized in that: In step 1, the corresponding wind turbine coordinates (x' i ,y' i ) is calculated as follows: x i ′=x i sinα-y i cosα and' i =x i thing+and i sina.

3. The wind farm yaw control optimization method based on sequential optimization algorithm according to claim 1 is characterized in that: During the initial optimization process in step 3, Each inner cycle fan WT i The yaw angle at the jth traversal is β j : The improved Gaussian wake model used is used to characterize the wake velocity loss, wake deflection, additional turbulence intensity, wake secondary turning effect and wake superposition. In the above formula, It represents the ratio of the wake velocity loss to the inflow wind speed, Δu is the wind speed loss, u ∞ is the velocity of the free flow; C T It represents the thrust coefficient of the wind turbine at the current inflow wind speed, which can be determined by looking up the table; d is the diameter of the wind rotor; γ represents the yaw angle of the wind turbine; δ is the wake deflection distance; z h Indicates the hub height; σ y and σ z Represent the standard deviation of the wake loss size of the Gaussian distribution in the y and z directions, respectively, and are determined by the following formula; In the above formula, k y and k z are the expansion coefficients of the wake in the y and z directions, I is the turbulence intensity of the inflow, α * and β * are the sub-item coefficients respectively; To simulate the secondary turning effect of the wake, the vortex in the wake is considered as a linear viscous anti-rotating Lamb-Oxen vortex, and the reverse vortex circulation of the wake is calculated based on the Biot-Saffar theorem: The yaw axial induction factor a and free flow velocity U of the wind turbine ∞ , the rotor diameter D and the tip speed ratio λ to determine the velocity circulation in its wake; According to the Lamb-Ossen vortex circulation induction theorem in the curl model, the spanwise induced velocity is calculated as follows: In the above formula, y0 and z h are the spanwise position and height of the hub, representing the radius of the vortex core; After obtaining the spanwise velocity component V, its influence on the "secondary steering effect" of the wake is approximated as the "equivalent yaw angle γ eff " is linearly superimposed on the yaw angle of the downstream wind turbine, γ eff The calculation formula is as follows:

4. The wind farm yaw control optimization method based on sequential optimization algorithm according to claim 1 is characterized in that: During the refinement optimization in step 4, the step length of the inner loop is dβ j The calculation formula is as follows: Inner circulation fan WT i The yaw angle value at the jth traversal is:

Images