Gradient optimization control method and system for wake flow of wind power plant

Abstract

The invention discloses a wind power plant wake flow echelon optimization control method and system, and the method comprises the steps: collecting flow field wind condition information, and determining a data collection interval according to the flow field wind condition information; in each preset fatigue monitoring period, regularly collecting the yaw working time of each unit to monitor the fatigue state: obtaining the maximum value in the accumulated yaw working time of each unit, and judging whether the maximum value exceeds the fatigue time threshold value of the fatigue monitoring period or not; if yes, the yaw angles of all the units return to zero; if not, whether the flow field wind condition information meets the starting condition of the wake flow deflection optimization solver or not is judged; if the starting condition is met, a new yaw angle of the whole set is calculated and output, and otherwise, the yaw angle of the whole set is kept unchanged. According to the method, optimization of the whole-field generating capacity is considered, the working characteristics and the fatigue degree of the yaw system of the unit are taken into consideration, and the working load of the yaw system can be effectively reduced while stable improvement of the whole-field generating capacity can be guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of wind farm wake control, and in particular to a hierarchical optimization control method and system for wind farm wakes. Background Art

[0002] During the operation of a wind farm, the wake effect severely restricts the power generation efficiency of the wind farm and the service life of the turbines. To effectively solve this problem, a variety of wake control technologies have been developed, among which the wake deflection technology and the axial flow induction control technology are prominent. In terms of the current research situation and popularity, since the wake deflection technology can significantly improve the overall power generation of wind farms that are greatly affected by wakes, the attention paid to it exceeds that of the axial flow induction control technology. The core principle of the traditional steady-state wake deflection technology is to adjust the yaw angle of the upstream wind turbine so that the wake generated by the wind turbine deviates from the downstream wind turbine, thereby reducing the power generation power loss of the downstream turbines caused by the wake effect and achieving the maximization of the overall power generation of the whole field. In actual operation, the steady-state wake deflection technology calculates the optimal yaw angle of the wind turbine based on a pre-set look-up table, combined with environmental parameters such as the layout, wind direction, and wind speed of the wind farm, and then conducts unified yaw control on the wind turbine. However, this technology has obvious limitations. It only focuses on a single indicator of the overall power generation of the whole field, and completely ignores the working characteristics, fatigue degree of the yaw system of the turbine and the changes in external working conditions, etc.

[0003] In the "Wind Farm Optimization Scheduling Method and System Based on Wind Turbine Wake Model Optimization" disclosed in the Chinese patent literature, with the publication number CN114169614B and the publication date of December 13, 2022, SCADA data is used to analyze the correction of the wake model, and combined with an intelligent optimization algorithm, aiming at minimizing the power calculation error of the wind farm, the key parameters of the model are optimized, so that the proposed wake optimization method can fully consider the actual situation of the wind farm to customize the parameters of the model. After optimizing the wake model using this method, the calculation accuracy of the model in an actual wind farm can be greatly improved, so as to more accurately model the wake effect in the wind farm, significantly improve the power prediction accuracy of the wind farm and the reliability of the wake control strategy, and then based on the optimized wake model, improve the overall power generation efficiency and overall power generation of the wind farm through wake optimization control and other means. However, this technology only focuses on a single indicator of the overall power generation of the whole field, and completely ignores the working characteristics, fatigue degree of the yaw system of the turbine and the changes in external working conditions, etc., which easily makes the yaw system operate at a high load for a long time, accelerates the fatigue damage of components, greatly shortens the service life of the yaw system, and increases the operation and maintenance costs and safety risks of the wind farm. Summary of the Invention

[0004] The present invention aims to overcome the problem in the prior art that only a single index of the total power generation of the whole field is concerned, without considering the working characteristics, fatigue degree, etc. of the yaw system of the unit, which easily causes the yaw system to be in a high-load operation state for a long time, accelerates the fatigue damage of components, greatly shortens the service life, and increases the operation and maintenance costs and safety risks of the wind farm. A hierarchical optimization control method and system for the wake of a wind farm are provided.

[0005] To achieve the above object, the present invention adopts the following technical solutions: A hierarchical optimization control method for the wake of a wind farm, comprising: Collecting the flow field wind condition information in the wind farm area, and determining the data collection interval according to the flow field wind condition information; During each preset fatigue monitoring period, regularly collecting the yaw working time of each unit for fatigue state monitoring: obtaining the maximum value in the cumulative yaw working time of each unit, and judging whether it exceeds the fatigue time threshold of this fatigue monitoring period; If so, the yaw angles of all units are reset to zero; If not, judging whether the flow field wind condition information meets the startup condition of the wake deflection optimization solver; if the startup condition is met, calculating and outputting the new yaw angles of all units in the field, otherwise maintaining the yaw angles of all units in the field unchanged.

[0006] In the present invention, not only the optimization of the total power generation of the whole field is considered, but also the working characteristics and fatigue degree of the yaw system of the unit are taken into account. By real-time monitoring the operating state of the wind turbine, the working parameters of the yaw system, and the environmental changes of the wind farm, the yaw angle of the wind turbine can be dynamically adjusted according to the actual situation. While ensuring the stable increase of the total power generation of the whole field, the working load of the yaw system is effectively reduced, the service life of the yaw system is extended, and a more reliable guarantee is provided for the efficient and stable operation of the wind farm. At the same time, the data collection interval is adaptively adjusted according to the actual flow field wind condition information to ensure that the operation of each unit better fits the real-time changes of the whole wind farm flow field; this adaptive data collection interval setting mode greatly enhances the flexibility of the wind condition data collection input and provides a comprehensive, real-time and accurate data basis for the wake deflection system.

[0007] Preferably, after the monitoring of each fatigue monitoring period is completed, the cumulative yaw working time of all units is cleared; Calculating the difference between the maximum cumulative yaw working time and the fatigue time threshold in this fatigue monitoring period; Taking the standard fatigue time threshold minus the difference as the fatigue time threshold of the next fatigue monitoring period.

[0008] Preferably, the judgment of whether the flow field wind condition information meets the startup condition of the wake deflection optimization solver includes: Judge whether the wind speed in the field area is greater than the cut-in wind speed of all units. If not, keep the yaw angles of all units in the whole field unchanged and turn off the wake deflection optimization solver; If so, judge whether the change of the wind direction in the field area is greater than the wind direction change threshold; if it is greater, start the wake deflection optimization solver and output the new yaw angles of all units in the whole field, otherwise keep the yaw angles of all units in the whole field unchanged.

[0009] Preferably, when the change of the wind direction in the field area is less than or equal to the wind direction change threshold, keep the yaw angles of all units in the whole field unchanged and turn off the wake deflection optimization solver; Judge whether the total power generation in the whole field is greater than the original reference value without wake deflection control; If it is greater, keep the yaw angles of all units in the whole field unchanged, otherwise start the wake deflection optimization solver and output the new yaw angles of all units in the whole field.

[0010] Preferably, the start-up times of the wake deflection optimization solver are negatively correlated with the data acquisition interval, negatively correlated with the wind direction change threshold, and positively correlated with the cumulative yaw working time.

[0011] Preferably, the acquisition of the flow field wind condition information in the wind farm field area includes: determining the first-order upstream units under each wind direction at the minimum wind direction change interval; determining the wind direction set corresponding to each first-order unit; Regularly obtain the wind directions measured by all first-order units and calculate the average wind direction; Determine the wind direction set into which the average wind direction falls, and use the wind measurement information of the first-order units corresponding to this wind direction set as the flow field wind condition information.

[0012] Preferably, the data acquisition interval is greater than the yaw response execution time of all units in the whole field and greater than the maximum wake conduction time in the determined wind direction in the flow field wind condition information; The yaw response execution time is the single maximum allowable yaw range divided by the yaw speed; The maximum wake conduction time is the maximum wake conduction distance in the determined wind direction divided by the average flow velocity of the flow field.

[0013] Preferably, the maximum wake conduction time is: establish a rotating wind direction coordinate system with the determined wind direction as the positive x-axis direction; Select the upstream first row of units and the downstream last row of units perpendicular to the determined wind direction; Calculate the x-axis coordinate vector of the last row of units, and subtract the minimum x-axis coordinate value in the first row of units to obtain the wake conduction distance vector; Divide the wake conduction distance vector by the average flow velocity of the flow field to obtain the wake conduction time vector, and use the maximum element in this vector as the maximum wake conduction time.

[0014] Preferably, the standard fatigue time threshold is the duration of the fatigue monitoring period multiplied by a threshold coefficient; The threshold coefficient is determined by the ratio of the maximum allowable working time within the designed service life cycle of the yaw system of the unit to its designed service life cycle.

[0015] A hierarchical optimization control system for the wake of a wind farm includes a central controller for controlling the wake deflection of the yaw systems of all units in the wind farm area; the central controller includes: A flow field wind condition identification module for dynamically identifying the free incoming flow direction of the wind farm; A data acquisition and processing module for collecting the flow field wind condition information in the area and processing the data; A wake deflection control module for controlling the start or stop of the wake deflection optimization solver and issuing wake deflection commands for all units in the whole field according to the set values of the yaw angles of all units in the whole field; A wake deflection optimization solver for optimizing and solving the yaw angles of all units in the whole field under different wind conditions.

[0016] The present invention has the following beneficial effects: It not only considers the optimization of the total power generation of the whole field, but also takes into account the working characteristics and fatigue degree of the yaw system of the unit. By real-time monitoring the operating state of the wind turbine, the working parameters of the yaw system, and the environmental changes of the wind farm, it can dynamically adjust the yaw angle of the wind turbine according to the actual situation. While ensuring the stable increase of the total power generation of the whole field, it effectively reduces the working load of the yaw system, extends the service life of the yaw system, and provides a more reliable guarantee for the efficient and stable operation of the wind farm. Description of the Drawings

[0017] Figure 1 is a flowchart of the hierarchical optimization control method for the wake of the wind farm in the present invention.

[0018] Figure 2 is a comparison diagram of the working time and action times of the yaw system between the control method of the present invention and other control methods.

[0019] Figure 3 is a comparison diagram of the power generation increase between the control method of the present invention and other control methods. Detailed Embodiments

[0020] The following further describes the present invention in conjunction with the drawings and specific embodiments.

[0021] The core principle of traditional steady-state wake deflection technology is to adjust the yaw angle of the upstream wind turbine, so that the wake generated by the wind turbine deviates from the downstream wind turbine, thereby reducing the power generation loss of the downstream unit caused by the wake effect, and achieving the maximization of the full-field power generation. It works closely around the goal of optimizing the full-field power generation. By changing the wake diffusion direction of the upstream unit and optimizing the airflow field distribution of the full-field units, the downstream units can operate in a better wind energy environment, and their power generation also increases significantly accordingly.

[0022] However, the steady-state wake deflection technology has obvious limitations. It only focuses on the single index of full-field power generation, and completely ignores the working characteristics, fatigue degree of the yaw system of the unit and the changes in external working conditions. During the long-term operation of the wind turbine's yaw system, frequent and fixed yaw operations will cause the yaw components to bear large mechanical stresses and wear. Since the steady-state wake deflection technology adopts a fixed yaw strategy and cannot be flexibly adjusted according to the real-time state of the yaw system and external working conditions, this makes the yaw system in a high-load operation state for a long time, accelerating the fatigue damage of the components, greatly shortening the service life of the yaw system, and increasing the operation and maintenance costs and safety risks of the wind farm.

[0023] To solve the above problems, the present invention provides a Figure 1 hierarchical optimization control method for the wake of a wind farm as shown in Collect the flow field wind condition information in the wind farm area, and determine the data collection interval according to the flow field wind condition information; During each preset fatigue monitoring period, regularly collect the yaw working time of each unit for fatigue state monitoring: obtain the maximum value in the cumulative yaw working time of each unit, and judge whether it exceeds the fatigue time threshold of this fatigue monitoring period; If so, reset the yaw angle of all units to zero; If not, judge whether the flow field wind condition information meets the start condition of the wake deflection optimization solver; if the start condition is met, calculate and output the new yaw angle of the full-field units, otherwise keep the yaw angle of the full-field units unchanged.

[0024] It should be noted that in the present invention, not only the optimization of the overall power generation is considered, but also the working characteristics and fatigue degree of the yaw system of the unit are taken into account. By real-time monitoring the operating state of the wind turbine, the working parameters of the yaw system, and the environmental changes of the wind farm, the yaw angle of the wind turbine can be dynamically adjusted according to the actual situation. While ensuring the stable increase of the overall power generation of the whole field, the working load of the yaw system can be effectively reduced, the service life of the yaw system can be extended, and more reliable guarantee for the efficient and stable operation of the wind farm is provided. At the same time, the data acquisition interval is adaptively adjusted according to the actual flow field wind condition information to ensure that the operation of each unit is more in line with the real-time changes of the whole wind farm flow field; this adaptive data acquisition interval setting mode greatly enhances the flexibility of the wind condition data acquisition input and provides a comprehensive, real-time and accurate data basis for the wake deflection system.

[0025] It is worth noting that the wake deflection system control method designed in the present invention monitors the incoming flow wind condition of the field area in real time, adaptively adjusts the data acquisition interval according to different incoming flow wind conditions, and actively turns on or off the wake deflection optimization solver according to the wind direction change situation and the fatigue state of the yaw systems of all the units in the whole field. By giving up some optimal solutions and instead seeking sub-optimal solutions, the index balance between the maximum overall power generation and the minimum fatigue of the unit equipment is achieved.

[0026] Specifically, step S1: The flow field wind condition recognition module and the data acquisition and processing module are both in the active state, collect the flow field wind condition information in the wind farm area, and determine the data acquisition interval according to the flow field wind condition information; the two cooperate to complete the collection of the flow field wind condition information.

[0027] Step S2: Preset the fatigue monitoring period of the yaw system, and determine the standard fatigue time threshold according to the time length of the fatigue monitoring period, and regularly (according to the data acquisition interval) monitor the fatigue state of the yaw system; when a fatigue monitoring period ends, perform the following operations: S201: Clear the accumulated yaw working time of all the units in the field area; S202: Calculate the difference between the maximum accumulated yaw working time of this round of fatigue monitoring period and the fatigue time threshold (i.e., the maximum allowable working time within a fatigue monitoring period); S203: Subtract the above difference from the standard fatigue time threshold to calculate the fatigue time threshold of the yaw system in the next round of fatigue monitoring period.

[0028] If the fatigue monitoring period has not ended, go to the next step S3.

[0029] Step S3: Query the cumulative yaw working time of each unit and calculate the maximum value among them. The yaw working time refers to the time required for each unit to yaw to the predetermined target position. Calculate the cumulative yaw working time of each unit and select the maximum value from the cumulative yaw working times of all units. Determine whether this maximum value exceeds the fatigue time threshold of the yaw system; if the determination result is yes, set the yaw angle of all units in the field to zero, turn off the wake deflection optimization solver, and enter Step S7; if the determination result is no, then enter Step S4.

[0030] Step S4: Query whether the wind speed in the field area at this moment is greater than the cut-in wind speed of all units. If the determination result is no, then the yaw angle of all units in the field remains unchanged, maintain the previous state value, turn off the wake deflection optimization solver, and enter Step S7; if the determination result is yes, then enter Step S5.

[0031] Step S5: Query whether the wind direction change in the field area at this moment is greater than the set wind direction change threshold. If the determination result is yes, then execute Step S6; if the determination result is no, then the yaw angle of all units in the field remains unchanged, maintain the previous state value, and perform the following operations: S501: Turn off the wake deflection optimization solver; S502: Calculate whether the total power generation in the field is greater than the original reference value without wake deflection control. If the determination result is yes, then output the yaw angle of all units in the field and execute Step S7; otherwise, execute Step S6. The original reference value refers to the total power generation of all wind turbines obtained by summing the power generation of each wind turbine, ignoring the wake effect of upstream units on downstream units and aiming at the optimal power generation of each unit rather than the optimal power generation of the whole field.

[0032] Step S6: Start the wake deflection optimization solver and output the yaw angle of all units in the field.

[0033] Step S7: The yaw system of all units in the field executes the yaw command and returns to Step S1 to prepare for the next cycle of fatigue monitoring.

[0034] As a specific embodiment, after each fatigue monitoring cycle is completed, clear the cumulative yaw working time of all units; Calculate the difference between the maximum cumulative yaw working time and the fatigue time threshold within this fatigue monitoring cycle; Use the standard fatigue time threshold minus the difference as the fatigue time threshold for the next fatigue monitoring cycle.

[0035] It should be noted that when calculating the difference between the maximum cumulative yaw working time and the fatigue time threshold within the fatigue monitoring period, there is a positive or negative sign; if the difference is positive, it means that the maximum cumulative yaw working time of the current yaw system exceeds the fatigue time threshold, and the excess part is regarded as over-drawing the fatigue time threshold tolerance of the next round and needs to be repaid within the next fatigue monitoring period; conversely, if the difference is negative, it means that the maximum cumulative yaw working time does not exceed the fatigue time threshold, and the corresponding threshold tolerance compensation can be obtained within the next fatigue monitoring period. The standard fatigue time is determined according to the time length of the fatigue monitoring period, which reflects the maximum allowable working time of the yaw system within the fatigue monitoring period. In the first fatigue monitoring period, its fatigue time threshold is the standard fatigue time threshold, and the fatigue time threshold in subsequent fatigue monitoring periods needs to be calculated based on the dynamic difference.

[0036] It is worth noting that the adaptive allocation of the fatigue time threshold across fatigue monitoring periods in the present invention can be calculated based on the dynamic difference between the real-time cumulative yaw working time and the standard fatigue time threshold, and adopts a two-way adjustment mechanism of over-drawing repayment and surplus compensation. It not only allows the fatigue balance to be achieved through the reduction of the fatigue time threshold in subsequent fatigue monitoring periods after the current fatigue monitoring period exceeds the threshold, but also can convert the unexhausted fatigue time threshold into the incremental tolerance of the next fatigue monitoring period. Thus, on the premise of ensuring system safety, the short-term high-load impact can be effectively alleviated through dynamic allocation, and at the same time, the overall fatigue fault tolerance of the system can be improved by using the threshold capacity compensation mechanism, and finally, the fatigue flexible regulation and optimal management of the yaw system throughout its life cycle can be realized.

[0037] As a specific embodiment, determining whether the flow field wind condition information meets the startup condition of the wake deflection optimization solver includes: determining whether the wind speed in the field area is greater than the cut-in wind speed of all wind turbines. If not, the yaw angles of all wind turbines in the whole field are maintained unchanged, and the wake deflection optimization solver is turned off; If so, it is determined whether the change in the wind direction in the field area is greater than the wind direction change threshold; if it is greater, the wake deflection optimization solver is started and the new yaw angles of all wind turbines in the whole field are output, otherwise the yaw angles of all wind turbines in the whole field are maintained unchanged.

[0038] Furthermore, in the case where the change in the wind direction in the field area is less than or equal to the wind direction change threshold, the yaw angles of all wind turbines in the whole field are maintained unchanged, and the wake deflection optimization solver is turned off; then it is determined whether the total power generation in the whole field is greater than the original reference value without wake deflection control; if it is greater, the yaw angles of all wind turbines in the whole field are maintained unchanged, otherwise the wake deflection optimization solver is started and the new yaw angles of all wind turbines in the whole field are output.

[0039] It should be noted that the wake deflection control of the present invention has remarkable efficiency and balance. By real-time monitoring of the flow field wind conditions in the wind farm, the data acquisition interval can be adaptively adjusted according to the response time of the whole-field system. At the same time, combined with the fatigue state of the yaw systems of all the turbines in the field, the wake deflection optimization solver is actively turned on or off. During the optimization process, some optimal solutions can be cleverly abandoned and sub-optimal solutions are sought instead, thus achieving a good balance between maximizing the overall power generation of the field and minimizing the fatigue of the turbine equipment. In addition, compared with the traditional wake deflection control method, the solution rate of the present invention is higher, and the optimization calculation can be completed in a shorter time, effectively improving the overall operation efficiency and equipment service life of the wind farm, and providing strong support for the intelligent and efficient operation of the wind farm.

[0040] Specifically, after the determination condition in step S3 is "yes", after step S4 is determined to be "no", and after step S5 is determined to be "no", the central controller will give up starting to solve for the optimal yaw angle of the whole field and instead set the yaw angle of the whole field to 0 or maintain the previous state value (seeking sub-optimal solutions). Especially when the determination in step S502 is yes, even if the previous state value is maintained (seeking sub-optimal solutions), the optimized overall power of the field is still greater than the reference value without using the wake control method currently, so there is no need to start the wake deflection optimization solver.

[0041] Due to the lack of the top-level supervision and control method of the present invention in the traditional wake deflection control method, for a wind farm with real-time control, the wake deflection optimization solver will be frequently started for calculation, calculating some tasks that can be completely avoided. For example, in step S4, if the wind speed is lower than the cut-in wind speed of the turbine at a certain moment, but the change in the wind direction of the whole field is greater than the set threshold, the wake deflection optimization solver will still start the calculation. The consequences of starting the calculation are: 1. The turbine executes the yaw command but the turbine will not generate electricity (the wind speed is too small); 2. Futilely increasing the yaw actions of the yaw system of the turbine, increasing the fatigue of the turbine; 3. Wasting the computing resources of the computer, and the wake deflection optimization requires consuming resources and time. This will lead to a longer waiting time before the control task gives the execution command to the yaw mechanism.

[0042] As a specific embodiment, collecting the flow field wind conditions in the wind farm area includes: determining the first-order upstream turbines under each wind direction at the minimum wind direction change interval; determining the wind direction set corresponding to each first-order turbine; Regularly obtaining the wind directions measured by all the first-order turbines and calculating the average wind direction; Determining the wind direction set into which the average wind direction falls, and using the wind measurement information of the first-order turbines corresponding to this wind direction set as the flow field wind conditions information.

[0043] It should be noted that as a prerequisite for the wake deflection collaborative control technology, the incoming flow wind condition in the field area is an important boundary input condition, which reflects the real-time distribution of the flow field in the wind farm area. The flow field wind condition information mainly includes the following time-series variables: wind speed, wind direction, turbulence, air density, wind shear, etc. These information jointly form the time-series matrix of the flow field wind condition information. Among them, the wind direction, as the core variable of the wake deflection collaborative control, determines the upstream and downstream sorting rules of all the turbines in the field.

[0044] Using the Cartesian coordinate system, establish the absolute geographical coordinate system of the entire field turbine positions and the rotating wind direction coordinate system respectively. The x-axis direction of the rotating wind direction coordinate points to the wind direction (by default, the due north is 0°, and the due east is 90°), and the y-axis is orthogonal to it. Assume that the geographical coordinates of turbine i are [x i , y i , and the wind direction change angle is w d , then its position coordinates [x rot,i , y rot,i in the rotating wind direction coordinate system can be calculated as follows: x rot,i = -(x i sinw d + y i cosw d ), y rot,i = -x i cosw d + y i sinw d . For the same physical turbine position, different wind direction angles will determine different wind direction rotation coordinate positions [x rot,i , y rot,i of the turbine. Therefore, in the x-axis direction of the wind direction, there is an upstream and downstream sorting problem of the x rot,i of each turbine position in each rotating wind direction coordinate system. Due to the influence of the wake effect of the upstream turbine, the wind condition measured by the downstream turbine cannot truly reflect the free flow distribution of the flow field in the wind farm area. Therefore, how to determine and obtain the wind condition information of the first-order turbines is crucial.

[0045] Specifically, by applying a 360° wind direction change range and with the minimum wind direction change interval (such as 1°), find the first-order turbines under each wind direction and make calibrations; determine the wind direction sets of all the first-order turbines; regularly obtain the wind directions measured by all the first-order turbines, and perform average calculations, and determine which wind direction set the average wind direction falls into; use the wind measurement information of the first-order turbine belonging to this wind direction set as the boundary condition input for the wake deflection collaborative control of the wind farm at this time, which is the flow field wind condition information.

[0046] For the determination of the first-order turbines, when a certain wind direction angle w d, First, calculate the rotational wind direction coordinates of all the turbines in the whole field at this wind direction angle; second, sort the x rot,i values of all the turbines in the whole field to find the x rot,i with the minimum coordinate value, thereby confirming the first-order turbine at wind direction w d . For the determination of the wind direction set of the first-order turbine, assuming that the minimum wind direction change interval is 1°, then there are a total of 360 w d , and the corresponding 360 first-order turbines. It should be noted that the first-order turbine can appear repeatedly at different wind directions w d ; therefore, each first-order turbine may correspond to multiple wind directions, and the set composed of these wind directions is the wind direction set corresponding to the first-order turbine.

[0047] It should be noted that the dynamic identification of the flow field wind conditions of the present invention has significant advantages in global accuracy and effectiveness. By obtaining the turbine numbers of the first-order turbines at any wind direction angle, the free flow wind condition information under a uniform flow field can be accurately obtained. This information is the key basis for solving wake deflection, and its accuracy directly affects the global optimization effect of wake deflection control. During the operation of a wind farm, the dynamic changes in wind conditions pose high requirements for the real-time performance and accuracy of wake deflection control. The present invention can capture the changes in wind direction and wind speed in real time, ensuring that the wake deflection control is always optimized based on the latest flow field state. This dynamic identification ability not only improves the adaptability of wake deflection control but also provides strong support for the intelligent operation of the wind farm, effectively improving the overall power generation efficiency and equipment service life of the wind farm.

[0048] As a specific embodiment, the startup times of the wake deflection optimization solver are negatively correlated with the data acquisition interval, negatively correlated with the wind direction change threshold, and positively correlated with the cumulative yaw working time.

[0049] It should be noted that different from the traditional steady-state wake deflection technology, the present invention introduces three control variables: the fatigue time threshold of the yaw system, the data acquisition interval, and the wind direction change threshold. By adaptively and dynamically regulating these three variables, the relationship between the total power generation of the whole field and the yaw fatigue of each turbine can be balanced as needed.

[0050] Optionally, the standard fatigue time threshold is the duration of the fatigue monitoring period multiplied by the threshold coefficient; the threshold coefficient is determined by the proportion of the maximum allowable working time of the turbine yaw system in its designed service life cycle. This proportion can be obtained through statistical analysis of historical operation data or through industry standards. Once this threshold coefficient is determined, it represents the maximum fatigue time quota allowed for the turbine to perform wake deflection control.

[0051] Optionally, the wind direction change threshold can trigger the start or shutdown of the wake deflection optimization solver. Once the wind direction change exceeds this threshold, the wake deflection optimization solver will start calculating the yaw angles of all the turbines in the new wind direction to cope with the change in the wind farm flow field. Conversely, the wake deflection optimization solver will be in a dormant or shutdown state. The wind direction change threshold in the present invention can follow the threshold setting in the single turbine control logic or adopt a common wind direction change threshold.

[0052] It should be noted that both the data acquisition interval and the full-field wind direction change threshold can be adjusted bidirectionally. Specifically, when the data acquisition interval is adjusted towards the lower value, that is, when the data acquisition frequency is increased, the number of starts of the wake deflection optimization solver will increase; conversely, when the set value is adjusted towards the higher value, the number of starts of the solver will decrease. Similarly, when the full-field wind direction change threshold is adjusted towards the lower value, the sensitivity of the wake deflection system to wind direction changes will increase, and the number of starts of the wake deflection optimization solver will increase. Conversely, when the set value is adjusted towards the higher value, the number of starts of the solver will decrease. It should be noted that the number of starts of the wake deflection optimization solver is positively correlated with the number of operations of the yaw system and is also positively correlated with the cumulative yaw operation time of the yaw system. The number of starts of the wake deflection optimization solver is restricted by the fatigue time threshold of the yaw system. Within the restricted range, it is expected that the number of starts of the wake deflection optimization solver is as large as possible, so that the increase in the total power generation of the whole field will be more.

[0053] As a specific embodiment, the data acquisition interval is greater than the yaw response execution time of all the turbines in the field and is also greater than the maximum wake conduction time in the determined wind direction in the flow field wind condition information; the yaw response execution time is calculated by dividing the maximum allowable single yaw range by the yaw speed; the maximum wake conduction time is calculated by dividing the maximum wake conduction distance in the determined wind direction by the average flow velocity of the flow field.

[0054] It should be noted that the data acquisition frequency cannot be too high. It is necessary to ensure that the data acquisition interval is greater than the yaw response execution time of all the turbines in the field and also consider that it needs to be greater than the maximum wake conduction time in the determined wind direction. Similarly, the acquisition frequency cannot be too low, otherwise the increase in power generation due to wake deflection will be limited, and even the power generation may decrease instead. Therefore, the design principle for the lower limit of the acquisition frequency is to be as close as possible to the upper limit.

[0055] Specifically, the yaw response execution time T yaw_max can be calculated by dividing the maximum allowable single yaw range by the yaw speed; for the maximum wake conduction time in the determined wind direction, it can be calculated by the following steps: establish a rotating wind direction coordinate system with the determined wind direction as the positive x-axis direction. The specific establishment method can refer to the description in the embodiment where the flow field wind condition information in the wind farm area is collected.

[0056] Select the upstream first-row turbines and the downstream last-row turbines that are perpendicular to the determined wind direction (there can be multiple turbines in the first row and multiple turbines in the last row); the determination method of the first-row turbines and the last-row turbines can refer to the determination method of the first-sequence turbines in the embodiment where the wind condition information is collected.

[0057] Calculate the x-axis coordinate vectors (the x-axis coordinate vectors composed of the x-axis coordinates of multiple turbines) X of the first-row turbines and the last-row turbines respectively in the rotating wind direction coordinate system row_1 , X row_N ; Select the turbine position with the smallest x coordinate in the first-row turbines Min(X row_1 ); Subtract the smallest x-axis coordinate value in the first-row turbines to obtain the wake conduction distance vector; respectively obtain the wind speed vectors of the first and last-row turbines and calculate the average to obtain the average flow velocity of the flow field; divide the wake conduction distance vector by the average flow velocity of the flow field to obtain the wake conduction time vector, and take the largest element in this vector as the maximum wake conduction time T flow_max ; Take the maximum value of the yaw response execution time T yaw_max and the maximum wake conduction time T flow_max as the minimum value of the data acquisition interval.

[0058] It should be noted that in the wind farm wake control technology system, the data acquisition link plays a decisive role in the precise operation of the system. The traditional data acquisition mode mostly adopts an equal-interval input mechanism to collect wind condition information within a fixed time period. However, this method has obvious limitations when facing the dynamic change of the wind direction, is difficult to flexibly adjust the data acquisition frequency, and cannot provide timely and accurate data support for the wake deflection system. The wake deflection system of the present invention shows a high degree of adaptability in setting the data acquisition interval. The system can adaptively adjust the data acquisition interval of the system according to the full-field wake propagation path and duration under different wind direction angles, ensuring that the operation of each turbine is more in line with the real-time changes of the entire wind farm flow field. This adaptive data acquisition interval setting mode greatly enhances the flexibility of the wind condition data acquisition input, provides a comprehensive, real-time and accurate data basis for the wake deflection system. Based on this high-quality data support, the system can make more timely and accurate decisions, optimize the control method, and then significantly improve the power generation efficiency of the wind farm, effectively reduce the wake influence, and provide a solid guarantee for the efficient and stable operation of the wind farm.

[0059] In addition to a method for hierarchical optimization control of wind farm wakes, the present invention also provides a hierarchical optimization control system for wind farm wakes, including a central controller for performing wake deflection control on the yaw systems of all turbines in the wind farm area; the central controller includes: A flow field wind condition recognition module for dynamically recognizing the free incoming flow direction of the wind farm; Data collection and processing module, collecting flow field and wind condition information in the field and performing data processing; The wake deflection control module controls the start or shut down of the wake deflection optimization solver, and issues the wake deflection command for all units in the field according to the yaw angle setting value of all units in the field; The wake deflection optimization solver optimizes the yaw angles of all turbines under different wind conditions.

[0060] It should be noted that the present invention mainly designs a central controller for the wake deflection of all turbines in the field, and introduces a top-level supervisory control architecture and mechanism to control the startup and shutdown of the wake deflection optimization solver under different working conditions and the input of external parameters. The main task of the wake deflection optimization solver is to complete the solution of the optimal yaw angle of all turbines in the field under different working conditions. The effect of the present invention can be achieved by using the existing wake deflection optimization solver, so it will not be described in detail.

[0061] When the functions in this application are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that can be executed by a processor. Based on this, the contribution made by the technical solution of this application compared to the prior art, or a specific part of the technical solution, can be presented in the form of a software product. The computer software product is stored in a storage medium and contains a series of instructions to drive a computer device (such as a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of this application. The above-mentioned storage medium includes but is not limited to: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store program codes.

[0062] The wake phenomenon of a wind farm specifically refers to an airflow area with reduced speed and greatly enhanced turbulence intensity formed after the airflow passes through the wind turbine blades when the wind turbine is capturing wind energy. Due to the existence of the wake, the quality of wind energy obtained by the wind turbine in the downstream position is significantly reduced, and the power generation capacity is greatly reduced. According to a large amount of authoritative research data, the wake effect can usually cause the overall power generation loss of a wind farm to reach 10%-20%. This loss not only directly weakens the power generation efficiency of the wind farm, but also indirectly raises the cost of power generation, which has an extremely adverse impact on the economic benefits and market competitiveness of the wind farm. What is more serious is that the sharp increase in turbulence intensity in the wake area will impose additional fatigue loads on the structural components of the downstream wind turbines, accelerate the wear process of the equipment, shorten the normal service life of the equipment, and significantly increase the operation and maintenance costs and safety risks of the wind farm.

[0063] In the development and construction of offshore wind farms, this problem is particularly prominent. Usually, in the early stage of offshore wind farm development, micro-siting of wind turbines is carried out to optimize the overall layout of the whole field, with the aim of reducing the wake influence between turbines. However, most offshore wind farms are located near important coastal waterways and ports. The sea area used in these areas is strictly restricted, which makes it difficult to optimize the layout of turbines within the wind farm, resulting in generally large wakes within the wind farm and a situation of "inherent deficiencies". Under such objective conditions, relying solely on the layout optimization in the early stage can no longer effectively solve the wake problem. Only through "postnatal" wake collaborative control technology can the wake influence between turbines be alleviated. However, the existing steady-state wake deflection technology has obvious limitations. It only focuses on the single index of the total power generation of the whole field, and completely ignores the working characteristics, fatigue level of the turbine yaw system and the changes in external working conditions.

[0064] The present invention designs a hierarchical optimization control method for wind farm wakes and solidifies it into a set of wind farm-level control systems, aiming to make up for the deficiencies of the steady-state wake deflection technology; it not only considers the optimization of the total power generation of the whole field, but also takes into account the working characteristics and fatigue level of the turbine yaw system. By real-time monitoring the operating state of wind turbines, the working parameters of the yaw system and the environmental changes of the wind farm, it can dynamically adjust the yaw angle of wind turbines according to the actual situation, effectively reducing the working load of the yaw system and extending the service life of the yaw system while ensuring the stable increase of the total power generation of the whole field, providing a more reliable guarantee for the efficient and stable operation of the wind farm.

[0065] In the embodiment of the present invention, taking an offshore wind farm with a total installed capacity of 300 MW as an example to illustrate the situation after adopting the wake deflection control method of the present invention, the fatigue monitoring period of the yaw system in this embodiment is preset to 1 hour. As Figure 2 Shown is a comparison chart of the total yaw working time and the number of yaw actions within a total of twenty-four fatigue monitoring periods in a day. Compared with the traditional steady-state wake deflection technology mentioned in the background art, the control method of the present invention has achieved a significant decrease in the yaw working time and the number of yaw operations within the fatigue monitoring period of the yaw system, with the decrease amplitudes reaching 48.5% and 74.6% respectively. As Figure 3As shown, compared with the case where wake deflection optimization control is not implemented (the original reference value), both the method of the present invention and the traditional steady-state wake deflection technology have achieved an increase in the total power generation of the whole field relative to the original reference value. Among them, the power generation increase ratio of the method of the present invention is 3.94%, and the power generation increase ratio of the traditional steady-state wake deflection technology is 4.23%, with a difference of only 0.29% between the two. The original reference value refers to ignoring the wake influence of the upstream unit on the downstream, not aiming for the optimal total power generation of the whole field, but aiming for the optimal control of the power generation of a single unit, and summing up the power generation of all wind turbines to obtain the total power generation of the whole field. Generally speaking, the wake deflection control method of the present invention is very close to the traditional steady-state wake deflection technology in terms of the effectiveness of increasing the total power generation of the whole field, but it has more significant advantages in reducing the fatigue degree of the yaw system of the unit.

[0066] The above embodiments are further elaborations and explanations of the present invention for the convenience of understanding, and are not any limitations to the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A stepwise optimization control method for the wake of a wind farm, characterized in that: include: Collect the flow field and wind condition information within the wind farm area, and determine the data collection interval based on the flow field and wind condition information; In each preset fatigue monitoring cycle, the yaw working time of each unit is collected regularly to monitor the fatigue status: the maximum value of the accumulated yaw working time of each unit is obtained to determine whether it exceeds the fatigue time threshold of the fatigue monitoring cycle; If yes, the yaw angles of all units are reset to zero; If not, it is determined whether the flow field wind condition information meets the start-up conditions of the wake deflection optimization solver; if the start-up conditions are met, a new yaw angle of the entire unit is calculated and output, otherwise the yaw angle of the entire unit is maintained unchanged.

2. A stepwise optimization control method for a wind farm wake according to claim 1, characterized in that: After each fatigue monitoring cycle is completed, the accumulated yaw working time of all units is reset to zero; Calculate the difference between the maximum accumulated yaw working time and the fatigue time threshold value during the fatigue monitoring period; The difference is subtracted from the standard fatigue time threshold value as the fatigue time threshold value for the next fatigue monitoring cycle.

3. A stepwise optimization control method for a wind farm wake according to claim 1 or 2, characterized in that: The determination of whether the flow field wind condition information meets the start-up conditions of the wake deflection optimization solver includes: Determine whether the wind speed in the field is greater than the cut-in wind speed of all units. If not, maintain the yaw angle of all units in the field unchanged and close the wake deflection optimization solver; If so, determine whether the wind direction change in the field is greater than the wind direction change threshold; if so, start the wake deflection optimization solver and output a new yaw angle for all units in the field; otherwise, maintain the yaw angle of all units in the field unchanged.

4. A stepwise optimization control method for a wind farm wake according to claim 3, characterized in that: When the wind direction change in the field is less than or equal to the wind direction change threshold, the yaw angle of all units in the field is maintained unchanged, and the wake deflection optimization solver is closed; Determine whether the total field power generation is greater than the original reference value without wake deflection control; If it is greater than, the yaw angle of the entire unit is maintained unchanged; otherwise, the wake deflection optimization solver is started and a new yaw angle of the entire unit is output.

5. The stepwise optimization control method for the wake of a wind farm according to claim 3, characterized in that: The number of startup times of the wake deflection optimization solver is negatively correlated with the data collection interval, negatively correlated with the wind direction change threshold, and positively correlated with the accumulated yaw working time.

6. A stepwise optimization control method for a wind farm wake according to claim 1, 2, 4 or 5, characterized in that: The collecting of wind condition information of the wind farm area includes: determining the upstream first-order unit in each wind direction with the minimum wind direction change interval; determining the wind direction set corresponding to each first-order unit; Obtain the wind direction measured by all first-order units at regular intervals and calculate the average wind direction; The wind direction set into which the average wind direction falls is determined, and the wind measurement information of the first-order unit corresponding to the wind direction set is used as the flow field wind condition information.

7. A stepwise optimization control method for a wind farm wake according to claim 1, 2, 4 or 5, characterized in that: The data collection interval is greater than the yaw response execution time of all units in the field, and greater than the maximum conduction time of the wake in the wind direction determined in the flow field wind condition information; The yaw response execution time is the single maximum allowable yaw range divided by the yaw speed; The maximum wake conduction time is determined by dividing the maximum wake conduction distance in the wind direction by the average flow velocity of the flow field.

8. A stepwise optimization control method for a wind farm wake according to claim 7, characterized in that: The maximum wake conduction time is: establish a rotating wind direction coordinate system with the wind direction determined as the positive direction of the x-axis; Select the first row of upstream units and the last row of downstream units that are perpendicular to the determined wind direction; Calculate the x-axis coordinate vector of the last row of units, and subtract the smallest x-axis coordinate value of the first row of units to obtain the wake conduction distance vector; The wake conduction time vector is obtained by dividing the wake conduction distance vector by the average flow velocity of the flow field, and the largest element in the vector is taken as the maximum wake conduction time.

9. The stepwise optimization control method for the wake of a wind farm according to claim 2, characterized in that: The standard fatigue time threshold is the duration of the fatigue monitoring cycle multiplied by the threshold coefficient; The threshold coefficient is determined by the ratio of the maximum allowable working time within the design service period of the unit yaw system to its design service period.

10. A stepwise optimization control system for a wind farm wake, applicable to the stepwise optimization control method according to any one of claims 1 to 9, characterized in that: A central controller for performing wake deflection control on the yaw systems of all units in a wind farm area; the central controller includes: The flow field wind condition identification module dynamically identifies the free flow direction of the wind farm; Data collection and processing module, collecting flow field and wind condition information in the field and performing data processing; The wake deflection control module controls the start or shut down of the wake deflection optimization solver, and issues the wake deflection command for all units in the field according to the yaw angle setting value of all units in the field; The wake deflection optimization solver optimizes the yaw angles of all turbines under different wind conditions.

Images