Wind power plant generating capacity improving method based on active wake flow yaw control optimization

Abstract

The invention discloses a wind power plant generating capacity improving method based on active wake flow yaw control optimization. The method comprises the following steps: 1) acquiring basic information data of a wind power plant; 2) establishing a CFD simulation model of the wind power plant according to the basic information data; 3) optimizing the arrangement scheme of the laser wind finding radar, and obtaining the operation data of the wind turbine generator in real time; 4) constructing a single-machine analysis wake flow model and a multi-machine wake flow superposition model, respectively comparing with the flow field simulation result data and the wind measurement result of the laser wind measurement radar, and correcting a wake flow speed distribution expression to obtain wind speed and wind direction prediction data at the hub center height of each machine site in the wind power plant; and 5) constructing a whole field optimization control model, taking maximum output as an optimization target to obtain an optimal yaw control strategy set of each wind turbine generator, and enabling the related wind turbine generators to execute yaw actions. The invention belongs to the technical field of wind power generation control, and solves the problem that the wake flow loss among wind power units in a wind power plant is difficult to reduce by the existing control strategy.

Description

Technical Field

[0001] The invention belongs to the technical field of wind power generation control, and relates to a method for improving power generation of a wind farm based on active wake yaw control optimization. Background Art

[0002] After the wind turbine captures energy from the incoming flow, a wake zone with reduced wind speed and increased turbulence will be formed downstream, resulting in uneven wind speed distribution in the downstream area, which directly affects the power generation performance of the downstream wind turbines and increases the load they bear, resulting in a reduction in the power generation efficiency and service life of the downstream wind turbines. Although the existing wind farms can achieve the best single-machine power generation performance by optimizing the control strategy of the units, the influence of the wake effect is becoming increasingly significant as the capacity of the single unit and the installed scale of the wind farm increase. In addition, the losses caused by the wake effect are further increased when the upstream wind farm is completed and put into operation.

[0003] Although the influence of the wake effect of some wind turbines has been avoided through micro-site selection in the early stage of wind farm construction, there is still room for further optimization. Therefore, it is urgent to develop a new method to maximize the use of wind energy, further improve power generation capacity, and achieve the best field-level operating benefits through fine control of the wake using active wake yaw. Summary of the invention

[0004] The purpose of the present invention is to provide a method for improving wind farm power generation based on active wake yaw control optimization, which solves the problem that high-precision wake calculation efficiency is low and the existing single-machine optimal control strategy is difficult to reduce the wake loss between each wind turbine in the wind farm.

[0005] The technical solution adopted by the present invention is a method for increasing wind farm power generation based on active wake yaw control optimization, which is implemented according to the following steps:

[0006] Step 1: Obtain basic information data of the wind farm;

[0007] Step 2: Establish a wind farm CFD simulation model based on basic information data;

[0008] Step 3: Optimize the layout of the laser wind radar to obtain wind turbine operation data in real time;

[0009] Step 4: Construct a single-machine analytical wake model and a multi-machine wake superposition model, compare them with the flow field simulation result data and the wind measurement results of the laser wind radar, correct the wake velocity distribution expression, and obtain the wind speed and wind direction prediction data at the hub center height of each machine site in the wind farm;

[0010] Step 5: Build an optimization control model for the entire field, take maximum output as the optimization goal, obtain the optimal yaw control strategy set for each wind turbine, and the relevant wind turbines perform yaw actions.

[0011] The beneficial effect of the present invention is that, through the laser wind measuring radar, combined with the constructed analytical model, the wind speed and wind direction information of each wind turbine station is predicted, and the maximum power generation of the whole field is taken as the optimization goal. The particle swarm algorithm and the high-performance solver are used to obtain the wake yaw control strategy, and the dynamic adjustment of the yaw angle of the wind turbine is realized to reduce the wake influence between the wind turbines and achieve the optimization of the whole field. The problem that the high-precision wake calculation efficiency in the prior art is low and the control strategy with the best single-machine power generation performance is difficult to reduce the wake loss between the wind turbines in the wind farm is solved, so as to reduce the influence of the wake effect between the units and significantly improve the operation efficiency of the wind farm. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a schematic diagram of the overall process of the method of the present invention;

[0013] Figure 2 It is a schematic diagram of the overall wind farm power optimization control process in the method of the present invention;

[0014] Figure 3 is a schematic diagram of the power load optimization control strategy in the method of the present invention;

[0015] Figure 4 This is the result of single-machine optimal control under 0° wind direction in Example 1 of the present invention;

[0016] Figure 5 This is the active wake yaw optimization control result of the entire field under 0° wind direction in Example 1 of the present invention;

[0017] Figure 6 This is the result of single-machine optimal control under 245° wind direction in Example 1 of the present invention;

[0018] Figure 7 This is the result of the active wake yaw optimization control for the entire field under a wind direction of 245° in Example 1 of the present invention. DETAILED DESCRIPTION

[0019] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] The method of the present invention is based on active wake yaw control, takes the maximum power output value of the entire field as the optimization target, obtains the optimal yaw control strategy set for the wind turbines in the entire field, and controls the relevant wind turbines in a targeted manner, thereby reducing the impact of the wake effect on the downstream wind turbines and improving the power generation efficiency of the entire wind farm.

[0021] Reference Figure 1 The method of the present invention is implemented according to the following steps:

[0022] Step 1: Obtain basic information data of the wind farm.

[0023] The basic information data mainly includes: terrain data, wind turbine coordinates, environmental data and wind turbine information; it can also include the geographical coordinates of the wind farm area, meteorological data, hub height, rotor diameter, power curve, wind energy utilization coefficient C of each wind turbine P , thrust coefficient C T , historical operation SCADA data of each wind turbine unit, etc.

[0024] Step 2: Based on the aforementioned basic information data, taking into account the terrain and surrounding mountain shielding factors, a high-precision wind farm CFD simulation model is established by combining fluid mechanics with a rotating actuator disk model. The specific process is:

[0025] 2.1) Use the existing topographic map data with a scale of 1:2000 to obtain the digital elevation model (DEM); use the GLC30 geomorphic data (existing literature) to obtain the roughness distribution data of the project wind farm and the surrounding surface;

[0026] 2.2) Modeling the wind turbine generator set using the actuator disk model to obtain a wind turbine generator set model;

[0027] 2.3) According to the topographic data of the wind farm (i.e. the two types of data obtained in step 2.1), three-dimensional modeling and mesh processing are performed at a distance of not less than 1.5 times in the direction of flat land and not less than 2.5 times in the direction of mountains, and the mesh of the wind farm itself and the surrounding mountain areas are encrypted to obtain a more refined CFD simulation model of the wind farm.

[0028] Step 3: Optimize the layout of the laser wind radar, install the laser wind radar in place, connect the meteorological data to the wind turbine main control system, and obtain the wind turbine operation data (such as wind speed, wind direction, etc.) in real time. The specific process is:

[0029] 3.1) Based on the analysis results of the wind farm CFD simulation model and the wind speed correlation analysis of each wind turbine, the wind turbines that are correlated with each wind turbine are screened according to the correlation threshold, and all wind turbines are grouped. Each group selects a wind turbine that is also correlated with other groups as the laser wind radar station site, and the layout plan of the laser wind radar is determined;

[0030] 3.2) According to the layout plan, deploy laser wind measuring radars at representative positions, install and wire the laser wind measuring radars according to the shape of the nacelle cover and the center line of the wind turbine shaft, and complete the installation and commissioning of the laser wind measuring radars;

[0031] 3.3) The information collected by the laser wind radar is input into the main control system of the wind turbine, and the measurement data of the laser wind radar is obtained and stored, providing meteorological data support for improving the prediction accuracy in subsequent steps.

[0032] Step 4: Construct a single-machine analytical wake model and a multi-machine wake superposition model, compare them with the flow field simulation result data and the wind measurement results of the laser wind radar, and correct the wake velocity distribution expression;

[0033] Combined with the characteristics of wind turbine arrangement, wind turbines in surrounding wind farms, terrain and other factors, and taking into full account the secondary deflection effect of wake and the accumulation of long-period prediction errors, the wind speed and wind direction prediction data at the hub center height of each machine site in the wind farm are obtained.

[0034] The specific process is:

[0035] 4.1) Single-machine analytical wake model The existing Gaussian wake model is used. According to the conservation of momentum in the wake flow, when the wake pressure returns to atmospheric pressure, the thrust expression of a single wind turbine is:

[0036]

[0037] Where T is the thrust of the wind turbine, ρ is the air density, is the average velocity on the central axis of the induced area, U ∞ is the free stream wind speed at the hub center height, Δu is the wake loss velocity profile, and A is the unit wind rotor swept area;

[0038] Assuming that the velocity loss distribution in the wake is Gaussian, the velocity loss expression in the wake is:

[0039]

[0040] Among them, C T is the thrust coefficient of the wind turbine, σ is the standard deviation of the Gaussian velocity profile, d0 is the rotor diameter, and r is the radial distance from the wake center, then:

[0041]

[0042] Among them, k W is the wake expansion rate, x is the longitudinal distance from the center of the wind rotor to the center of the downstream wake, and β is the ratio of the wake diameter to the wind rotor diameter when the wake pressure downstream of the wind rotor returns to atmospheric pressure;

[0043] 4.2) Compare the flow field simulation result data obtained by simulating the wind farm CFD simulation model in step 2.3) with the wind measurement data obtained by the laser wind radar, correct and obtain the optimal wake velocity distribution expression, comprehensively consider the incoming flow turbulence and far-field wake deflection characteristics, calculate the wake center offset under different yaw states, and obtain the optimal expression that adapts to the wind farm model and environment. The corrected single-machine wake velocity distribution expression is:

[0044] US =a S (U ∞ -Δu)+b S #

[0045] Among them, a S is the single-machine wake velocity correction coefficient, b S is the single-machine wake velocity correction;

[0046] 4.3) The Katic multi-machine wake superposition model between multiple wind turbines is used. Based on the law of conservation of kinetic energy, the average wind speed of downstream unit j under the influence of the wake of n upstream units is obtained, which is expressed as:

[0047]

[0048] in, is the inflow wind speed of downstream wind turbine j, U ∞ is the free stream wind speed at the hub center height, n is the total number of downstream units j affected by upstream units, U ij represents the equivalent wind speed of the upstream unit i at the downstream unit j. According to the law of conservation of momentum, U ij The expression is:

[0049]

[0050] Where A is the swept area of ​​the wind wheel, A=πD 2 / 4;

[0051] 4.4) Considering the characteristics of the unit arrangement, the surrounding wind farm units and terrain, and combining the high-precision flow field numerical simulation results with the laser wind radar wind measurement data, the optimal wake stacking velocity distribution is corrected and obtained. The corrected multi-machine wake velocity distribution expression is:

[0052]

[0053] Among them, a M is the multi-machine wake velocity correction factor, b M is the multi-machine wake velocity correction;

[0054] 4.5) Considering the influence of the secondary deflection effect of the wake on the wake of the downstream wind turbines and the accumulation of long-period prediction errors, the total velocity loss in the wake superposition area is accurately estimated under the given yaw sequence and initial point conditions, and the wind speed and wind direction prediction data at the hub center height of each wind turbine in the wind farm are obtained.

[0055] Step 5: Build an optimization control model based on the maximum output of the whole field, take the maximum output as the optimization goal, input the meteorological forecast data at the station, and use the particle swarm algorithm and solver (the high-performance configuration of the solver is preferably: CPU: 2* Gold 6338 processor, 2GHz, 64 cores and 128 threads; memory: 128G), quickly calculate the output of the entire wind farm and the wake distribution under different yaw angles, and obtain the optimal yaw control strategy set for each wind turbine; the entire field optimization control model sends the yaw control strategy to the wind turbine main control system, and the relevant wind turbines perform yaw actions and execute Figure 3 The power load optimization control strategy is operated to achieve the purpose of improving the overall power generation revenue and reducing the load of wind turbines. Figure 2 , implement the power optimization control of the whole wind farm,

[0056] The specific process is:

[0057] 5.1) According to the wind farm CFD simulation model obtained in step 2, the wind speed and wind direction forecast data of each wind turbine are obtained, and the whole field optimization control model with maximum output as the optimization target is constructed, as shown in formula (8):

[0058]

[0059] Among them, P i is the power of the i-th wind turbine, ρ is the air density, U p is the wind energy utilization coefficient, A is the swept area of ​​the wind rotor, u i is the wind speed at the hub center height of the i-th wind turbine, γ i is the yaw angle of the i-th wind turbine;

[0060] 5.2) Through the particle swarm algorithm, the maximum power output and yaw value of each wind turbine corresponding to the wind speed and wind direction are quickly traversed to obtain the yaw control strategy set under the maximum power output value of the whole field, which is expressed as:

[0061] γ opt =(γ1,γ2,…γ i ,…γ n )#

[0062] Among them, γ opt Achieve the optimal yaw control strategy set for n wind turbines to achieve the entire field power generation;

[0063] 5.3) The whole-field optimization control model sends the yaw control strategy to the wind turbine main control system, and the relevant wind turbines perform yaw actions and adjust the operating yaw angles of the relevant wind turbines to achieve optimal field-level power generation.

[0064] Reference Figure 3, the relevant wind turbines implement a yaw control strategy according to the following operating conditions:

[0065] When v > v cut-in , all wind turbines start to operate, where v is the wind speed at the hub center height of the wind turbine, and v cut-in is the cut-in wind speed of the wind turbine;

[0066] When v < v rate , all wind turbines operate with power load optimization control, and v rate is the rated wind speed of the wind turbine;

[0067] When v rate < v < v1, the under-loaded wind turbines operate with power load optimization control, and some fully-loaded wind turbines operate with the control method that minimizes the load at the full-load power, where v1 is the wind speed when each wind turbine is fully loaded;

[0068] When v1 < v < v cut-out , all wind turbines operate with the control method that minimizes the full-load power load, and v cut-out is the cut-out wind speed of the wind turbine;

[0069] When v cut-out < v, the loads of all wind turbines are too high, and all of them perform protective shutdown.

[0070] Example 1

[0071] The active wake yaw control is implemented on the wind turbines of a wind farm in Qinghai using the method of the present invention.

[0072] This wind farm consists of 20 wind turbines of a certain type with a capacity of 2.5 MW. The wind turbine rotor diameter is 131 m, and the hub height is 90 m. The annual average wind speed at the hub center height of this wind farm is 6.58 m / s, and the turbulence intensity is 0.06. The initial active yaw angles of the wind turbines are all 0, and the wind wheels are perpendicular to the incoming wind. After the active wake yaw control, the power generation of the whole field increases by 1.94% at this wind speed. Among them, the power generation at the 0° wind direction (north wind) with a large wake influence increases by 34.53%, and the power generation at the 180° (south wind) with a large wake influence increases by 34.59%; the wind energy of this wind farm is concentrated between the west direction and the southwest direction. Taking the 245° wind direction (southwest by west) as an example, the power generation increases by 2.39%.

[0073] Table 1 and Table 2 respectively give the yaw angles γ of each unit under the active wake yaw control at the 0° and 245° wind directions, Figure 4 , Figure 5 , Figure 6 , Figure 7 are the results of single-unit optimal control and whole-field active wake yaw optimization control at the 0° and 245° wind directions, respectively.

[0074] Table 1. Yaw angle γ of each unit under active wake yaw control at 0° wind direction

[0075]

[0076] Table 2. Yaw angle γ of each unit under active wake yaw control at 245° wind direction

[0077]

[0078] The results of single-machine optimal control and whole-field active wake yaw optimization control under different wind directions are compared as follows: Figure 4 It is the result of single machine optimal control under 0° wind direction; Figure 5 It is the result of the whole-field active wake yaw optimization control under 0° wind direction; Figure 6 It is the result of single machine optimal control under 245° wind direction; Figure 7 This is the result of the whole-field active wake yaw optimization control under a wind direction of 245°. It can be seen that the whole-field active wake yaw optimization control method of the present invention reduces the influence of the wake effect between wind farm units and significantly improves the operating efficiency of the wind farm.

[0079] Example 2

[0080] Also based on a wind farm in Qinghai in Example 1, Example 2 adopts the method of the present invention and takes the rated wind speed of the wind turbines in the wind farm as an example. When the wind speed at the center height of the hub is 9.5m / s and the turbulence intensity is 0.06, after active wake yaw control, the power generation of the entire field is increased by 2.14% at this wind speed.

[0081] Example 3

[0082] Also based on a wind farm in Qinghai in Example 1, Example 3 adopts the method of the present invention to take the working condition greater than the rated wind speed of the wind turbine set in the wind farm as an example. When the wind speed at the center height of the hub is 11m / s and the turbulence intensity is 0.06, after active wake yaw control, the power generation of the entire field is increased by 1.20% at this wind speed.

Claims

1. A method for increasing wind farm power generation based on active wake yaw control optimization, characterized in that: Follow these steps to implement: Step 1: Obtain basic information data of the wind farm; Step 2: Establish a wind farm CFD simulation model based on basic information data; Step 3: Optimize the layout of the laser wind radar to obtain wind turbine operation data in real time; Step 4: Construct a single-machine analytical wake model and a multi-machine wake superposition model, compare them with the flow field simulation result data and the wind measurement results of the laser wind radar, correct the wake velocity distribution expression, and obtain the wind speed and wind direction prediction data at the hub center height of each machine site in the wind farm; Step 5: Build an optimization control model for the entire field, take maximum output as the optimization goal, obtain the optimal yaw control strategy set for each wind turbine, and the relevant wind turbines perform yaw actions.

2. The method for increasing wind farm power generation based on active wake yaw control optimization according to claim 1, characterized in that: In step 1, the basic information data includes: terrain data, wind turbine coordinates, environmental data and wind turbine information.

3. The method for increasing wind farm power generation based on active wake yaw control optimization according to claim 1, characterized in that: In step 2, the specific process is: 2.1) Use the existing topographic map data with a scale of 1:2000 to obtain the digital elevation model DEM; use the GLC30 landform data to obtain the roughness distribution data of the project wind farm and the surrounding surface; 2.2) Modeling the wind turbine generator set using the actuator disk model to obtain a wind turbine generator set model; 2.3) According to the topographic data of the wind farm, three-dimensional modeling and grid processing are carried out at a distance of not less than 1.5 times in the flat direction and not less than 2.5 times in the mountainous direction. The grid of the wind farm itself and the surrounding mountain areas is encrypted to obtain the CFD simulation model of the wind farm.

4. The method for increasing wind farm power generation based on active wake yaw control optimization according to claim 1, characterized in that: In step 3, the specific process is: 3.1) Based on the analysis results of the wind farm CFD simulation model and the wind speed correlation analysis of each wind turbine, the wind turbines that are correlated with each wind turbine are screened according to the correlation threshold, and all wind turbines are grouped. Each group selects a wind turbine that is also correlated with other groups as the laser wind radar station site, and the layout plan of the laser wind radar is determined; 3.2) According to the layout plan, deploy laser wind measuring radars at representative positions, install and wire the laser wind measuring radars according to the shape of the nacelle cover and the center line of the wind turbine shaft, and complete the installation and commissioning of the laser wind measuring radars; 3.3) The information collected by the laser wind radar is input into the main control system of the wind farm to obtain and store the measurement data of the laser wind radar to provide meteorological data support.

5. The method for increasing wind farm power generation based on active wake yaw control optimization according to claim 1, characterized in that: In step 4, the specific process is: 4.1) Single-machine analytical wake model The existing Gaussian wake model is used. According to the conservation of momentum in the wake flow, when the wake pressure returns to atmospheric pressure, the thrust expression of a single wind turbine is: Where T is the thrust of the wind turbine, ρ is the air density, is the average velocity on the central axis of the induced area, U ∞ is the free stream wind speed at the hub center height, Δu is the wake loss velocity profile, dA is the unit wind rotor swept area, Assuming that the velocity loss distribution in the wake is Gaussian, the velocity loss expression in the wake is: Among them, C T is the thrust coefficient of the wind turbine, σ is the standard deviation of the Gaussian velocity profile, d0 is the rotor diameter, and r is the radial distance from the wake center, then: Among them, k W is the wake expansion rate, x is the longitudinal distance from the center of the wind rotor to the center of the downstream wake, and β is the ratio of the wake diameter to the wind rotor diameter when the wake pressure downstream of the wind rotor returns to atmospheric pressure; The free stream wind speed at the hub center height minus the speed loss in the wake, that is, U ∞ -Δu; 4.2) Compare the flow field simulation result data obtained by the wind farm CFD simulation model with the wind measurement data obtained by the laser wind radar, correct and obtain the optimal wake velocity distribution expression, comprehensively consider the incoming flow turbulence and far-field wake deflection characteristics, calculate the wake center offset under different yaw states, and obtain the optimal expression that adapts to the wind farm model and environment. The corrected single-machine wake velocity distribution expression is: U S =a S (U ∞ -Δu)+b S # Among them, a S is the single-machine wake velocity correction coefficient, b S is the single-machine wake velocity correction; 4.3) The Katic multi-machine wake superposition model between multiple wind turbines is used. Based on the law of conservation of kinetic energy, the average wind speed of downstream unit j under the influence of the wake of n upstream units is obtained, which is expressed as: in, is the inflow wind speed of downstream wind turbine j, U ∞ is the free stream wind speed at the hub center height, n is the total number of downstream units j affected by upstream units, U ij represents the equivalent wind speed of the upstream unit i at the downstream unit j. According to the law of conservation of momentum, U ij The expression is: Where A is the swept area of ​​the wind wheel, A=πD 2 / 4; 4.4) Considering the characteristics of the unit arrangement, the surrounding wind farm units and terrain, and combining the high-precision flow field numerical simulation results with the laser wind radar wind measurement data, the optimal wake stacking velocity distribution is corrected and obtained. The corrected multi-machine wake velocity distribution expression is: Among them, a M is the multi-machine wake velocity correction factor, b M is the multi-machine wake velocity correction; 4.5) Considering the influence of the secondary deflection effect of the wake on the wake of the downstream wind turbines and the accumulation of long-period prediction errors, the total velocity loss in the wake superposition area is estimated under the given yaw sequence and initial point conditions, and the wind speed and wind direction prediction data at the hub center height of each wind turbine in the wind farm are obtained.

6. The method for increasing wind farm power generation based on active wake yaw control optimization according to claim 1, characterized in that: In step 5, the specific process is: 5.1) According to the wind farm CFD simulation model obtained in step 2, the wind speed and wind direction forecast data of each wind turbine are obtained, and the optimization control model of the whole field is constructed, as shown in formula (8): Among them, P i is the power of the i-th wind turbine, ρ is the air density, C p is the wind energy utilization coefficient, A is the swept area of ​​the wind rotor, u i is the wind speed at the hub center height of the i-th wind turbine, γ i is the yaw angle of the i-th wind turbine; 5.2) Through the particle swarm algorithm, the maximum power output and yaw value of each wind turbine corresponding to the wind speed and wind direction are quickly traversed to obtain the yaw control strategy set under the maximum power output value of the whole field, which is expressed as: c opt =(γ1,γ2,…γ i ,…c n )# Among them, γ opt Achieve the optimal yaw control strategy set for n wind turbines to achieve the entire field power generation; 5.3) The whole-field optimization control model sends the yaw control strategy to the main control system of the wind turbine. The relevant wind turbines perform yaw actions and adjust the operating yaw angles of the relevant wind turbines. The relevant wind turbines implement yaw control strategies according to different working conditions, thereby achieving optimal field-level power generation.

7. The method for increasing wind farm power generation based on active wake yaw control optimization according to claim 1, characterized in that: The yaw control strategy is implemented by the relevant wind turbines according to the following working conditions: When v>v cut-in When all wind turbines start to run, v is the wind speed at the center height of the wind turbine hub, v cut-in is the cut-in wind speed of the wind turbine; When v <v rate When all wind turbines are operated with power load optimization control, v rate is the rated wind speed of the wind turbine; When v rate When v < v1, the wind turbines that are not operating at full capacity adopt power load optimization control operation, and some of the wind turbines operating at full capacity operate in a control mode that minimizes the load while maintaining the full load power. v1 is the wind speed at which each wind turbine operates at full capacity. When v1 <v<v cut-out When all wind turbines are operated in full power load minimum control mode, v cut-out is the cut-out wind speed of the wind turbine; When v cut-out <When v is less than cut-out , the loads of all wind turbines are too high, and all of them perform protective shutdowns.

Images