Multi-fan cooperative yaw optimization method for offshore wind plant

Abstract

The invention provides an offshore wind power plant multi-fan cooperative yaw optimization method, which comprises the following steps: establishing a single-machine double-Gaussian yaw wake flow model and a multi-machine yaw wake flow adaptive combination superposition model based on given wind speed, wind direction and fan parameters, and setting the models as the overall working condition of a wind power plant; relevant parameters of a particle swarm optimization algorithm are set according to the overall working condition, and iterative circulation is conducted on a fan yaw scheme through the particle swarm optimization algorithm with penalty factors inserted; and outputting the fan yaw scheme meeting the convergence condition as a multi-fan collaborative yaw optimization scheme. On the basis of a yaw fan double-Gaussian wake flow model with better accuracy and stronger robustness, a yaw multi-wake flow self-adaptive combination superposition model is provided, global collaborative optimization is performed on fan yaw by using a particle swarm optimization algorithm, and a yaw strategy which does not meet requirements is identified and eliminated by introducing a penalty factor, so that the yaw rate of the fan is improved. Therefore, the running smoothness of the fan is ensured, and the power generation efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of wind power generation, and particularly relates to a method for collaborative yaw optimization of multiple wind turbines in an offshore wind farm. Background Art

[0002] During the actual operation of a wind turbine, the wake wind field is affected by various complex meteorological conditions, among which the influence of yaw is the most significant. Yaw of the wind turbine, as an indispensable part of the wind power generation system, ensures that the wind turbine can operate efficiently and safely and effectively utilize wind energy resources. In an offshore wind farm, the yaw control of the wind turbine is crucial for improving power generation efficiency and ensuring the smooth operation of the wind turbine.

[0003] Current wind turbine yaw optimization methods are mostly based on the analytical wake model, and the wake superposition model is used to solve the situation where a certain wind turbine in a large wind farm is located in the mixed wake of multiple upstream wind turbines. The most famous and widely used analytical wake model is the Jensen model, and many scholars have improved it to make it have higher applicability and accuracy. The Gaussian wake model considering the thrust coefficient and environmental turbulence intensity proposed by Ishihara et al. also adds a correction term to make the model applicable to the near wake region and the far wake region. At the same time, the change of additional turbulence intensity in the wake region is considered in this model, and a double Gaussian model applicable to the calculation of turbulence intensity is proposed. In order to be able to use the analytical wake model for yaw optimization, some researchers have improved the Gaussian wake model so that it can calculate the offset distance of the wind turbine wake under yaw conditions, the wind speed loss in the wake region, and the additional turbulence intensity, etc. The commonly used wake superposition models are the geometric sum model, the linear superposition model, the energy conservation model, and the sum of squares model. At present, the discussion on the accuracy of each model has not been unified yet.

[0004] In the traditional single - environment wind turbine yaw optimization, only the single - wind - condition optimization under different incoming wind directions and wind speeds is often considered. Directly applying it to an offshore wind farm with variable wind conditions may cause the wind turbine to yaw too frequently, which will further exacerbate the fatigue load borne by the wind turbine and seriously affect the service life of the wind turbine structure. For example, when the wind speed of a certain wind turbine increases from 5 m / s to 6 m / s, the yaw angle may suddenly change from - 30° to + 30°; similarly, when the wind direction changes from north (N) to north - northwest (NNW), a similar sudden change in the yaw angle may also occur.

[0005] To overcome this limitation, the present invention proposes a method for collaborative yaw optimization of wind turbines in a complex offshore wind farm environment, which aims to ensure that the yaw of the wind turbine follows the incoming wind speed and wind direction smoothly.

[0006] It should be noted that the information disclosed in the above background art section is only used to strengthen the understanding of the background of the present invention and does not constitute any limitation to the present invention. Summary of the Invention

[0007] In view of the above-mentioned disadvantages of the prior art, the present invention provides a multi-wind turbine collaborative yaw optimization method for an offshore wind farm. By means of a yaw wind turbine double Gaussian wake model with better accuracy and stronger robustness, a yaw multi-wind turbine wake adaptive combined superposition model is proposed, and the particle swarm optimization algorithm is used to globally collaborate and optimize the yaw of the wind turbines. By introducing a penalty factor, the yaw strategies that do not meet the requirements are identified and excluded to ensure the smooth operation of the wind turbines and improve the power generation efficiency.

[0008] The present invention provides a multi-wind turbine collaborative yaw optimization method for an offshore wind farm, including:

[0009] Based on the given wind speed, wind direction and wind turbine parameters, a single-wind turbine double Gaussian yaw wake model and a multi-wind turbine yaw wake adaptive combined superposition model are established and set as the overall working conditions of the wind farm;

[0010] According to the overall working conditions, the relevant parameters of the particle swarm optimization algorithm are set, and the yaw scheme of the wind turbines is iteratively cycled through the particle swarm optimization algorithm with the penalty factor inserted;

[0011] The yaw scheme of the wind turbines that meets the convergence conditions is output as the multi-wind turbine collaborative yaw optimization scheme.

[0012] In an embodiment of the present invention, the wind turbine parameters further include a smoothness index that limits the fluctuation range of the wind turbine yaw angle.

[0013] In an embodiment of the present invention, the step of iteratively cycling the yaw scheme of the wind turbines through the particle swarm optimization algorithm with the penalty factor inserted includes:

[0014] An initial yaw angle scheme for multiple wind turbines in the wind farm is generated through the particle swarm optimization algorithm, and the total power generation of the wind farm is calculated;

[0015] The penalty factor is calculated according to the yaw angle scheme of the multiple wind turbines in the wind farm;

[0016] The total power generation and the penalty factor of multiple groups of yaw angle schemes for multiple wind turbines in the wind farm are calculated through iterative cycling;

[0017] The total power generation of the wind farm is multiplied by the penalty factor as the optimized power generation of the wind farm;

[0018] Based on maximizing the optimized power generation of the wind farm, the corresponding yaw angle scheme for multiple wind turbines in the wind farm is determined.

[0019] In an embodiment of the present invention, the step of calculating the penalty factor according to the yaw angle scheme of the multiple wind turbines in the wind farm includes:

[0020] Based on the yaw angle scheme of the multiple wind turbines in the wind farm, the yaw angle of each wind turbine is determined;

[0021] Determine whether the yaw angle of each wind turbine exceeds the yaw angle fluctuation range of the wind turbine;

[0022] If there is no yaw angle of the wind turbine exceeding, output the basic coefficient of the penalty factor;

[0023] If there is a yaw angle of the wind turbine exceeding, calculate the proportion of the yaw angle of each wind turbine exceeding the yaw angle fluctuation range of the wind turbine and the proportion of the number of wind turbines in the wind farm whose yaw angle exceeds the yaw angle fluctuation range.

[0024] In an embodiment of the present invention, when there is a yaw angle of the wind turbine exceeding, the parameter of the output penalty factor is less than its basic coefficient, and the larger the proportion of the yaw angle of each wind turbine exceeding the yaw angle fluctuation range of the wind turbine, the smaller the parameter of the output penalty factor, and the larger the proportion of the number of wind turbines in the wind farm whose yaw angle exceeds the yaw angle fluctuation range, the smaller the parameter of the output penalty factor.

[0025] In an embodiment of the present invention, the convergence condition includes maximizing the total power generation of the wind farm within the allowable yaw angle fluctuation range of the wind turbine.

[0026] In an embodiment of the present invention, the yaw single double Gaussian wake model is:

[0027]

[0028] k * = 0.11C t 1.07 I 0.2 (18)

[0029] ε = 0.23C t 0.25 I 0.17 (19)

[0030] a = 0.93C t -0.75 I 0.17 (20)

[0031] b = 0.42C t 0.6 I 0.2 (21)

[0032] c = 0.15C t -0.25 I -0.7 (22)

[0033] Wherein, U ∞denote the free stream; ΔU denotes the deficit velocity at the position (x, y, z) downstream of the wind turbine with the wind turbine as the origin; D denotes the diameter of the wind turbine blade; θ denotes the yaw angle of the wind turbine; I denotes the turbulence intensity; Ct denotes the thrust coefficient of the wind turbine; α k and α m take 0.7 and 0.65 respectively.

[0034] In an embodiment of the present invention, the yaw multi-wind turbine wake adaptive combined superposition model is:

[0035] V = ωV SS +(1 - ω)V GS (23)

[0036]

[0037] In the formula, V denotes the wake wind speed; V SS denotes the wake wind speed obtained based on the sum-of-squares superposition model; V GS denotes the wake wind speed obtained based on the geometric superposition model; x1 and x2 respectively denote the x / D values corresponding to the weight coefficients of 1 for the sum-of-squares superposition model and the geometric superposition model.

[0038] In an embodiment of the present invention, the execution steps of the particle swarm optimization algorithm include:

[0039] Initializing the particle swarm based on the set relevant parameters to generate the yaw scheme of each particle in the particle swarm;

[0040] Calculating the total power generation of the wind farm corresponding to the particle swarm by inserting the inertia factor and using it as the fitness of the particle swarm;

[0041] Comparing the current fitness of the particle swarm with the historical optimal value, and updating the yaw scheme of the wind turbine with the optimal global total power generation of the wind farm and the yaw scheme of the wind turbine with the optimal local total power generation of the wind farm;

[0042] Judging whether the yaw scheme of the wind turbine reaches the set convergence condition;

[0043] If not satisfied, continue to calculate the fitness of the particle swarm by inserting the inertia factor;

[0044] If satisfied, output the yaw scheme of the wind turbine with the optimal global total power generation of the wind farm as the multi-wind turbine collaborative yaw optimization scheme.

[0045] In an embodiment of the present invention, the step of calculating the total power generation of the wind farm corresponding to the particle swarm by inserting the inertia factor and using it as the fitness of the particle swarm includes providing an inertia factor that changes with time, which is used to calculate the velocity of the particle swarm and update the position of the particle swarm; where the inertia factor, the velocity of the particle swarm, and the position of the particle swarm are:

[0046]

[0047] v i = w p v i + c1r d (pb i - x i ) + c2r d (gb - x i ) (25)

[0048] x i = x i + v i (26)

[0049] Where: wp - inertia factor; w1 - initial inertia factor; w2 - final inertia factor; Niter - total number of iterations; j - current iteration step; vi - velocity of the i-th particle, m / s; c1 - individual learning factor; c2 - social learning factor; pbi - local optimal position of the particle, m; gb - global optimal position of the particle, m; xi - position of the particle, m; rd - random number between 0 and 1.

[0050] In an embodiment of the present invention, after the step of calculating the total power generation of the wind farm corresponding to the particle swarm by inserting the inertia factor, it further includes multiplying the total power generation of the wind farm by the penalty factor to obtain the optimized power generation of the wind farm based on the allowable fluctuation range of the yaw angle of the wind turbine, and using it as the fitness of the particle.

[0051] In an embodiment of the present invention, by setting the base coefficient of the penalty factor, the yaw angle schemes of the wind turbines corresponding to the total power generation of the wind farm are screened. In the multi-wind turbine collaborative yaw optimization scheme, the proportion of the total power generation factor of the wind farm is reduced, and the proportion of the yaw angle fluctuation range factor of the wind turbine is increased.

[0052] Advantages of the present invention: By setting the smoothness index, the yaw fluctuation range of the wind turbine under continuous wind direction and wind speed changes is effectively limited, thus ensuring the smoothness and stability of the wind turbine operation. At the same time, based on the yaw wind turbine double Gaussian wake model with better accuracy and stronger robustness, a yaw multi-wind turbine wake self-adaptive combined superposition model is proposed, and the particle swarm algorithm is used to globally optimize the yaw of the wind turbine, fully considering the mutual influence between the wind turbines in the wind farm, and significantly improving the power generation efficiency of the entire wind farm. In addition, by introducing the penalty factor mechanism, the yaw strategies that do not meet the requirements are identified and excluded, enhancing the robustness and reliability of the system. It not only optimizes the operation management of the wind farm, but also provides strong technical support for the sustainable development and efficient operation of the offshore wind power industry.

[0053] It should be understood that the above general description and the following detailed description are merely exemplary and explanatory, and do not limit the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The drawings herein are incorporated into and constitute a part of this specification, showing embodiments consistent with the present invention, and are used together with the specification to explain the principles of the present invention. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts. In the drawings:

[0055] Figure 1 is a flowchart of the multi - turbine collaborative yaw optimization method for the offshore wind farm of the present invention;

[0056] Figure 2 is the power curve and thrust coefficient curve of a wind turbine in an embodiment of the present invention;

[0057] Figure 3 is a schematic diagram of the positions of multiple wind turbines in an embodiment of the present invention;

[0058] Figure 4 is a schematic diagram of the yaw angle range of multiple wind turbines after single - turbine yaw optimization in an embodiment of the present invention;

[0059] Figure 5 is a schematic diagram of the yaw angle range of multiple wind turbines after the multi - turbine collaborative yaw optimization method is adopted in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0060] The following specific examples illustrate the embodiments of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, without conflict, the following embodiments and the features in the embodiments can be combined with each other. It should also be understood that the terms used in the embodiments of the present invention are for the purpose of describing specific specific implementation manners, rather than for limiting the protection scope of the present invention.

[0061] Please refer to Figures 1 to 5It should be noted that the structures, proportions, sizes, etc. shown in the drawings of this specification are only used to cooperate with the content disclosed in the specification for those familiar with this technology to understand and read, and are not used to limit the conditions under which the present invention can be implemented. Therefore, they do not have substantial technical significance. Any modification of the structure, change in the proportional relationship, or adjustment of the size, without affecting the effects that the present invention can produce and the purposes that can be achieved, should still fall within the scope covered by the technical content disclosed in the present invention. At the same time, the terms such as the positions and quantitative relationships cited in this specification are only for the convenience of clear narration and are not used to limit the scope of implementation of the present invention. The change or adjustment of their relative relationships, without substantial change in the technical content, should also be regarded as the scope within which the present invention can be implemented.

[0062] An offshore wind farm refers to a power generation site in a marine area where multiple wind turbines are centrally arranged to utilize the rich wind energy resources in the sea to convert wind energy into electrical energy. Yaw of the wind turbine refers to a key operation in the wind power generation system, which means that the nacelle of the wind turbine (including components such as the wind rotor) rotates around the vertical center line of the tower to adjust the orientation of the wind rotor. Its purpose is to keep the wind rotor of the wind turbine always aligned with the wind direction, capture wind energy to the maximum extent, and ensure the efficient, stable, and safe operation of the wind turbine. Wake effect is used in the field of wind power generation. The wake refers to an air flow area formed downstream of the wind turbine after the wind rotor obtains energy from the air flow during the operation of the wind turbine. Due to the blocking and stirring of the wind turbine blades, the flow velocity of this air flow is lower than the upstream incoming flow velocity, and at the same time, it is accompanied by characteristics such as an increase in turbulence intensity and changes in wind direction and wind speed distribution. The existence of the wake not only affects the incoming flow wind conditions of downstream wind turbines, reducing the wind energy that can be captured by them, but also may cause downstream wind turbines to bear more complex and variable loads, interfering with the coordinated operation and power generation efficiency of the wind turbine group in the entire wind farm.

[0063] In the traditional single - environment wind turbine yaw optimization, often only the single - wind - condition optimization under different incoming flow wind directions and wind speeds is considered. Directly applying it to an offshore wind farm with variable wind conditions may lead to overly frequent yaw of the wind turbine, thereby exacerbating the fatigue load borne by the wind turbine and seriously affecting the service life of the wind turbine structure. For example, when the wind speed of a certain wind turbine increases from 5 m / s to 6 m / s, the yaw angle may suddenly change from - 30° to + 30°; similarly, when the wind direction changes from north (N) to north - northwest (NNW), a similar sudden change in the yaw angle may also occur.

[0064] Please refer to Figure 1 , the present invention provides a multi - wind - turbine collaborative yaw optimization method for an offshore wind farm, including:

[0065] Establish a single - machine double - Gaussian yaw wake model and a multi - machine yaw wake adaptive combined superposition model based on the given wind speed, wind direction, and wind turbine parameters, and set them as the overall working conditions of the wind farm;

[0066] Set the relevant parameters of the particle swarm optimization algorithm according to the overall working conditions, and perform iterative loops on the wind turbine yaw scheme through the particle swarm optimization algorithm with an inserted penalty factor;

[0067] Output the wind turbine yaw scheme that meets the convergence condition as a multi-wind turbine collaborative yaw optimization scheme.

[0068] Specifically, in the embodiment of the present invention, the wind speed, wind direction and wind turbine parameters can be given according to meteorological measurements of the wind farm site and known wind turbine data. Its single-machine double-Gaussian yaw wake model can be modeled according to the given parameters above, and relevant formulas can be constructed in combination with relevant wind tunnel experiment data and simulation results, which can more accurately describe the characteristics of the wake when the wind turbine yaws. The multi-machine yaw wake adaptive combined superposition model combines the characteristics of the geometric superposition model and the sum-of-squares superposition model, and its weight coefficient can vary with the spatial position to achieve accurate calculation of the wake wind speed and ensure the smooth transition of the flow field. In this way, the overall working conditions of the wind farm are established. By proposing a new wind turbine yaw wake model and a multi-machine yaw wake adaptive combined superposition model, a basis is provided for the multi-wind turbine yaw optimization.

[0069] Furthermore, according to the set overall working conditions, configure the relevant parameters of the Particle Swarm Optimization (PSO) algorithm, such as the number of particles, the number of iteration steps, the inertia factor, the individual learning factor, the social learning factor, the limited range of the yaw angle, and the allowable fluctuation range, etc. During the optimization process, by calculating the total power generation of each particle and introducing a penalty factor (for example, the penalty factor can be whether the wind turbine is in the optimal power generation condition, please refer to the appendix Figure 2 , which shows the relationship between the power curve and the thrust coefficient curve of a certain wind turbine and the average wind speed. The actual wind speed received can be changed by adjusting the wind turbine yaw angle), the efficacy data corresponding to the wind turbine yaw scheme is processed and screened for the second time. The total power generation data corresponding to the wind turbine yaw scheme can be multiplied by the parameter of the penalty factor to obtain the corrected data considering the corresponding conditions of the penalty factor, so as to obtain the effective data under the condition of considering the corresponding conditions of the penalty factor. That is to say, the equipment condition cost corresponding to the penalty factor is compared with the power generation income of the wind turbine yaw angle selection for comparison and reference, and a more optimized wind turbine yaw scheme is comprehensively selected.

[0070] More specifically, the particle swarm optimization algorithm is used to evaluate the advantages and disadvantages of each particle (each particle corresponds to a set of wind turbine yaw angle schemes), and the particle swarm optimization algorithm is used to iteratively update the speed and position of the particles (the speed corresponds to the adjustment direction and step size of the yaw angle, and the position corresponds to the yaw angle scheme of each wind turbine in the wind farm) to find the optimal wind turbine yaw scheme.

[0071] Furthermore, please refer to the appendix Figure 2, which shows the power curve and thrust coefficient curve of a certain type of wind turbine adopted. The power curve represents the output power of the wind turbine at different wind speeds, with the horizontal axis being the wind speed and the vertical axis being the power; the thrust coefficient curve reflects the change of the thrust coefficient of the wind turbine at different wind speeds, with the horizontal axis also being the wind speed and the vertical axis being the thrust coefficient. Through the power curve, the power generation capacity of the wind turbine at various wind speeds can be understood, and its performance under different working conditions can be determined; the thrust coefficient curve helps to analyze the wind capture and utilization efficiency of the wind turbine, as well as the thrust on the wind turbine blades at different wind speeds, which is of great significance for the layout and operation control of wind turbines in a wind farm. Especially when conducting yaw optimization, it is necessary to consider the performance characteristics of the wind turbine at different wind speeds to ensure that the optimization strategy can effectively improve the power generation efficiency and ensure the safe and stable operation of the wind turbine.

[0072] In this way, when the particle swarm optimization algorithm reaches the convergence condition (for example, the maximum power generation that can be achieved under non-ideal working conditions represented by the penalty factor), the current wind turbine yaw scheme can be output as a multi-wind turbine collaborative yaw optimization scheme. It can ensure that the yaw angle of the wind turbine changes smoothly with the wind speed and wind direction, avoiding the wind turbine being in non-ideal working conditions represented by the penalty factor (such as mechanical wear and performance degradation caused by frequent large yaw adjustments to match the wind speed), so as to extend the service life of the wind turbine structure, improve the power generation efficiency and operation stability of the entire wind farm.

[0073] In one embodiment, the wind turbine parameters further include a smoothness index that limits the fluctuation range of the wind turbine yaw angle.

[0074] Specifically, by setting the smoothness index, it is used to limit the fluctuation range of the wind turbine yaw angle. During the operation of the wind farm, the wind turbine needs to adjust the yaw angle according to the changes of the wind direction and wind speed to maintain the best wind energy capture efficiency. However, if the yaw angle changes too violently, it will not only cause an increase in the fatigue load borne by the wind turbine, shortening its structural service life, but also may have a negative impact on the stable operation of the entire wind farm. Therefore, by setting an allowable yaw angle fluctuation range, it is ensured that the wind turbine can make a smooth transition when adjusting the yaw angle. Specifically, when the wind direction or wind speed changes, the yaw control system of the wind turbine will refer to the smoothness index and calculate a reasonable yaw angle adjustment range to avoid sudden changes in the yaw angle. For example, when the wind speed increases from 5 m / s to 6 m / s or the wind direction changes from north (N) to north-northwest (NNW), the smoothness index limits the sudden change of the wind turbine yaw angle from -30° to +30°, thereby reducing the risk of mechanical wear and performance degradation.

[0075] Similarly, the ride comfort index can also be incorporated into the particle swarm optimization algorithm as a constraint condition. During the iterative process of the particle swarm optimization algorithm, not only the total power generation of the wind farm needs to be considered, but also the yaw angle changes of each wind turbine must meet the requirements of the ride comfort index. In this way, the optimized yaw scheme of the wind turbine can ensure the power generation efficiency while extending the service life of the wind turbine and improving the operation stability and long-term economic benefits of the entire wind farm.

[0076] In one embodiment, the steps of iteratively cycling the yaw scheme of the wind turbine through the particle swarm optimization algorithm with inserted penalty factors include:

[0077] Generate an initial multi-wind turbine yaw angle scheme for the wind farm through the particle swarm optimization algorithm, and calculate the total power generation of the wind farm;

[0078] Calculate the penalty factor according to the multi-wind turbine yaw angle scheme of the wind farm;

[0079] Calculate the total power generation and penalty factor of multiple groups of multi-wind turbine yaw angle schemes of the wind farm through iterative cycling;

[0080] Multiply the total power generation of the wind farm by the penalty factor as the optimized power generation of the wind farm;

[0081] Determine the corresponding multi-wind turbine yaw angle scheme of the wind farm based on maximizing the optimized power generation of the wind farm.

[0082] Specifically, in the embodiment of the present invention, a group of random multi-wind turbine yaw angle schemes for the wind farm are initialized through the particle swarm optimization algorithm, where each particle represents a possible combination of wind turbine yaw angles. Based on the given wind speed, wind direction, and the parameters of the wind turbine, the established yaw single-machine double-Gaussian wake model and yaw multi-machine wake self-adaptive combined superposition model are used to simulate the wake characteristics of each wind turbine and their mutual influence at different yaw angles. By calculating the power generation of each wind turbine at a specific yaw angle and comprehensively considering the influence of the wake on the power generation efficiency of the downstream wind turbine, the total power generation of the wind farm is obtained.

[0083] Furthermore, according to the generated multi-wind turbine yaw angle scheme of the wind farm, the corresponding penalty factor is calculated. By introducing the penalty factor, it can ensure that the change of the wind turbine yaw angle meets the requirements of the ride comfort index, and avoid excessive wear and performance degradation of the wind turbine structure caused by the drastic fluctuation of the yaw angle. Specifically, for each wind turbine, the number or degree of times that the yaw angle exceeds the allowable fluctuation range under different working conditions is counted, and then these data are summarized to calculate the penalty factor of the entire wind farm.

[0084] Similarly, the particle swarm optimization algorithm continuously updates the positions and velocities of particles through iterative loops. The positions and velocities of the particles correspond to the angular parameters of each group of wind turbine yaw schemes and the change in the yaw angle of the wind turbines in the update scheme, thereby generating multiple groups of different yaw angle schemes for multi-wind turbines in a wind farm. In each iteration, the total power generation of the wind farm corresponding to each group of schemes and the penalty factor are recalculated, and this process is repeated to gradually explore the optimal yaw angle scheme.

[0085] More specifically, multiply the total power generation of the wind farm obtained in each iteration by the corresponding penalty factor to obtain the optimized power generation considering the yaw angle limit of the wind turbines. The optimized power generation not only reflects the power generation efficiency of the wind farm but also comprehensively considers the smoothness of the yaw angles of the wind turbines, ensuring the feasibility and reliability of the optimized scheme in actual operation.

[0086] In this way, continuously track and update the optimal yaw angle scheme for multi-wind turbines in a wind farm during the iterative process. When the particle swarm optimization algorithm reaches the convergence condition, such as the number of iterations reaches the preset value or the change in the optimized power generation tends to be stable, that is, under the principle of maximizing the optimized power generation, output the yaw angle scheme of the wind turbines at this time. This can effectively limit the fluctuation range of the yaw angles of the wind turbines while ensuring the power generation efficiency of the wind farm, extend the service life of the wind turbines, and improve the operation stability and economic benefits of the entire wind farm.

[0087] It should be noted that using the particle swarm algorithm to collaboratively optimize the yaw angles of multi-wind turbines in a wind farm takes into account the mutual influence between the wind turbines in the entire wind farm. During the iterative process, the power generation of the wind farm is expressed as the total power generation after considering the wake multiplied by the penalty factor, making the iterative results more in line with the actual operation requirements. Through collaborative optimization, the yaw strategy that maximizes the power generation of the wind farm is found, thereby improving the power generation efficiency of the entire wind farm.

[0088] By introducing the penalty factor as the optimization objective during the iterative process, this mechanism enables the system to automatically exclude yaw strategies that do not meet the requirements during the optimization process, thereby enhancing the robustness and reliability of the system. In practical applications, even in the face of extreme weather conditions or wind turbine failures, the system can ensure the overall operation efficiency and stability of the wind farm through collaborative optimization and the penalty factor mechanism.

[0089] In one embodiment, the steps of calculating the penalty factor according to the yaw angle scheme of multi-wind turbines in a wind farm include:

[0090] Determine the yaw angle of each wind turbine based on the yaw angle scheme of multi-wind turbines in a wind farm;

[0091] Judge whether the yaw angle of each wind turbine exceeds the yaw angle fluctuation range of the wind turbine;

[0092] If there is no yaw angle of the wind turbine exceeding the limit, output the base coefficient of the penalty factor.

[0093] If there is a yaw angle of the wind turbine exceeding the limit, calculate the ratio of the yaw angle of each wind turbine exceeding the yaw angle fluctuation range of the wind turbine and the ratio of the number of wind turbines exceeding the yaw angle fluctuation range in the wind farm.

[0094] Specifically, in the embodiment of the present invention, first, according to the yaw angle scheme of multiple wind turbines in the wind farm, clarify the yaw angle of each wind turbine under specific working conditions, which affects the subsequent calculation of the penalty factor. Next, judge the yaw angle of each wind turbine one by one to see if it exceeds the yaw angle fluctuation range of the wind turbine. If the yaw angle exceeds the allowable range, it may cause an increase in the fatigue load borne by the wind turbine, affecting the service life and operation stability of the wind turbine. If there is no yaw angle of the wind turbine exceeding the allowable range, directly output the base coefficient of the penalty factor. The base coefficient can usually be set to 1, indicating that when there is no exceeding the fluctuation range, no additional penalty is imposed on the total power generation, and the total power generation at this time is the optimized power generation. If there is a yaw angle of the wind turbine exceeding the allowable range, it is necessary to further calculate the ratio of the yaw angle of each wind turbine exceeding the yaw angle fluctuation range of the wind turbine and the ratio of the number of wind turbines exceeding the yaw angle fluctuation range in the wind farm.

[0095] More specifically, the calculation method of the penalty factor includes: for a single wind turbine, calculate the degree to which its yaw angle exceeds the allowable fluctuation range, that is, the ratio of the exceeding angle to the maximum angle of the allowable fluctuation range. For example, if the allowable fluctuation range of the wind turbine yaw angle is ±30°, and the actual yaw angle is +35°, then the exceeding ratio is (35° - 30°) / 30° = 5° / 30° ≈ 0.167. For multiple wind turbines in the wind farm: by counting the number of wind turbines exceeding the yaw angle fluctuation range in the entire wind farm, and then calculating the ratio of these wind turbine numbers to the total number of wind turbines in the entire wind farm. For example, there are 25 wind turbines in the wind farm, and 5 of them exceed the yaw angle fluctuation range, then the ratio is 5 / 25 = 0.2. Then combine the exceeding ratio of a single wind turbine and the ratio of the number of wind turbines to calculate the final penalty factor. The calculation formula of the penalty factor can be a simple product or a more complex function, specifically depending on the design of the optimization algorithm. For example, the penalty factor can be the difference between the weighted calculation of the exceeding ratio of a single wind turbine and the ratio of the number of wind turbines and the base coefficient. Through the above steps, a reasonable penalty factor can be obtained, which is used to evaluate and screen different yaw angle schemes in the particle swarm optimization algorithm to ensure that the final optimization scheme is both efficient and reliable.

[0096] In one embodiment, when the yaw angle of a wind turbine exceeds the limit, the parameter of the output penalty factor is smaller than its basic coefficient, and the greater the proportion of the yaw angle of each wind turbine exceeding the yaw angle fluctuation range of the wind turbine, the smaller the parameter of the output penalty factor, and the greater the proportion of wind turbines exceeding the yaw angle fluctuation range of the wind turbine in the wind farm, the smaller the parameter of the output penalty factor.

[0097] Specifically, in an embodiment of the present invention, in the particle swarm optimization algorithm, the basic coefficient of the penalty factor is usually set to 1, which means that in an ideal situation, the yaw angles of all wind turbines do not exceed the allowable fluctuation range. At this time, the total power generation of the wind farm is not subject to any penalty and is directly used as the optimized power generation to evaluate the pros and cons of the current yaw angle scheme. When the yaw angle of a certain wind turbine exceeds the allowable fluctuation range, the greater the excess ratio, the greater the impact on the penalty factor. Specifically, the excess ratio refers to the ratio of the difference between the actual yaw angle of the wind turbine and the maximum angle of the allowable fluctuation range to the maximum angle of the allowable fluctuation range. The larger the excess ratio, the more drastic the yaw angle adjustment of the wind turbine, and the higher the risk of mechanical wear and performance degradation that may be caused. Therefore, it is necessary to impose a greater penalty on the total power generation to reduce the possibility of the scheme being selected during the optimization process. In addition to the excess ratio of a single wind turbine, the proportion of the number of wind turbines in the wind farm that exceeds the allowable range is also a key factor affecting the penalty factor. The proportion of the number of wind turbines that exceeds the allowable range refers to the proportion of the number of wind turbines in the wind farm whose yaw angle exceeds the allowable range to the total number of wind turbines in the entire wind farm. The larger the proportion of wind turbines that exceed the limit, the more wind turbines in the wind farm have yaw angles that do not meet the smoothness requirements, and the greater the impact on the overall operating stability and wind turbine life. Therefore, when calculating the penalty factor, it is necessary to comprehensively consider the proportion of wind turbines that exceed the limit and impose a more severe penalty on the total power generation to ensure that the optimized solution can meet the overall smoothness requirements of the wind farm.

[0098] More specifically, the calculation of the penalty factor is a comprehensive evaluation process that combines the excess ratio of a single wind turbine and the ratio of the number of wind turbines in a wind farm. The specific calculation formula can be designed as: Penalty factor = 1 / (1+α×excess ratio of a single wind turbine+β×the ratio of the number of wind turbines in a wind farm). Among them, α and β are weight coefficients, which are used to adjust the degree of influence of the excess ratio of a single wind turbine and the ratio of the number of wind turbines in a wind farm on the penalty factor. The penalty factor calculated in this way can fully reflect the smoothness of the yaw angle scheme of multiple wind turbines in a wind farm. The smaller the penalty factor, the more wind turbine yaw angles in the scheme exceed the allowable range, and the correspondingly lower the optimized power generation, so it is screened out in the particle swarm optimization algorithm; conversely, the closer the penalty factor is to the basic coefficient 1, the more the scheme meets the smoothness requirements, the higher the optimized power generation, and the more likely it is to be selected as the final optimal scheme.

[0099] In this way, by inserting a penalty factor into the convergence condition of the particle swarm optimization algorithm, the particle swarm optimization algorithm can fully consider the smoothness of the yaw angle of the wind turbine while searching for the maximum power generation, avoid damage to the wind turbine caused by frequent large yaw movements, extend the service life of the wind turbine, and improve the operation efficiency and economic benefits of the wind farm.

[0100] In one embodiment, the convergence condition includes maximizing the total power generation of the wind farm within the allowable yaw angle fluctuation range of the wind turbine.

[0101] Specifically, in the embodiment of the present invention, during the iterative optimization process of the particle swarm optimization algorithm, the setting of the convergence condition is crucial for finding the yaw scheme of the wind turbine that meets the smoothness requirement and has the maximum power generation. Specifically, the convergence condition includes maximizing the total power generation of the wind farm within the allowable yaw angle fluctuation range of the wind turbine, which means that the optimization process not only pursues the increase of power generation but also ensures that the change of the yaw angle of the wind turbine meets the requirement of the smoothness index, and avoids damage to the wind turbine caused by frequent large yaw movements.

[0102] During the iterative process, the particle swarm optimization algorithm continuously updates the positions and velocities of the particles, and each particle represents a possible yaw angle scheme of the wind turbine. For each scheme, the total power generation of the wind farm is calculated through the established single-yaw double-Gaussian wake model and the adaptive combined superposition model of multi-yaw wakes. At the same time, according to whether the yaw angle of each wind turbine exceeds the allowable fluctuation range, the corresponding penalty factor is calculated. The introduction of the penalty factor makes the optimization objective change to maximizing the optimized power generation of the wind farm under the premise of meeting the smoothness requirement, that is, the product of the total power generation and the penalty factor.

[0103] When it is found during the iterative process that the optimized power generation of a certain set of yaw angle schemes reaches the maximum value and this value is no longer exceeded in subsequent iterations, or the change of the optimized power generation tends to be stable and the fluctuation amplitude is less than the set threshold, it is considered that the algorithm has reached the convergence condition. At this time, the corresponding yaw angle scheme is the optimal solution that meets the convergence condition, which can maximize the total power generation of the wind farm while ensuring the smooth change of the yaw angle of the wind turbine.

[0104] It should be noted that when judging whether the convergence condition is met, the overall operation status of the wind farm also needs to be comprehensively considered. For example, the stability of the optimized power generation in several consecutive iterations is statistically analyzed, and whether the yaw angle of the wind turbine continuously remains within the allowable fluctuation range. Only when all these conditions are met can it be ensured that the found optimal yaw angle scheme is feasible and effective in the actual operation of the wind farm, so as to realize the efficient and stable operation of the wind farm.

[0105] Thus, through the proposed new yaw wind turbine wake model with both accuracy and stability, it provides effective support for the collaborative yaw optimization of wind turbines in a wind farm. The proposed yaw multi-turbine wake adaptive combined superposition model gives full play to the advantages of each superposition model. Based on this, the proposed method for collaborative yaw optimization of multiple wind turbines in an offshore wind farm conducts collaborative optimization of wind turbine yaw by setting a smoothness index and using a particle swarm algorithm to improve the power generation efficiency of the entire wind farm and reduce the fatigue effect load.

[0106] In one embodiment, the yaw single-turbine double-Gaussian wake model is:

[0107]

[0108]

[0109] k * =0.11C t 1.07 I 0.2 (31)

[0110] ε=0.23C t 0.25 I 0.17 (32)

[0111] a=0.93C t -0.75 I 0.17 (33)

[0112] b=0.42C t 0.6 I 0.2 (34)

[0113] c=0.15C t -0.25 I -0.7 (35)

[0114] In the formula, U ∞ represents the free incoming flow; ΔU represents the deficit velocity at the position (x, y, z) downstream of the wind turbine with the wind turbine as the origin; D represents the wind turbine blade diameter; θ represents the wind turbine yaw angle; I represents the turbulence intensity; Ct represents the wind turbine thrust coefficient; α k and α m are taken as 0.7 and 0.65 respectively.

[0115] Specifically, in the embodiments of the present invention, based on the numerical simulation of the yaw single-turbine wake and the results of wind tunnel tests, existing yaw wind turbine wake models, including the Jensen model, the Gaussian model, the BPA model, etc., are compared and analyzed. The deficiencies of the existing wake models are deeply explored, and a double-Gaussian wake model for yaw wind turbines with better accuracy and stronger robustness is constructed. Formulas (1) to (9) constitute the double-Gaussian wake model for the yaw single-turbine, which is used to describe the wind speed deficit distribution in the wake region of the wind turbine under yaw conditions. Among them: The symbols in the formulas, such as the free-stream wind speed, the downstream position of the wind turbine, the wind turbine blade diameter, the yaw angle, the turbulence intensity, the wind turbine thrust coefficient, etc., are all key parameters affecting the wind speed deficit in the wake. Through these formulas, the wind speed deficit in the wake caused by the yaw of the wind turbine at different positions can be calculated, providing a basis for the subsequent multi-turbine wake superposition model and collaborative optimization method.

[0116] In one embodiment, the adaptive combined superposition model for the yaw multi-turbine wake is as follows:

[0117] V = ωV SS +(1 - ω)V GS (36)

[0118]

[0119] In the formula, V represents the wake wind speed; V SS represents the wake wind speed obtained based on the sum-of-squares superposition model; V GS represents the wake wind speed obtained based on the geometric superposition model; x1 and x2 respectively represent the x / D values corresponding to the weight coefficient of 1 for the sum-of-squares superposition model and the geometric superposition model. For example, x1 and x2 can be respectively set to 6 and 10 to achieve the optimal utilization of these two wake superposition models and ensure the smooth transition of the flow field.

[0120] Specifically, in the embodiments of the present invention, an adaptive combined superposition model is constructed for the sum-of-squares superposition model and the geometric superposition model, and its weight coefficient varies with the spatial position. Formulas (10) and (11) are the calculation formulas of the adaptive combined superposition model for the yaw multi-turbine wake, which are used to comprehensively consider the advantages of the sum-of-squares superposition model and the geometric superposition model, and adaptively combine the weight coefficients of the two models according to different spatial positions to achieve a more accurate calculation of the wake wind speed.

[0121] In one embodiment, the execution steps of the particle swarm optimization algorithm include:

[0122] Initialize the particle swarm based on the set relevant parameters to generate the yaw scheme of each particle in the particle swarm;

[0123] Calculate the total power generation of the wind farm corresponding to the particle swarm by inserting the inertia factor and use it as the fitness of the particle swarm;

[0124] Compare the current fitness of the particle swarm with the historical optimal value, and update the yaw scheme of the wind turbines with the optimal total power generation of the entire wind farm and the yaw scheme of the wind turbines with the optimal total power generation of the local area of the wind farm;

[0125] Determine whether the yaw scheme of the wind turbines meets the set convergence conditions;

[0126] If not, continue to calculate the fitness of the particle swarm by inserting the inertia factor;

[0127] If it is satisfied, output the yaw scheme of the wind turbines with the optimal total power generation of the entire wind farm as the multi-wind turbine collaborative yaw optimization scheme.

[0128] Specifically, in the embodiment of the present invention, the relevant parameters of the particle swarm optimization algorithm may include, for example, the number of particles, the number of iteration steps, the inertia factor, the individual learning factor, the social learning factor, the yaw angle limit range, etc. By performing the initialization operation of the particle swarm, the yaw scheme of the wind turbines corresponding to each particle in the particle swarm is generated. Each particle represents a possible combination of yaw angles of the wind turbines. When initializing, the positions and velocities of the particles are randomly generated within the specified range to ensure the diversity of the particles and provide a good starting point for subsequent optimization searches. The inertia factor is used to balance the global and local nature of the particle search. Initially, the inertia factor is relatively large, enabling the particles to conduct extensive searches globally and avoid falling into local optima prematurely. As the iteration progresses, the inertia factor gradually decreases, and the particles tend to conduct fine searches in the local area to improve the search accuracy. When calculating the fitness, it is necessary to comprehensively consider the influence of the yaw angle of the wind turbines on the wake flow and the overall power generation efficiency of the wind farm. Each particle records its own best position (individual optimum) experienced, and at the same time, the group also records the best position among all particles (global optimum).

[0129] Furthermore, in each iteration, by comparing the fitness of the current particle with the fitness of the individual optimum and the global optimum, it is decided whether to update these optimal positions. If the fitness of the current particle is better, the corresponding optimal positions are updated to guide the particles towards better solutions. The convergence conditions usually include reaching the maximum number of iterations, the change in the fitness value being stable within a certain range, and maximizing the total power generation within the yaw angle fluctuation range of the wind turbines. By checking whether these conditions are met, it is determined whether the algorithm stops iterating. If not, continue to calculate the fitness of the particle swarm by inserting the inertia factor, and repeat the above steps until the convergence conditions are met. While ensuring the smooth change of the yaw angle of the wind turbines, the total power generation of the wind farm is maximized, the operation efficiency and economic benefits of the wind farm are improved, the service life of the wind turbines is extended, and the stable operation of the wind farm is ensured.

[0130] Thus, an initial particle swarm is randomly generated based on the set parameters, and each particle represents a set of yaw angle schemes for the wind turbines. Calculate the total power generation P of the wind farm corresponding to each particle, and calculate the penalty factor in combination with the proportion of the units whose yaw angles exceed the allowable fluctuation range. Multiply the total power generation P by the penalty factor to obtain the corrected fitness P', which is used as the optimization objective function. Adjust the particle swarm velocity by using an inertia factor that changes with time, update the particle positions, and ensure that the yaw angles are within the preset limit range. After the update, recalculate the particle fitness P', and compare the current particle swarm, the historical local optimum, and the global optimum solution to update the individual optimum position and the global optimum position. Finally, check whether the algorithm meets the convergence condition (such as reaching the maximum number of iterations or the fitness change threshold). If not converged, repeat the iterative optimization; if converged, terminate the loop and output the yaw angle scheme corresponding to the global optimum solution.

[0131] In one embodiment, the step of calculating the total power generation of the wind farm corresponding to the particle swarm by inserting the inertia factor and using it as the fitness of the particle swarm includes providing an inertia factor that changes with time, which is used to calculate the velocity of the particle swarm and update the position of the particle swarm; wherein the inertia factor, the velocity of the particle swarm, and the position of the particle swarm are:

[0132]

[0133] v i =w p v i +c1r d (pb i -x i )+c2r d (gb-x i ) (38)

[0134] x i =x i +v i (39)

[0135] In the formula: wp - inertia factor; w1 - initial inertia factor; w2 - final inertia factor; Niter - total number of iterations; j - current iteration step; vi - velocity of the i-th particle, m / s; c1 - individual learning factor; c2 - social learning factor; pbi - local optimum position of the particle, m; gb - global optimum position of the particle, m; xi - position of the particle, m; rd - random number between 0 and 1.

[0136] Specifically, in the embodiments of the present invention, formulas (12) to (14) are the key formulas for calculating the particle velocity and updating the position in the particle swarm algorithm. Formula (12) gives the calculation method of the inertia factor wp, which changes with time. The initial inertia factor w1, the final inertia factor w2, the total number of iterations Niter, and the current iteration step j jointly determine the size of the inertia factor, thereby affecting the movement trend of the particle in the search space. Formulas (13) and (14) calculate the particle velocity and update the particle position respectively, which involve parameters such as the individual learning factor c1, the social learning factor c2, the local optimal position pbi of the particle, the global optimal position gb of the particle, and the random number rd. These parameters work together to enable the particle to explore new solution spaces and gradually converge to better solutions during the search process, and are used to optimize the yaw angle of the wind turbines in the wind farm to achieve goals such as maximizing the power generation efficiency and smooth change of the yaw angle. During the iteration process, the inertia factor of the particle swarm optimization algorithm changes with time to balance the global search and local search capabilities, thereby improving the optimization efficiency and accuracy.

[0137] In one embodiment, after the step of calculating the total power generation of the wind farm corresponding to the particle swarm by inserting the inertia factor, it further includes multiplying the total power generation of the wind farm by a penalty factor to obtain the optimized power generation of the wind farm based on the allowable fluctuation range of the yaw angle of the wind turbine, and using it as the fitness of the particle.

[0138] Specifically, in the embodiments of the present invention, the introduction of the penalty factor is to ensure that the change of the yaw angle of the wind turbine meets the requirements of the smoothness index and avoid damage to the wind turbine caused by frequent large yaw. After calculating the total power generation of the wind farm, it is necessary to further consider the smoothness of the yaw angle of the wind turbine. By multiplying the total power generation by the penalty factor, the optimized power generation is obtained. This optimized power generation not only reflects the power generation efficiency of the wind farm, but also comprehensively considers the fluctuation range of the yaw angle of the wind turbine, ensuring the feasibility and reliability of the optimization scheme in actual operation.

[0139] In this way, the optimized power generation calculated with the penalty factor is used as the fitness of the particle to evaluate the quality of the particle. In the particle swarm optimization algorithm, the particle with a higher fitness value is more likely to be selected as the new individual optimal or global optimal solution. In this way, while searching for the maximum power generation, the smoothness of the yaw angle of the wind turbine can be fully considered, avoiding damage to the wind turbine caused by frequent large yaw, prolonging the service life of the wind turbine, and improving the operation efficiency and economic benefits of the wind farm.

[0140] In one embodiment, by setting the base coefficient of the penalty factor, the yaw angle scheme of the wind turbine corresponding to the total power generation of the wind farm is screened. In the multi-wind turbine collaborative yaw optimization scheme, the proportion of the total power generation factor of the wind farm is reduced, and the proportion of the yaw angle fluctuation range factor of the wind turbine is increased.

[0141] Specifically, in the embodiment of the present invention, the basic coefficient of the penalty factor is usually set to 1, representing an ideal condition, that is, the penalty factor value when the yaw angles of all wind turbines do not exceed the allowable fluctuation range. At this time, the total power generation of the wind farm is not subject to any penalty and is directly used as the optimized power generation to evaluate the pros and cons of the current yaw angle scheme. By setting the basic coefficient of the penalty factor, the optimization target can be adjusted in the multi-wind turbine collaborative yaw optimization scheme.

[0142] More specifically, by reducing the weight of the total power generation factor of the wind farm, it means that in the optimization process, we no longer simply pursue the maximum power generation, but pay more attention to the smoothness of the yaw angle of the wind turbine. Similarly, for wind turbine equipment with different model characteristics, the value of the basic coefficient can be adjusted in a targeted manner. For example, the parameter value of the basic coefficient can be adjusted according to the size of the acceptable fluctuation range of the wind turbine yaw angle. At the same time, the weight of the wind turbine yaw angle fluctuation range factor is increased to ensure that the optimized solution can reduce the mechanical wear of the wind turbine in actual operation, extend the service life of the wind turbine, and improve the operational stability and economic benefits of the wind farm.

[0143] Please see attached Figures 3 to 5 In one embodiment, the impeller diameter of a certain wind turbine model is 240.55m, the hub height is 143.8m, the rated power is 12MW, the cut-in wind speed is 3m / s, and the cut-out wind speed is 25m / s. Figure 3 There are 25 wind turbines arranged as shown. The spacing between wind turbines in the east-west direction is longer, while the spacing in the north-south direction is shorter. For this reason, four working wind directions are defined, and the wind speeds are 5, 10, 15, 20, and 25 m / s for each wind direction, in total, to consider the situation where the yaw angle of the wind turbine fluctuates smoothly with the wind speed. In addition, coordinated yaw optimization is performed for all wind directions. The optimization results for a single wind turbine are shown in the attached figure. Figure 4 The maximum yaw fluctuation range limited by the optimization result is 30°, and the wind speed interval is 5m / s, which makes the yaw result fluctuate greatly.

[0144] Please see attached Figure 3 , each position point marked in the coordinate system represents the position point of a wind turbine, and is arranged according to its actual layout in the wind farm, which intuitively presents the layout structure of the wind farm and helps to understand the relative position relationship and spacing between wind turbines. When performing multi-wind turbine yaw collaborative optimization, the position relationship between wind turbines affects the wake effect, which will vary with the distance and relative position between wind turbines. By clarifying the position of the wind turbine, the wake model can be established more accurately, and the mutual influence between wind turbines can be analyzed, so as to formulate a reasonable collaborative optimization strategy to improve the power generation efficiency and operation stability of the entire wind farm.

[0145] Please see attached Figure 4, which shows the yaw results of a single wind turbine. Under different wind speeds and directions, the variation of the yaw angle of the wind turbine. The maximum yaw fluctuation range is 30°, and the wind speed interval is taken as 5 m / s, resulting in large fluctuations in the yaw results. It reflects the yaw effect of a single wind turbine under different working conditions. Although it can improve the power generation efficiency of the wind turbine to a certain extent, due to the lack of consideration of the mutual influence between multiple wind turbines and the continuity of wind speed and direction changes, the yaw angle fluctuates greatly, which may exacerbate the fatigue load of the wind turbine and affect its service life.

[0146] Please refer to the appendix Figure 5 , which shows the results after the multi-wind turbine yaw collaborative optimization method. Under different wind speeds and directions, the variation of the yaw angles of multiple wind turbines. The effect of the collaborative yaw optimization of the wind turbines is significant, and the yaw angles change smoothly with the wind speed and direction, with small fluctuations. It intuitively reflects the actual application effect of the collaborative optimization method, proving that this method can effectively solve the problems existing in the yaw optimization of a single wind turbine. By considering the mutual influence between wind turbines in the entire wind farm and the continuity of wind speed and direction changes, the smooth adjustment of the yaw angles of the wind turbines is realized, which not only reduces the fatigue load of the wind turbines and extends the service life of the wind turbine structure, but also has important guiding significance for the actual operation and management of offshore wind farms, and can help wind farm operators formulate more reasonable control strategies to achieve a double improvement in economic benefits and equipment reliability.

[0147] In this way, through the use of the multi-wind turbine yaw collaborative optimization method in the embodiments of the present invention for the overall optimization of the yaw of wind turbines in all wind directions and all wind speeds, the execution results of the optimized yaw angle scheme of the wind turbines are as shown in the appendix Figure 5 shown. Generally speaking, the effect of the collaborative yaw optimization of the wind turbines is significant, and the yaw angles of the wind turbines change smoothly with the wind speed and direction, with small fluctuations, which can reduce the fatigue load borne by the wind turbines to a certain extent.

[0148] In summary, a multi-wind turbine collaborative yaw optimization method provided by the present invention not only focuses on the yaw control of a single wind turbine, but also considers the collaborative optimization of the entire wind farm. By real-time monitoring and analyzing the operating states and environmental conditions of each wind turbine in the wind farm, it is possible to achieve refined management of the wind farm, improve the operating efficiency and economic benefits of the wind farm, reduce the operation and maintenance costs, and provide a strong guarantee for the long-term stable operation of the wind farm. At the same time, compared with existing methods, it has obvious advantages in aspects such as the smoothness of wind turbine operation, the power generation efficiency of the wind farm, the robustness and reliability of the system, and the operation and management of the wind farm. These advantages will contribute to the sustainable development and progress of the offshore wind power industry.

[0149] The above embodiments are only illustrative of the principles and effects of the present invention and are not intended to limit the present invention. Any person familiar with this technology can modify or change the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or changes made by those with ordinary knowledge in the technical field without departing from the spirit and technical idea disclosed by the present invention should still be covered by the claims of the present invention.

Claims

1. A multi-wind turbine collaborative yaw optimization method for an offshore wind farm, characterized in that Including: Based on the given wind speed, wind direction and fan parameters, establish a single - machine double - Gaussian yaw wake model and a multi - machine yaw wake adaptive combined superposition model, and set it as the overall working condition of the wind farm; According to the overall working condition, set the relevant parameters of the particle swarm optimization algorithm, and perform iterative loops on the fan yaw scheme through the particle swarm optimization algorithm with an inserted penalty factor; Output the fan yaw scheme that meets the convergence condition as the multi - fan collaborative yaw optimization scheme.

2. The multi-fan collaborative yaw optimization method according to claim 1, characterized in that The fan parameters also include a smoothness index that limits the fluctuation range of the fan yaw angle.

3. The multi-fan collaborative yaw optimization method according to claim 1, wherein The steps of performing iterative loops on the fan yaw scheme through the particle swarm optimization algorithm with an inserted penalty factor include: Generate an initial multi - fan yaw angle scheme for the wind farm through the particle swarm optimization algorithm, and calculate the total power generation of the wind farm; Calculate the penalty factor according to the multi - fan yaw angle scheme of the wind farm; Calculate the total power generation and penalty factor of multiple groups of multi - fan yaw angle schemes for the wind farm through iterative loops; Multiply the total power generation of the wind farm by the penalty factor as the optimized power generation of the wind farm; Determine the corresponding multi - fan yaw angle scheme for the wind farm based on maximizing the optimized power generation of the wind farm.

4. The multi-fan collaborative yaw optimization method according to claim 3, wherein The steps of calculating the penalty factor according to the multi - fan yaw angle scheme of the wind farm include: Determine the yaw angle of each fan based on the multi - fan yaw angle scheme of the wind farm; Judge whether the yaw angle of each fan exceeds the fan yaw angle fluctuation range; If there is no fan whose yaw angle exceeds, output the basic coefficient of the penalty factor; If there is a fan whose yaw angle exceeds, calculate the proportion of the yaw angle of each fan that exceeds the fan yaw angle fluctuation range and the proportion of the number of fans in the wind farm whose yaw angles exceed the fan yaw angle fluctuation range.

5. The multi-fan collaborative yaw optimization method according to claim 4, wherein When there is a fan whose yaw angle exceeds, the parameter of the output penalty factor is less than its basic coefficient, and the greater the proportion of the yaw angle of each fan that exceeds the fan yaw angle fluctuation range, the smaller the parameter of the output penalty factor, and the greater the proportion of the number of fans in the wind farm whose yaw angles exceed the fan yaw angle fluctuation range, the smaller the parameter of the output penalty factor.

6. The multi-fan collaborative yaw optimization method according to claim 1, wherein The convergence condition includes setting the maximization of the total power generation of the wind farm within the allowable fan yaw angle fluctuation range.

7. The multi-fan collaborative yaw optimization method according to claim 1, characterized in that The yaw single - machine double - Gaussian wake model is: Where, U ∞ represents the free stream; ΔU represents the deficit velocity at the position (x, y, z) downstream of the wind turbine with the wind turbine as the origin; D represents the wind turbine blade diameter; θ represents the wind turbine yaw angle; I represents the turbulence intensity; Ct represents the wind turbine thrust coefficient; αk and α m are taken as 0.7 and 0.65 respectively.

8. The multi-fan collaborative yaw optimization method according to claim 1, characterized in that The yaw multi - machine wake adaptive combined superposition model is: V = ωV SS +(1 - ω)V GS (10) where V represents the wake wind speed; V SS represents the wake wind speed obtained based on the sum-of-squares superposition model; V GS represents the wake wind speed obtained based on the geometric superposition model; x1 and x2 respectively represent the x / D values corresponding to the weight coefficients of 1 for the sum-of-squares superposition model and the geometric superposition model.

9. The multi-fan collaborative yaw optimization method according to claim 1, wherein The execution steps of the particle swarm optimization algorithm include: Based on the set relevant parameters, perform particle swarm initialization and generate the fan yaw scheme corresponding to each particle in the particle swarm; Calculate the total power generation of the wind farm corresponding to the particle swarm by inserting an inertia factor and use it as the fitness of the particle swarm; Compare the current fitness of the particle swarm with the historical optimal value, and update the fan yaw scheme with the optimal total power generation of the wind farm globally and the fan yaw scheme with the optimal total power generation of the wind farm locally; Judge whether the fan yaw scheme reaches the set convergence condition; If not satisfied, continue to calculate the fitness of the particle swarm by inserting an inertia factor; If satisfied, output the fan yaw scheme with the optimal total power generation of the wind farm globally as the multi - fan collaborative yaw optimization scheme.

10. The multi-fan collaborative yaw optimization method according to claim 9, characterized in that, The step of calculating the total power generation of the wind farm corresponding to the particle swarm by inserting an inertia factor and using it as the fitness of the particle swarm includes providing an inertia factor that varies with time, which is used to calculate the velocity of the particle swarm and update the position of the particle swarm; wherein the inertia factor, the velocity of the particle swarm, and the position of the particle swarm are as follows: x i = x i + v i (13) In the formula: wp - inertia factor; w1 - initial inertia factor; w2 - final inertia factor; Niter - total number of iterations; j - current iteration step; vi - velocity of the i-th particle, m / s; c1 - individual learning factor; c2 - social learning factor; pbi - local optimal position of the particle, m; gb - global optimal position of the particle, m; xi - particle position, m; rd - random number between 0 and 1.

11. The multi-fan collaborative yaw optimization method according to claim 9, characterized in that, After the step of calculating the total power generation of the wind farm corresponding to the particle swarm by inserting an inertia factor, it further includes multiplying the total power generation of the wind farm by a penalty factor to obtain the optimized power generation of the wind farm based on the allowable fluctuation range of the yaw angle of the wind turbine, and using it as the fitness of the particle.

12. The multi-fan collaborative yaw optimization method according to claim 5, characterized in that, By setting the base coefficient of the penalty factor, screening the yaw angle scheme of the wind turbine corresponding to the total power generation of the wind farm, in the multi-wind turbine collaborative yaw optimization scheme, reducing the proportion of the total power generation factor of the wind farm and increasing the proportion of the wind turbine yaw angle fluctuation range factor.

Images