A Field-level Cooperative Control Strategy for Offshore Wind Farms Based on Wake Tracking

Abstract

The present invention discloses a field-level collaborative control strategy for an offshore wind farm based on wake tracking, including a nacelle-mounted lidar wind measurement system, a wake tracking module, an optimizer, and a field-level controller. First, the original wind information is measured by the nacelle-mounted lidar wind measurement system, and the wake tracking module is used to perform wind field inversion to complete the parameter evaluation of the environmental input wind and the identification of wake characteristic parameters. Based on the aero-hydro-servo-elastic dynamics simulation model of the offshore wind farm, the optimal wake center position under different environmental conditions is solved by an intelligent optimization algorithm, and a multi-dimensional intelligent decision database LUT and an optimizer are established. The field-level collaborative PI controller is used to achieve wake redirection and intelligent control. The field-level collaborative control method of the present invention can realize online real-time tracking of the wake, dynamically improve the wake effect, and achieve efficiency increase and load reduction of the offshore wind farm; the controller has the characteristics of good control effect and fast execution speed.

Description

Technical Field

[0001] The present invention belongs to the technical field of coordinated control of offshore wind farms, and particularly relates to an offshore wind farm field-level coordinated control strategy based on wake tracking. Background Art

[0002] With the rapid development of the economic society, the power demand has increased significantly, and it is necessary to vigorously develop renewable energy. As an important part of renewable energy, offshore wind energy has rich reserves. Moreover, offshore wind power is relatively close to the coastal cities with intensive power consumption, which is convenient for power transmission and consumption, can effectively save land resources, and reduce visual and noise pollution. Therefore, offshore wind power generation is considered to be one of the effective ways to solve energy demand and environmental problems.

[0003] However, existing offshore wind farms are often arranged closely due to sea area restrictions, thus forming a wake effect that brings power loss and fatigue loads. Research shows that the wake effect can cause a 40% power deficit and an 80% increase in fatigue loads to downstream wind turbines, ultimately reducing the power generation capacity and fatigue life of the wind farm, and affecting the economic benefits and safe operation of the entire wind farm.

[0004] To improve the wake effect, existing research has proposed an offshore wind farm field-level coordinated control strategy, which uses advanced sensing and control methods to achieve coordinated operation control of wind turbines, reduce the wake influence, and achieve power improvement and load reduction at the wind farm field level. The specific control methods include wake redirection control and axial induction factor control.

[0005] Chinese patent document with publication number CN111980857A discloses a closed-loop control method for a wind farm, including: calculating the preliminary optimal yaw angle of each wind turbine in the wind farm under different incoming flows through the engineering wake model of the wind farm; measuring the on-site wake flow data of the wind farm; and correcting the preliminary optimal yaw angle of each wind turbine under different incoming flows based on the on-site wake flow data of the wind farm as feedback.

[0006] Chinese patent document with publication number CN108953060A discloses a wind farm field-level yaw control method based on a lidar anemometer, including: measuring the wind speed and wind direction data at corresponding positions through lidars arranged at predetermined positions and in a predetermined number within the wind farm; processing the wind speed data to obtain a local single-machine wake model of the wind farm; obtaining the wake model of the entire wind farm through the local wake model; calculating the current yaw angle correction factor of the wind turbine based on the wake model of the entire wind farm; measuring the advance wind direction information through the front-row lidar as the compensation signal for the yaw of the first row of wind turbines; and performing yaw control of the wind turbine based on the wind direction signal of the wind speed and wind direction indicator, the current yaw correction signal, and the compensation signal.

[0007] However, the atmospheric wind field is complex and the wake field structure varies. It is difficult to achieve accurate wake assessment in an offshore wind farm using existing technical methods. Currently, the existing offshore wind farm coordinated control technology does not have a mature intelligent wake control ability. Summary of the Invention

[0008] The present invention provides an offshore wind farm field-level coordinated control strategy based on wake tracking. The real-time wake tracking is achieved through lidar wind measurement technology, and the wake redirection and intelligent control are realized by using the field-level coordinated control technology, so as to improve the power generation of the offshore wind farm system and achieve the efficiency increase and load reduction of the offshore wind farm.

[0009] An offshore wind farm field-level coordinated control strategy based on wake tracking includes:

[0010] (1) Using the nacelle lidar wind measurement system to measure the wind speed distribution data of the front side flow field of each offshore wind turbine;

[0011] (2) Establishing a wake tracking module, and using the wind speed distribution data obtained in step (1) to complete the wind information assessment of the incoming wind without wake influence and the identification of wake characteristic parameters under wake influence through three-dimensional wind field inversion. The wake characteristic parameters include wake depth, wake width, and wake center position;

[0012] (3) According to the layout characteristics and flow field distribution of the actual offshore wind farm, establishing an aero-hydro-servo-elastic dynamics simulation model of the offshore wind farm; using the historical wind speed distribution data measured by the lidar wind measurement system to correct the parameters of the simulation model, and using the simulation model to predict the operating conditions of the wind farm under the input wind conditions and control strategies;

[0013] (4) Establishing a multi-dimensional intelligent decision-making database for an offshore wind farm based on wake redirection to realize the online real-time search for the optimal control strategy under different environmental wind conditions;

[0014] Establishing a multi-objective optimization function that comprehensively considers the maximization of wind farm power and the minimization of load, obtaining the optimal wake center position information of the wind farm under different environmental wind condition inputs through an intelligent optimization algorithm, and recording the above parameter information into the multi-dimensional intelligent decision-making database;

[0015] (5) Establishing an offshore wind farm field-level coordinated PI controller, using the look-up table method to online search the multi-dimensional intelligent decision-making database to obtain the optimal wake center position information of the wind farm under the corresponding wind condition input, taking it as the reference value of the field-level coordinated PI controller, and according to the measured wake information value measured by the wake tracking module, changing the yaw misalignment degree of each unit and the incoming environmental wind through the field-level coordinated PI controller to achieve wake redirection and control optimization.

[0016] Further, the wake tracking module aims to utilize the nacelle-mounted lidar measurements in an offshore wind farm to obtain the wind speed distribution of the front-side flow field. Through data processing such as north correction of the wind direction, conversion of the horizontal wind speed, inversion of the three-dimensional wind field, and identification of wake parameters, the wind information assessment of the incoming wind under the influence of no wake and the identification of wake characteristic parameters under the influence of wake are completed to obtain wake parameter information.

[0017] The specific process of step (2) is as follows:

[0018] (2-1) Inversion of the three-dimensional wind field

[0019] The nacelle-mounted lidar measurement system in an offshore wind farm can accurately remotely measure the wind conditions in front of its installation position, obtain the wind profile of the wind turbine and complex wake states, and is mainly used for the performance optimization and overall control of the wind farm. In the process of field-level collaborative control of the wind farm described in the present invention, the nacelle of the offshore wind turbine often also faces problems such as continuous change of the spatial attitude position over time, which brings difficulties to the lidar in wind direction measurement. Therefore, it is necessary to design a corresponding wind speed algorithm during the inversion of the three-dimensional wind field to eliminate errors.

[0020] To accurately evaluate the wind speed distribution information inside the offshore wind farm, first, north correction of the wind direction is carried out, and the pointing of the laser beam in the earth coordinate system is inverted through the coordinate rotation transformation matrix algorithm; the coordinate transformation formula of the laser beam pointing is:

[0021] L LOS =(H1H2H3) -1 L LOS_Lidar (1)

[0022] where L LOS is the pointing of the laser beam in the earth coordinate, H1, H2, and H3 are the coordinate rotation transformation matrices, and L LOS_Lidar is the pointing of the laser beam in the nacelle coordinate;

[0023] Define the directions of the nacelle coordinate system: X0 is the rotor axis pointing towards the nacelle head direction, Y0 is the rotor radius pointing towards the horizontal left side, and Z0 is coaxial with the wind turbine tower axis and points towards the nacelle top direction; the laser beam pointing is described by the azimuth angle and the pitch angle θ0, where the azimuth angle is the angle between the projection of the laser beam on the X0Y0 plane and X0, and the pitch angle θ0 is the angle between the laser beam and the X0Y0 plane; the changes in the attitude formed by the wind turbine nacelle during the collaborative control process are represented by the roll angle pitch angle θ, and yaw angle ψ; therefore, the pointing L LOS_Lidar of the laser beam in the nacelle coordinate and the coordinate rotation transformation matrix are expressed as follows:

[0024]

[0025]

[0026]

[0027]

[0028] The actual pointing of the laser beam can be directly obtained by substituting equations (2), (3), (4), and (5) into equation (1). Further solving gives the line-of-sight wind speed v of the laser beam. LOS :

[0029] v LOS = v wind ·L LOS (6)

[0030] During the actual scanning measurement of the marine nacelle-mounted lidar, line-of-sight wind speed information in multiple azimuths at any height can be obtained. At the same height and circular scanning layer, the line-of-sight wind speed is a trigonometric function under ideal conditions and is defined as:

[0031] f(v LOS ) = a cos(θ - b) + c (7)

[0032] The parameters a, b, and c are determined by the wind vector information at the measured height. Specifically:

[0033]

[0034] Among them, v1, v2, and v3 are the velocity components of the measured wind farm in the north-south direction, east-west direction, and vertical direction in the earth coordinate system, which determine the parameter values of the above line-of-sight wind speed function;

[0035] By fitting the line-of-sight wind speed described by equations (7) and (8) using the least squares method, and minimizing the function error in equation (9), the parameters a, b, and c of the curve function are solved, completing the inversion of the wind speed distribution of the three-dimensional wind farm;

[0036]

[0037] Among them, N is the number of test points of the nacelle-mounted lidar in a certain measurement period; based on the above wind field inversion, the wind information of the incoming flow wind of the wind farm without wake influence is obtained, including wind speed, wind direction, and turbulence intensity, and at the same time, it provides a data basis for the identification of wake characteristic parameters;

[0038] (2 - 2) Identification of wake characteristic parameters

[0039] Based on the original data measured by the nacelle lidar of an offshore wind farm, the three-dimensional wind field inversion is completed through the north correction of the wind direction and the conversion of the horizontal wind speed, and the wind speed distribution in front of the wind turbine is obtained. Using the wind speed data at the hub height of the wind turbine, the wake parameter identification based on the inverse Gaussian distribution is completed to obtain the characteristic parameters of the wake. Specifically, an inverse Gaussian function of the wake velocity distribution received in front of the wind turbine is established, and the formula is as follows:

[0040]

[0041] where U is the wind speed at the position of (x, z h ) of the wind turbine hub height at time t, U0 is the ambient average wind speed at time t, C s is the wake ratio coefficient used to quantify the wake depth, e is the natural constant, μ is the wake center position at the hub height, σ is the wake standard deviation used to quantify the wake width, and π is the pi; based on the layout of the offshore wind farm, the present invention assumes that the wake center position is approximately located at the hub height of the downstream wind turbine.

[0042] Using the wind speed data at the corresponding distances in front of the wind turbine measured by the lidar, the parameter fitting of the inverse Gaussian function is completed by the least squares method to obtain the wake characteristic parameters, specifically including the wake depth, wake width and wake center position information.

[0043] In step (3), the aero-hydro-servo-elastic dynamics simulation model of the offshore wind farm can be used to predict the dynamic operation characteristics of the offshore wind farm under different environmental conditions and control inputs, specifically including the flow field distribution and the dynamic responses of the unit power, load, etc.

[0044] The aero-hydro-servo-elastic dynamics simulation model of the offshore wind farm specifically includes a wind farm aerodynamic simulation model, an offshore wind turbine model and a wake model, etc.

[0045] The aerodynamic simulation model of the wind farm uses the Taylor frozen turbulence hypothesis to complete the dynamic propagation simulation of the three-dimensional flow field of the wind farm in space; the offshore wind turbine model adopts a brake disc model, and the modeling is completed according to the performance parameters of the actual unit and the dynamic responses of the unit power and load under specific environmental conditions are obtained; the wake model is used to predict the key wake characteristics related to the wind farm output power and the unit load, including the wake wind speed deficit, wake expansion and wake meandering, etc.

[0046] In step (4), the multi-dimensional intelligent decision-making database LUT for wake redirection of an offshore wind farm aims to record the optimal wake center position information in front of each unit under different environmental conditions of the offshore wind farm. The wake redirection in the present invention is achieved through active yaw coordination control of the wind farm. By misaligning the upstream wind turbines, the wake trajectory of the downstream is changed, thereby improving the power loss and load increase caused by the wake effect of the downstream wind turbines. In order to achieve multi-objective optimization under the dynamically changing environmental conditions of an actual offshore wind farm, the multi-dimensional intelligent decision-making database LUT established in the present invention can be used to achieve online real-time search for the optimal wake center position, so as to realize wake redirection by using coordinated yaw control.

[0047] Specifically, in order to establish the multi-dimensional intelligent decision-making database LUT for wake redirection of the offshore wind farm, the aerodynamic-hydro-servo-elastic dynamics simulation model of the offshore wind farm described in step (3) is used to obtain the flow field distribution, and the dynamic responses such as the power and load of the units under different environmental conditions and control inputs of the offshore wind farm. According to the time-domain changes of the power and load of the offshore wind farm system within a unit time, the total output power and the fatigue loads at the key parts of the units are calculated respectively.

[0048] The specific process of establishing a multi-objective optimization function that comprehensively considers maximizing the power of the wind farm and minimizing the load is as follows:

[0049] (4-1) Calculate the output power of the offshore wind turbine i:

[0050]

[0051] In the formula, P i is the output power (W) of the generator of unit i in the wind farm, T q,i (t) is the instantaneous torque (N-m) of the generator of unit i at time t, ω i is the instantaneous rotational speed (rpm) of the generator of unit i at time t, t k is the current time, and ΔT is the time domain period currently being solved. In the present invention, the inflow angle of the environmental wind can be changed through active yaw control to optimize the operating performance of the wind farm.

[0052] According to the above power calculation formula for a single unit, the total power generation of the wind farm system is the sum of the power generations of all the wind turbines in the wind farm system It is:

[0053]

[0054] (4-2) Calculate the equivalent fatigue load on the wind turbine i:

[0055] The time-domain variation of the loads on the key parts of the wind turbine generator i in the wind farm is obtained by using the aero-hydro-servo-elastic dynamics simulation model of the wind farm, including the load variations at the bottom of the tower and the root of the blade; through the rain-flow counting method, the time-domain varying loads on the wind turbine generator i are equivalently analyzed to obtain the equivalent fatigue load DEL of the key parts:

[0056]

[0057] In the formula, DEL is the equivalent fatigue load, is the equivalent number of cycles within the time series j, N j,i is the number of occurrences of the i-th working condition within the time series j, L j,i is the load range of the i-th working condition within the time series j, and m is the slope of the material S-N curve;

[0058] (4-3) To maximize the total output power of the wind farm and reduce the fatigue loads on the key parts of the units, the following multi-objective optimization function is established:

[0059]

[0060] In the formula, is the total power generation of the wind farm system after normalization, DEL norm is the maximum fatigue load on the key parts of each wind turbine generator after normalization, and α is the weight coefficient.

[0061] After establishing the multi-objective optimization function, the optimal yaw control angle that maximizes the objective function J under specific environmental working conditions is solved through an intelligent optimization algorithm, and the optimal wake center position in front of the rotor of each unit under this yaw control input is recorded through the aero-hydro-servo-elastic dynamics simulation model of the wind farm; finally, the above environmental wind condition parameters and the corresponding optimal wake center position information are saved as a multi-dimensional intelligent decision database for online real-time search of the optimal wake center position in front of each unit under different environmental wind conditions and field-level collaborative control.

[0062] Specifically, in step (4-3), the intelligent optimization algorithm includes, but is not limited to, genetic algorithm, particle swarm algorithm, and game theory algorithm.

[0063] In step (5), the offshore wind farm field-level collaborative PI controller includes a low-pass filter, a multi-dimensional intelligent decision database, a wake center position error calculation module, a PI controller, a yaw actuator, and a lidar wind measurement system.

[0064] Taking the measured values of the environmental wind conditions obtained by the lidar wind measurement system as the input, the average wind speed and direction of the external environmental wind input are obtained through a low-pass filter, and a multi-dimensional intelligent decision database is connected after the low-pass filter; the online search is realized by using the look-up table method to obtain the optimal wake center position information in front of each unit corresponding to the environmental wind conditions; the wake center position error calculation module is used to complete the error estimation of the actual wake center position and the expected wake center position in front of each unit; the center position deviation value obtained by the wake center position error calculation module is input into the PI controller, and the proportional and integral operations are performed on the deviation value to obtain the adjustment value for the active cooperative yaw control of the wind farm within a unit control period; the yaw misalignment degree between each unit and the environmental incoming wind is changed through the yaw actuator to achieve wake redirection and control optimization.

[0065] The error estimation of the actual wake center position and the expected wake center position in front of each unit completed by using the wake center position error calculation module specifically includes:

[0066] The measured wake center position information parameters are obtained through the three-dimensional wind field inversion of the lidar wind measurement system, the optimal wake center position information parameters are obtained by online searching the multi-dimensional intelligent decision database, and the center position deviation value is obtained by taking the difference between the two through the wake center position error calculation module.

[0067] Compared with the prior art, the present invention has the following beneficial effects:

[0068] 1. The present invention patent establishes a wake tracking module, which can complete the three-dimensional wind field inversion according to the real-time original data measured by the lidar wind measurement system, obtain the environmental wind condition parameters and the wake characteristic parameters in front of each wind turbine, so as to complete the real-time wake tracking, and can provide reliable measured input for the subsequent wake intelligent control.

[0069] 2. The present invention patent adopts a multi-objective optimization method, comprehensively considers the total output power of the offshore wind farm system and the fatigue load conditions of the key parts of the unit to establish a multi-objective optimization function, completes the prediction of the flow field distribution based on the offshore wind farm aerodynamic-hydro-servo-elastic dynamics simulation model, and uses the intelligent optimization algorithm to solve the optimal wake center position under different environmental wind conditions, and establishes a multi-dimensional intelligent decision database LUT. This intelligent database can be used for online search and rapid optimization of the offshore wind farm, and realizes the wake intelligent control of the wind farm system under changing environmental wind conditions.

[0070] 3. The present invention patent establishes a field-level collaborative PI controller, which uses the measured wake center position information obtained by the lidar wind measurement system as input and the optimal wake center position information obtained by searching the multi-dimensional intelligent decision database LUT as the target value, solves for the collaborative yaw control adjustment value within a unit control period, and issues it to each unit to achieve field-level collaborative control and wake redirection. The controller of the present invention has characteristics such as good control effect and fast execution speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 FIG. is the system structure diagram of a field-level collaborative control strategy for an offshore wind farm based on wake tracking according to the present invention;

[0072] Figure 2 FIG. is the structure diagram of the field-level collaborative PI controller for an offshore wind farm;

[0073] Figure 3 FIG. is the structure diagram of the PI controller;

[0074] Figure 4 FIG. is the establishment flow chart of the multi-dimensional intelligent decision database;

[0075] Figure 5 FIG. is the measured wind condition information of the lidar wind measurement system in the embodiment of the present invention;

[0076] Figure 6 FIG. shows the change in the center position of the wake received by the front sides of the WT2, WT3, WT5, and WT6 units of the aero-hydro-servo-elastic dynamics simulation model of the offshore wind farm in the embodiment of the present invention before and after optimization control;

[0077] Figure 7 FIG. shows the time-domain change in the operating performance of the offshore wind farm system before and after optimization control in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0078] The following further describes the present invention in detail with reference to the drawings and embodiments. It should be noted that the following embodiments are intended to facilitate the understanding of the present invention and do not limit it in any way.

[0079] As Figure 1 shown, the control system of the field-level collaborative control strategy for an offshore wind farm based on wake tracking includes a nacelle-mounted lidar wind measurement system, a wake tracking module, an optimizer, and a field-level controller. The lidar wind measurement system is used to complete the identification of environmental wind conditions and wake characteristic parameters. The optimal wake center position information reference value is obtained through the multi-dimensional intelligent decision database established in the optimizer. The PI collaborative controller is used to complete wake redirection and intelligent control through active yaw.

[0080] The specific steps are as follows:

[0081] Step 1: Establish a wake tracking module. Use the nacelle-mounted lidar wind measurement system of the offshore wind farm to measure the wind speed distribution in the front flow field of each offshore wind turbine. Based on the measured wind speed distribution data, complete the assessment of the wind information of the incoming flow without wake influence and the identification of wake characteristics under wake influence through three-dimensional wind field inversion, obtain wake parameter information and the wake center position, and complete real-time wake tracking.

[0082] Step 2: According to the layout characteristics and flow field distribution of the actual offshore wind farm, establish an aero-hydro-servo-elastic dynamics simulation model of the offshore wind farm, specifically including an aerodynamic simulation model of the wind farm, an offshore wind turbine model, a wake model, etc., and use the historical wind speed information measured by the above lidar to correct the model parameters, and use this model to predict the operating conditions of the wind farm under specific environmental input wind conditions and control strategies.

[0083] Step 3: The multi-dimensional intelligent decision-making database LUT of the offshore wind farm based on wake redirection is used to realize the online real-time search for the optimal control strategy under different environmental wind conditions. Establish a multi-objective optimization function that comprehensively considers the maximization of wind farm power and the minimization of loads. Through intelligent optimization algorithms, obtain the optimal wake center position information of the wind farm under different environmental wind condition inputs, and record the above parameter information as a multi-dimensional intelligent decision-making database.

[0084] Step 4: The field-level coordinated controller of the offshore wind farm uses the look-up table method to online search the multi-dimensional intelligent decision-making database LUT to obtain the optimal wake center position information of the wind farm under the corresponding wind condition input, and use it as the reference value of the field-level controller. According to the measured wake information value measured by the wake tracking module, change the yaw misalignment degree of each unit and the incoming environmental flow through the field-level coordinated PI controller, and realize wake redirection and control optimization.

[0085] As Figure 2 shown, the field-level coordinated PI controller of the offshore wind farm specifically includes a low-pass filter, a multi-dimensional intelligent decision-making database LUT, a wake center position error calculation module, a PI controller, a yaw actuator, and a wake tracking module based on the lidar wind measurement system, etc. Within a unit control period, measure the wind condition data through the lidar wind measurement system to complete three-dimensional wind field inversion, obtain the average wind speed and wind direction information of the environmental input wind, and input it into the multi-dimensional intelligent decision-making database LUT for online search to obtain the optimal wake center position information under the corresponding working conditions after low-pass filtering. Use the PI controller to complete wake redirection through the yaw actuator, so that the measured wake center position is consistent with the optimal wake center position information, and complete intelligent wake control.

[0086] As Figure 3As shown in the figure, the PI controller uses the above lidar wind measurement system and the multi-dimensional intelligent decision-making database LUT to solve the deviation value between the measured wake center position and the optimal wake center position, and obtains the adjustment amount of the field-level collaborative yaw control through proportional and integral operations to complete wake redirection.

[0087] As Figure 4 shown in the figure, based on the aero-hydro-servo-elastic dynamics simulation model of the offshore wind farm, the dynamic operation characteristics of the actual offshore wind farm under corresponding environmental conditions are predicted, and the mapping relationship model between different environmental condition parameters, unit control parameters, wind farm output power and unit load is obtained. A multi-objective optimization function considering reducing power loss and fatigue load is established, and the intelligent optimization algorithm is used to complete the optimization of the optimal yaw control strategy, record the optimal wake center position information under the corresponding control, and establish and save it as the multi-dimensional intelligent decision-making database LUT.

[0088] In a specific embodiment, the above offshore wind turbine model includes a mechanical model and a control model, etc., and different types of offshore wind turbines can be simulated by modifying the model parameters. In this embodiment, the NREL 5MW wind turbine is selected for simulation.

[0089] The above flow field model and wake model complete the prediction of the wind conditions received by each unit according to the input wind conditions information and wake propagation characteristics of the offshore wind farm environment, and the specific model parameters are corrected by the measured data obtained by the wind measurement lidar system. In this embodiment, the wind conditions information measured by the actual nacelle lidar is selected, and the measured original data is the wind conditions data of the nacelle lidar system installed on a unit in a wind farm measured on May 19, 2019 within 350 minutes. Based on the measured data, 3D wind field inversion is completed to obtain the line-of-sight wind speeds measured by different laser beams, and the time series of the wind speeds in front of the unit are solved to obtain the average wind speed and wind direction information of the environmental wind input to the wind farm. The time series of the wind information is as Figure 5 shown in the figure. It can be obtained from the above data processing that the average wind speed of the environmental wind input to the wind farm during the measurement time is 8.3892 m / s, and the average wind direction is 268.6361° (selecting the due north direction as 0°).

[0090] In a specific embodiment, the establishment of the above-mentioned offshore wind farm simulation environment is based on the characteristics of an actual offshore wind farm, specifically including the layout of wind turbines and the distribution of environmental wind resources, etc., so that the simulation results are close to the actual working conditions. Based on the above input wind condition information, the wind information measured by the wind lidar is used as the input working condition of the wind farm simulation model. A 2×3 offshore wind farm aero-hydro-servo-elastic dynamics simulation model is established, and the NREL 5MW wind turbine is selected for each unit model, and the rotor diameter of the unit is 126m. The operating characteristics of the wind farm within 800s are predicted through simulation, and specifically, the time-series changes of its output power and the time-series changes of the fatigue loads at key parts are calculated. Among them, an X-Y plane coordinate system is established at the hub height of each unit, and the layout position information of each unit in the offshore wind farm system is shown in Table 1 below. Among them, the 270° input wind direction is the OX axis direction (east-west direction), and the Y axis direction is the north-south direction.

[0091] Table 1 Layout positions of each unit in the offshore wind farm simulation model

[0092] Unit serial number Coordinate information (m) WT1 (0，-189) WT2 (630，-189) WT3 (1260，-189) WT4 (0，189) WT5 (630，189) WT6 (1260，189)

[0093] According to the time-domain change law of the obtained operating characteristics, the total output power of the wind farm system and the fatigue load conditions at key parts within 800s of a unit control period are solved. A multi-objective optimization function that comprehensively considers efficiency increase and load reduction is established as follows:

[0094]

[0095] In the formula, is the total power generation of the wind farm system after normalization, DEL norm is the maximum fatigue load suffered by the key parts of each wind turbine after normalization, and α is the weight coefficient. In this embodiment, the optimization weight coefficient α is selected as 0.6. The intelligent optimization algorithm is used to complete the optimization of the optimal yaw control angle to maximize the result of this objective function. Specifically, the genetic algorithm is selected in this embodiment to complete the optimization of the optimal solution. Observe the influence of the upstream wake on the front side of each unit in the offshore wind farm system under the corresponding environmental input wind conditions and active yaw coordinated control, record the wake center position information received by each unit, and save it as a multi-dimensional intelligent decision database LUT to achieve online real-time search and optimization.

[0096] Specifically, in this embodiment, simulation and multi-objective optimization control are carried out under the wind conditions measured by the lidar wind measurement system. The optimal wake center position information corresponding to the working conditions in the multi-dimensional intelligent decision-making database LUT is searched online, and wake redirection and intelligent control are completed through the field-level PI collaborative controller. Under closed-loop control, the optimal wake center position obtained by online search is used as the reference value of the system, and the measured wake center position obtained by three-dimensional wind field inversion is used as the observed value. Each unit in the wind farm system realizes intelligent control of the wake by actively and collaboratively yawing to track the optimal wake center position information in real time.

[0097] As Figure 6 shown, the changes in the center position of the wake received by the front sides of the WT2, WT3, WT5, and WT6 units in the aero-hydro-servo-elastic dynamics simulation model of the offshore wind farm before and after optimization control are presented. The vertical coordinate is the change in the Y-axis coordinate value of the center position of the wake received by the front sides of each unit in the XY coordinate system. The center position of the wake in front of each unit shows a dynamic meandering state downstream of the upstream wind turbine rotor, and the farther downstream the unit is, the more significant the wake meandering phenomenon is.

[0098] Figure 7 It shows the time-domain changes in the operating performance of the offshore wind farm system before and after optimization control, specifically including the total output power, the time-domain changes in the tower pitch moment, and the time-domain changes in the blade pitch moment, etc. Due to the wake delay propagation effect, the offshore wind farm system reaches a stable state approximately after 450 s. Therefore, this embodiment mainly focuses on the results within the time range of 450 s - 800 s. Through calculation, it can be obtained that the total output power of the wind farm system after optimization control can be increased by 10.33%, the tower fatigue load can be reduced by 9.74%, while the blade fatigue load slightly increases. Generally speaking, the optimization control strategy described in the present invention can significantly achieve multi-objective optimization of increasing efficiency and reducing load.

[0099] The above-described embodiments have elaborated in detail the technical solutions and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the present invention. Any modification, supplement, and equivalent replacement made within the scope of the principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A field-level collaborative control strategy for an offshore wind farm based on wake tracking, characterized in that, Including: (1) Measuring the wind speed distribution data of the flow field in front of each offshore wind turbine by using the nacelle lidar wind measurement system; (2) Establishing a wake tracking module, and using the wind speed distribution data obtained in step (1) to complete the evaluation of the wind information of the incoming wind without wake influence and the identification of the wake characteristic parameters under the influence of the wake through three-dimensional wind field inversion. The wake characteristic parameters include wake depth, wake width, and wake center position; (3) According to the layout characteristics and flow field distribution of the actual offshore wind farm, establishing an aero-hydro-servo-elastic dynamics simulation model of the offshore wind farm; using the historical wind speed distribution data measured by the lidar wind measurement system to correct the parameters of the simulation model, and using the simulation model to predict the operating conditions of the wind farm under the input wind conditions and control strategies; The aero-hydro-servo-elastic dynamics simulation model of the offshore wind farm specifically includes a wind farm aerodynamic simulation model, an offshore wind turbine model, and a wake model; the wind farm aerodynamic simulation model uses the Taylor turbulence freezing hypothesis to complete the dynamic propagation simulation of the three-dimensional flow field of the wind farm in space; the offshore wind turbine model adopts a brake disc model, and completes the modeling according to the performance parameters of the actual unit and obtains the dynamic response of the unit power and load under specific environmental conditions; the wake model is used to predict the key wake characteristics related to the output power of the wind farm and the load of the unit, including wake wind speed deficit, wake expansion, and wake meandering; (4) Establishing a multi-dimensional intelligent decision-making database for the offshore wind farm based on wake redirection, which is used to realize the online real-time search for the optimal control strategy under different environmental wind conditions; Establishing a multi-objective optimization function that comprehensively considers the maximization of wind farm power and the minimization of load, obtaining the optimal wake center position information of the wind farm under different environmental wind condition inputs through an intelligent optimization algorithm, and recording the above parameter information into the multi-dimensional intelligent decision-making database; (5) Establishing a field-level cooperative PI controller for the offshore wind farm, using the look-up table method to online search the multi-dimensional intelligent decision-making database to obtain the optimal wake center position information of the wind farm under the corresponding wind condition input, taking it as the reference value of the field-level cooperative PI controller, and according to the measured wake information value measured by the wake tracking module, changing the yaw misalignment degree of each unit and the incoming environmental wind through the field-level cooperative PI controller to achieve wake redirection and control optimization.

2. The field-level cooperative control strategy for an offshore wind farm based on wake tracking according to claim 1, wherein The specific process of step (2) is as follows: (2-1) Three-dimensional wind field inversion In order to accurately evaluate the wind speed distribution information inside the offshore wind farm, first perform the northward correction of the wind direction, and use the coordinate rotation transformation matrix algorithm to invert the pointing direction of the laser beam in the earth coordinate system; the coordinate transformation formula of the laser beam pointing is: L LOS =(H1H2H3) -1 L LOS_Lidar (1) where L LOS is the direction of the laser beam in the Earth coordinates, and H1, H2, H3 are coordinate rotation transformation matrices, and L LOS_Lidar is the direction of the laser beam in the aircraft cabin coordinates; Define the directions of the nacelle coordinate system: X0 points axially along the rotor towards the nacelle head, Y0 points radially along the rotor towards the left horizontally, and Z0 is coaxial with the wind turbine tower axis and points towards the nacelle top; the direction of the laser beam is described by the azimuth angle and the pitch angle θ0, where the azimuth angle is the angle between the projection of the laser beam on the X0Y0 plane and X0, and the pitch angle θ0 is the angle between the laser beam and the X0Y0 plane; the attitude changes of the wind turbine nacelle during the coordinated control process are represented by the roll angle , the pitch angle θ, and the yaw angle ψ; therefore, the direction L LOS_Lidar of the laser beam in the nacelle coordinate system and the coordinate rotation transformation matrix are expressed as follows: Further solving gives the line-of-sight wind speed v of the laser beam LOS : v LOS = v wind ·L LOS (6) When performing actual marine nacelle lidar scanning measurement, at the same height and circumferential scanning layer, the line-of-sight wind speed is a trigonometric function under ideal conditions and is defined as: f(v LOS ) = acos(θ - b) + c (7) The parameters a, b, and c are determined by the wind vector information at the measured height. Specifically: Among them, v1, v2, and v3 are the component velocities of the measured wind farm in the north-south direction, east-west direction, and vertical direction in the earth coordinate system, which determine the parameter sizes of the above line-of-sight wind speed function; The line-of-sight wind speed and wind vector information described by equations (7) and (8) are obtained through least-squares fitting, including: by minimizing the function error in equation (9), the parameters a, b, and c of the curve function are solved, and the inversion of the wind speed distribution of the three-dimensional wind farm is completed; where N is the number of test points of the nacelle-mounted lidar in a certain measurement period; based on the above wind field inversion, the wind information of the incoming flow of the wind farm without wake influence is obtained, including wind speed, wind direction, and turbulence intensity, and at the same time, a data basis is provided for the identification of wake characteristic parameters; (2-2) Identification of wake characteristic parameters An inverse Gaussian function of the wake velocity distribution received by the wind turbine is established, and the formula is as follows: Among them, U is the wind speed at the position of (x, z h ) at the hub height of the wind turbine at time t, U0 is the ambient average wind speed at time t, C s is the wake proportion coefficient used to quantify the wake depth, e is the natural constant, μ is the wake center position at the hub height, σ is the wake standard deviation used to quantify the wake width, and π is the pi; Using the wind speed data measured by the lidar at the corresponding distance in front of the wind turbine, the parameter fitting of the inverse Gaussian function is completed by the least-squares method, and the wake characteristic parameters are obtained, specifically including wake depth, wake width, and wake center position information.

3. The field-level cooperative control strategy for an offshore wind farm based on wake tracking according to claim 1, wherein In step (4), the specific process of establishing the multi-objective optimization function that comprehensively considers the maximization of wind farm power and the minimization of load is as follows: (4-1) Calculate the output power of the offshore wind turbine generator set i: Wherein, P i is the output power of the generator of Unit i in the wind farm, T q,i (t) is the instantaneous torque of the generator of Unit i at time t, ω i is the instantaneous speed of the generator of Unit i at time t, t k is the current time, and ΔT is the current solution time domain period; According to the power calculation formula of a single unit above, the total power generation of the wind farm system is solved as the sum of the power generations of all wind turbines in the wind farm system It is as follows: (4-2) Calculate the equivalent fatigue load received by the wind turbine generator set i: Using the aero-hydro-servo-elastic dynamics simulation model of the wind farm to obtain the time-domain variation of the loads received by the key parts of the wind turbine generator set i in the wind farm, including the load variations at the bottom of the tower and the root of the blade; through the rainflow counting method, the time-domain varying loads received by the wind turbine generator set i are equivalently analyzed to obtain the equivalent fatigue load DEL of the key parts; where DEL is the equivalent fatigue load, is the equivalent number of cycles in time series j, N j,i is the number of occurrences of the i-th working condition in time series j, L j,i is the load range of the i-th working condition in time series j, and m is the slope of the material S-N curve; (4-3) To maximize the total output power of the wind farm and reduce the fatigue load received by the key parts of the unit, a multi-objective optimization function is established as follows: In the formula, is the total generated power of the wind farm system after normalization, and DEL norm is the maximum fatigue load on the key parts of each wind turbine generator after normalization, and α is the weight coefficient.

4. The field-level cooperative control strategy for an offshore wind farm based on wake tracking according to claim 1, characterized in that In step (5), the offshore wind farm field-level cooperative PI controller includes a low-pass filter, a multi-dimensional intelligent decision database, a wake center position error calculation module, a PI controller, a yaw actuator, and a lidar wind measurement system; Taking the environmental wind condition measurement value obtained by the lidar wind measurement system as the input, the average wind speed and wind direction of the external environmental wind input are obtained through the low-pass filter, and a multi-dimensional intelligent decision database is connected after the low-pass filter; the online search is realized by using the look-up table method to obtain the optimal wake center position information in front of each unit under the corresponding environmental wind condition; the error estimation of the actual wake center position and the expected wake center position in front of each unit is completed by using the wake center position error calculation module; the center position deviation value obtained by the wake center position error calculation module is input into the PI controller, and the deviation value is subjected to proportional and integral operations to obtain the adjustment value of the active cooperative yaw control of the wind farm in a unit control period; the yaw misalignment degree between each unit and the incoming environmental wind is changed through the yaw actuator to realize wake redirection and control optimization.

5. The field-level cooperative control strategy for an offshore wind farm based on wake tracking according to claim 4, characterized in that The error estimation of the actual wake center position and the expected wake center position in front of each unit completed by using the wake center position error calculation module specifically includes: The measured wake center position information parameters are obtained through the three-dimensional wind field inversion of the lidar wind measurement system, the optimal wake center position information parameters are obtained by searching the multi-dimensional intelligent decision-making database online, and the deviation value of the center position is obtained by taking the difference between the two through the wake center position error calculation module.

Images