CN111425347B - Wind turbine maximum power point tracking control method based on torque gain coefficient optimization - Google Patents

Abstract

The invention discloses a maximum power point tracking control method for wind turbines based on torque gain coefficient optimization. Based on the MPPT control method for reducing torque gain, the method takes the wind energy capture efficiency of the wind turbines running under the PSF method as the The comprehensive measurement index of the influence of turbulent wind speed on MPPT, the function relationship between the optimal torque gain coefficient and the index is traversed offline; when running online, the comprehensive measurement index is periodically obtained, and the optimal setting of the torque gain coefficient is determined according to the function. The fixed value is estimated and updated; the wind energy capture efficiency corresponding to the PSF method can be obtained by constructing a virtual wind turbine running the PSF method in the main control PLC of the unit and the actual unit running synchronously. The invention can realize a single index characterization of the comprehensive influence of multiple indexes on the MPPT, simplify the construction complexity of the direct quantitative relationship; while ensuring the wind energy capture efficiency, the computing power resource requirements are greatly reduced.

Description

Maximum power point tracking control method of wind turbine based on torque gain coefficient optimization

technical field

The invention belongs to the field of wind turbine control, in particular to a maximum power point tracking control method of a wind turbine based on torque gain coefficient optimization.

Background technique

In order to improve the maximum power point tracking (MPPT) performance of wind turbines in the face of turbulent wind speed, a new method based on torque gain and tracking interval was developed based on the most widely used power signal feedback (PSF) method. Adjust the improved PSF method of the two ideas. Both of these two ideas are to sacrifice the wind energy capture efficiency in the low-energy wind speed region in exchange for the improvement of the tracking performance in the high-energy wind speed region. It is necessary to reasonably set the key adjustment parameters to balance the loss and the increase to maximize the overall efficiency. The research shows that the optimal value of the key adjustment parameters is affected by the characteristics of wind conditions (average wind speed, turbulence intensity, turbulence frequency), as well as the aerodynamic and structural parameters of the unit. How to periodically estimate and update the optimization settings of key parameters according to the changes of the above-mentioned influencing factors during the operation process has become a focus issue.

At present, there are two types of solutions for this problem: adaptive torque control (ATC) and constructing the quantitative relationship between the optimal value of key parameters and influencing factors and guiding online operation. Among them, the adaptive torque control is based on the method of reducing the torque gain, and determines the direction and magnitude of the next cycle disturbance according to the change of the wind energy capture efficiency after the disturbance of the key parameters. The latter, for specific units, directly constructs a clear nonlinear functional relationship between the optimal torque curve adjustment amount and the three wind conditions and unit parameters through offline traversal. During online operation, the optimal set value of key parameters can be estimated according to the wind condition information and functional relationship.

The adaptive algorithm does not need to know the parameters of wind turbines in advance, and has the advantages of strong versatility and rapid implementation in batches. However, in some scenarios with changing wind conditions, the search does not converge or the search direction is wrong. Performance improvements are limited. The method of directly constructing the functional relationship between the optimal torque curve adjustment and the three wind condition characteristics to guide the online optimization of parameters avoids the iterative search process, and can obtain higher wind energy capture efficiency and good wind condition adaptability. However, such methods require a lot of time and computing power for offline traversal work, and are not easy to implement quickly in batches, which limits their engineering practicability. Therefore, how to balance high wind energy capture efficiency and fast implementation is a problem that still needs to be further solved in current MPPT control methods.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a maximum power point tracking control method for wind turbines based on torque gain coefficient optimization, which can obtain higher wind energy capture efficiency under the premise of paying less computing power and time cost.

The technical solution for realizing the purpose of the present invention is: a maximum power point tracking control method for wind turbines based on torque gain coefficient optimization, comprising the following steps:

(1) Offline construction of functional relationships

Step 1-1: For the wind turbine to which the present invention is to be applied, obtain the aerodynamic and structural parameters required for building its FAST model;

Step 1-2: In the FAST software, complete the construction of the wind turbine simulation model according to the parameters;

湍流强度TI和湍流积分尺度L，生成对应于不同特征指标组合的湍流风速序列；Step 1-3: Use the turbulent wind speed simulation method to sequentially change the three characteristic indicators that characterize the turbulent wind condition: average wind speed

Turbulence intensity TI and turbulence integral scale L are used to generate turbulent wind speed series corresponding to different characteristic index combinations;

以及应用PSF法时的风能捕获效率

Step 1-4: Traverse the wind turbine optimal torque gain coefficient corresponding to each wind speed sequence based on FAST software

and wind energy capture efficiency when applying the PSF method

和PSF法下平均风能捕获效率

的函数关系

Step 1-5: Fit the optimal torque gain coefficient with the traversal results as sample data

and the average wind energy capture efficiency under the PSF method

functional relationship

(2) Build a virtual wind turbine

Step 2-1: Embed the FAST model of the wind turbine to which the present invention is to be applied into its actual main control PLC to construct a virtual unit that can run synchronously with the actual unit;

Step 2-2: Embed the PSF method code for the MPPT control of the virtual wind turbine in the main control PLC;

(3) Online operation

Step 3-1: Set the torque gain coefficient optimization period as _Tw , initialize the torque gain coefficient K _d of the actual wind turbine operation, initialize the virtual wind turbine and the actual unit starting speed ω _bgn as the minimum speed in the MPPT stage, set Set the current period as n=1;

Step 3-2: Read the current measured wind speed value, the actual wind turbine adopts the torque gain reduction method optimized by the torque gain coefficient for MPPT control, and the virtual wind turbine in the PLC adopts the PSF method for MPPT control, and the two run synchronously;

发电机电磁转矩T_e；Step 3-3: Record the operation data of the virtual wind turbine, including rotor speed ω _r , rotor acceleration

generator electromagnetic torque _Te ;

并代入离线构建的函数关系

中预估出最优转矩增益系数

否则，返回执行步骤3-2；Step 3-4: Determine whether the operation of the nth _Tw period is over; if so, calculate the average wind energy capture efficiency corresponding to the PSF method of the virtual wind turbine in the current period according to the recorded operation data

And substitute the function relationship built offline

The optimal torque gain coefficient is estimated from

Otherwise, return to step 3-2;

设定为实际风电机组第n+1个时段的转矩增益系数；Step 3-5: Apply the optimal torque gain coefficient given in Step 3-4

Set as the torque gain coefficient of the n+1th period of the actual wind turbine;

Step 3-6: n=n+1, jump to step 3-2, and enter the next operation cycle.

替代平均风速、湍流强度、湍流频率三个风况特征指标，构建其与风电机组MPPT阶段最优转矩增益系数之间的函数关系，与现有离线遍历、在线优化类的算法，如神经网络、响应面模型等相比，保证风能捕获效率的同时，大幅降低算力资源与遍历时间的要求，具有较强的工程实用性。Compared with the prior art, the present invention has the following significant advantages: (1) The present invention directly establishes the relationship between the optimal gain coefficient and the wind condition characteristics, thereby avoiding the adaptive iterative search process and the possible problem of non-convergence of the search. Strong adaptability to changes in wind conditions; (2) The power curve method introduced in the present invention corresponds to the wind energy capture efficiency

Substitute the three wind condition characteristic indicators of average wind speed, turbulence intensity and turbulence frequency, and construct the functional relationship between them and the optimal torque gain coefficient in the MPPT stage of the wind turbine, which is compatible with the existing offline traversal and online optimization algorithms, such as neural networks. Compared with the response surface model, the wind energy capture efficiency is guaranteed, and the requirements for computing resources and traversal time are greatly reduced, which has strong engineering practicability.

The present invention will be described in further detail below with reference to the accompanying drawings.

Description of drawings

FIG. 1 is a flow chart of a method for tracking and controlling the maximum power point of a wind turbine based on torque gain coefficient optimization according to the present invention.

Figure 2 is a schematic diagram of the statistical relationship between the optimal gain coefficient of the NREL CART3 wind turbine and its average wind energy capture efficiency corresponding to the application of the PSF method.

FIG. 3 is a comparison diagram of the implementation effect of the method of the present invention and other methods.

Detailed ways

The invention belongs to the method type of constructing the direct quantitative relationship between the optimal adjustment parameter and the MPPT influencing factor characterization index to guide the online optimization of the parameter, and has higher wind energy capture performance. However, the present invention further realizes the characterization of a single index of the comprehensive influence of multiple indexes on the MPPT, simplifies the difficulty of constructing a direct quantitative relationship, and can obtain higher wind energy capture efficiency under the premise of paying less computing power resources and time costs. It has engineering application value.

1, the present invention firstly constructs the functional relationship between the optimal torque gain coefficient and the wind energy capture efficiency corresponding to the PSF method, and periodically adjusts the torque curve gain coefficient during the actual operation of the wind turbine according to the functional relationship. In order to improve the MPPT control method in the actual operation of wind turbines, the problem that the wind energy capture efficiency corresponding to the PSF method cannot be directly known, is solved by constructing a virtual wind turbine applying the PSF method in the controller to run synchronously with the actual wind turbine.

Among them, the steps of offline construction of function relationship are as follows:

Step 1-1: For the wind turbine to which the present invention is to be applied, obtain the aerodynamic and structural parameters required for building its FAST model;

Step 1-2: in the FAST software, complete the construction of the simulation model of the wind turbine applying the present invention according to the parameters;

湍流强度TI和湍流积分尺度L，生成对应于不同特征指标组合的湍流风速序列。其中，

的变化范围为4～9m/s，步长为1m/s，湍流强度TI根据A、B、C湍流级别变化，积分尺度L变化范围为100～500m，步长为50m。一共可获得162种参数设置组合风况，对应每种风况各生成10条风速序列；Step 1-3: Using the turbulent wind speed simulation method, change the three characteristic indicators that characterize the turbulent wind condition in turn, namely the average wind speed

Turbulence intensity TI and turbulence integral scale L are used to generate turbulent wind speed series corresponding to different characteristic index combinations. in,

The variation range is 4~9m/s, the step size is 1m/s, the turbulence intensity TI changes according to the A, B, C turbulence levels, the integral scale L changes in the range of 100~500m, and the step size is 50m. A total of 162 kinds of parameter setting combined wind conditions can be obtained, and 10 wind speed sequences are generated corresponding to each wind condition;

以及应用PSF法时的风能捕获效率

其中Step 1-4: Traverse the wind turbine optimal torque gain coefficient corresponding to each wind speed sequence based on FAST software

and wind energy capture efficiency when applying the PSF method

in

即平均风能捕获效率最大时对应的K_d；In the formula, P _cap represents the actual captured power of the fan, P _wy is the maximum wind power contained in the air, v is the wind speed, _{ng is the gear ratio of the gearbox, T e represents} the electromagnetic torque, ω _r represents the rotational speed, and J _t represents the rotation Inertia, ρ is the air density, R is the radius of the rotor. Optimum torque gain coefficient

That is, K _d corresponding to the maximum average wind energy capture efficiency;

和PSF法下平均风能捕获效率

的函数关系

Step 1-5: Use the 1620 sets of results obtained by traversal as sample data to fit the optimal torque gain coefficient

and the average wind energy capture efficiency under the PSF method

functional relationship

The steps to construct a virtual wind turbine are as follows:

Step 2-1: Embed the FAST model of the wind turbine to which the present invention is to be applied into its actual main control PLC to construct a virtual unit that can run synchronously with the actual unit;

Step 2-2: Embed the PSF method code for the MPPT control of the virtual wind turbine in the main control PLC.

The online operation steps are as follows:

Step 3-1: Set the torque gain coefficient optimization period as _Tw , initialize the torque gain coefficient K _d of the actual wind turbine operation, initialize the virtual wind turbine and the actual unit starting speed ω _bgn as the minimum speed in the MPPT stage, set The current period is set as n=1; the torque gain coefficient optimization period Tw is set to 10min~1h, the torque gain coefficient K _d for initializing the actual operation of the wind turbine is _{0.9K opt} ~ 0.98K _opt , and K _opt is PSF torque gain coefficient of the method;

Step 3-2: Read the current measured wind speed value, the actual wind turbine adopts the torque gain reduction method optimized by the torque gain coefficient for MPPT control, and the virtual wind turbine in the PLC adopts the PSF method for MPPT control, and the two run synchronously;

发电机电磁转矩T_e；Step 3-3: Record the operation data of the virtual wind turbine, including rotor speed ω _r , rotor acceleration

generator electromagnetic torque _Te ;

并代入离线构建的函数关系

中预估出最优转矩增益系数

否则，返回执行步骤3-2；Step 3-4: Determine whether the operation of the nth _Tw period ends. If so, calculate the average wind energy capture efficiency corresponding to the PSF method of the virtual wind turbine in the current period according to the recorded operation data

And substitute the function relationship built offline

The optimal torque gain coefficient is estimated from

Otherwise, return to step 3-2;

设定为实际风电机组第n+1个时段的转矩增益系数；Step 3-5: Apply the optimal torque gain coefficient given in Step 3-4

Set as the torque gain coefficient of the n+1th period of the actual wind turbine;

Step 3-6: n=n+1, jump to step 3-2, and enter the next operation cycle.

The invention belongs to the method type of constructing the direct quantitative relationship between the optimal adjustment parameter and the MPPT influencing factor characterization index to guide the parameter online optimization, and can avoid the efficiency drop caused by the adaptive algorithm iterative search not converging or the wrong direction. In addition, the present invention realizes the characterization of a single index of the comprehensive influence of multiple indexes on the MPPT, and can simplify the construction complexity of the direct quantitative relationship. Therefore, it can greatly reduce the computing resource requirements while ensuring the wind energy capture efficiency, and has strong engineering practicability.

Below in conjunction with embodiment, the present invention is described in further detail:

Example

Taking the 0.6MW CART3 model of the Renewable Energy Laboratory (NREL) of the U.S. Department of Energy as the application object, through the simulation calculation and statistical analysis of the simulated wind speed sequence, the maximum power of the wind turbine based on the torque gain coefficient optimization proposed by the present invention is analyzed. The point tracking control method compares the wind energy capture efficiency with the traditional power curve method and adaptive torque control, and constructs the function relationship between the optimal torque curve adjustment amount and the three wind condition characteristics offline. The online application method consumes computing resources and fast Practical comparison to verify the effectiveness and superiority of the present invention.

(1) Simulation model

The simulation model adopts the open source professional wind turbine simulation software FAST provided by the Renewable Energy Laboratory (NREL) of the US Department of Energy. The wind turbine model corresponds to the 0.6MW CART3 model developed by NREL, and its related parameters are as follows.

Table 1 Main parameters of NREL 0.6MW CART3 wind turbine

(2) Simulation realization

According to the steps described in the content of the invention, the details are as follows:

Build functional relationships offline:

Step 1-1: For the CRAT3 wind turbine to which the present invention is to be applied, obtain the aerodynamic and structural parameters required for building its FAST model, as described in Table 1;

Step 1-2: in the FAST software, complete the construction of the simulation model of the wind turbine applying the present invention according to the parameters;

湍流强度TI和湍流积分尺度L)，生成对应于不同特征指标组合的湍流风速序列。其中，

的变化范围为4～9m/s，步长为1m/s，湍流强度TI根据A、B、C湍流级别变化，积分尺度L变化范围为100～500m，步长为50m。一共可获得162种参数设置组合风况，对应每种风况各生成10条风速序列；Step 1-3: Using the turbulent wind speed simulation method, change the three characteristic indicators (average wind speed) that characterize the turbulent wind condition in turn.

Turbulence intensity TI and turbulence integral scale L) are used to generate turbulent wind speed series corresponding to different characteristic index combinations. in,

The variation range is 4~9m/s, the step size is 1m/s, the turbulence intensity TI changes according to the A, B, C turbulence levels, the integral scale L changes in the range of 100~500m, and the step size is 50m. A total of 162 kinds of parameter setting combination wind conditions can be obtained, and 10 wind speed sequences are generated corresponding to each wind condition;

以及应用PSF法时的风能捕获效率

其中Step 1-4: Traverse the wind turbine optimal torque gain coefficient corresponding to each wind speed sequence based on FAST software

and wind energy capture efficiency when applying the PSF method

in

即平均风能捕获效率最大时对应的K_d；In the formula, P _cap represents the actual captured power of the fan, P _wy is the maximum wind power contained in the air, v is the wind speed, _{ng is the gear ratio of the gearbox, T e represents} the electromagnetic torque, ω _r represents the rotational speed, and J _t represents the rotation inertia. Optimum torque gain coefficient

That is, K _d corresponding to the maximum average wind energy capture efficiency;

和PSF法下平均风能捕获效率

的函数关系

对应CART3风电机组最优转矩增益系数

和PSF法下平均风能捕获效率

的关系，如图2所示。Step 1-5: Use the 1620 sets of results obtained by traversal as sample data to fit the optimal torque gain coefficient

and the average wind energy capture efficiency under the PSF method

functional relationship

Corresponding to the optimal torque gain coefficient of CART3 wind turbine

and the average wind energy capture efficiency under the PSF method

relationship, as shown in Figure 2.

的具体表达式为：

The specific expression is:

Virtual wind turbine construction:

Step 2-1: Embed the FAST model of the wind turbine to which the present invention is to be applied into its actual main control PLC to construct a virtual unit that can run synchronously with the actual unit;

Step 2-2: Embed the PSF method code for the MPPT control of the virtual wind turbine in the main control PLC.

Running online:

Step 3-1: Set the torque gain coefficient optimization period as Tw = 10min, initialize the torque gain coefficient K _d = _{0.9K opt} of the actual wind turbine operation, initialize the virtual wind turbine and the actual unit starting speed ω _bgn as MPPT The minimum speed of the stage, set the current period as n=1;

Step 3-2: Read the current measured wind speed value, the actual wind turbine adopts the torque gain reduction method optimized by the torque gain coefficient for MPPT control, and the virtual wind turbine in the PLC adopts the PSF method for MPPT control, and the two run synchronously;

发电机电磁转矩T_e；Step 3-3: Record the operation data of the virtual wind turbine, including rotor speed ω _r , rotor acceleration

generator electromagnetic torque _Te ;

并代入离线构建的函数关系

中预估出最优转矩增益系数

否则，返回执行步骤3-2；Step 3-4: Determine whether the operation of the nth _Tw period is over; if so, calculate the average wind energy capture efficiency corresponding to the PSF method of the virtual wind turbine in the current period according to the recorded operation data

And substitute the function relationship built offline

The optimal torque gain coefficient is estimated from

Otherwise, return to step 3-2;

设定为实际风电机组第n+1个时段的转矩增益系数；Step 3-5: Apply the optimal torque gain coefficient given in Step 3-4

Set as the torque gain coefficient of the n+1th period of the actual wind turbine;

Step 3-6: n=n+1, jump to step 3-2, and enter the next operation cycle.

(3) Computational resources consumed

Compared with the complex nonlinear function relationship between the optimal torque gain coefficient and the three wind characteristics of average wind speed, turbulence intensity and turbulence frequency, to obtain the same sample fineness (that is, after the application, the wind energy obtained at the statistical level) The capture efficiency is the same), the sample size required by the present invention to construct the relationship between the optimal torque gain coefficient and the wind energy capture efficiency corresponding to the PSF method, that is, the computing resource is only 1% of it. For this embodiment, it only takes about 2 days to complete the offline traversal of the sample size required by the present invention by using an ordinary i7 quad-core workstation.

(4) Comparative analysis of wind energy capture efficiency

For 50 measured turbulent wind speed sequences with a duration of 4 hours, the present invention applies the PSF method, the adaptive torque control and the improved method proposed by the present invention to simulate respectively. For each wind speed series, the percentage increase in P _favg of different MPPT methods relative to the PSF method was calculated. Table 2 gives the statistical average of 50 groups of calculation examples.

Table 2 Comparison of different MPPT control methods

It can be seen from Table 2 that, compared with the traditional PSF method and the adaptive torque method, the MPPT method of the wind turbine provided by the present invention can improve the wind energy capture efficiency.

Combining with a specific example, the torque gain coefficient, theoretical optimal torque gain coefficient and average wind energy capture efficiency given by each method in each 10min period are shown in detail, as shown in Figure 3. It can be seen that the adaptive torque method has the phenomenon of non-convergence of the search, and will give a gain coefficient estimation with a large deviation from the optimal value. Comparatively speaking, the estimated value of the gain coefficient given by the present invention is closer to the optimal value, and higher wind energy capture efficiency can be obtained. Specifically, the adaptive torque control is improved by 1.29% compared with the traditional PSF method, and the method proposed in the present invention further improves the efficiency of the adaptive torque by 0.70% on this basis. In this embodiment, the method of the present invention is compared with other existing MPPT control methods for wind energy capture efficiency and computing power resource consumption, and the effectiveness and superiority of the method are verified.

Claims (4)

1. a wind turbine maximum power point tracking control method based on torque gain coefficient optimization, is characterized in that, comprises the following steps: (1) Offline construction of functional relationships Step 1-1: For the wind turbine to be applied, obtain the aerodynamic and structural parameters required to build its FAST model; Step 1-2: In the FAST software, complete the construction of the wind turbine simulation model according to the parameters; Step 1-3: Use the turbulent wind speed simulation method to sequentially change the three characteristic indicators that characterize the turbulent wind condition: average wind speed

Turbulence intensity TI and turbulence integral scale L are used to generate turbulent wind speed series corresponding to different characteristic index combinations; Step 1-4: Traverse the wind turbine optimal torque gain coefficient corresponding to each wind speed sequence based on FAST software

and wind energy capture efficiency when applying the PSF method

in

In the formula, P _cap represents the actual captured power of the fan, P _wy is the maximum wind power contained in the air, v is the wind speed, _{ng is the gear ratio of the gearbox, T e represents} the electromagnetic torque, ω _r represents the rotational speed, and J _t represents the rotation Inertia, ρ represents air density, R is rotor radius; optimal torque gain coefficient

That is, the K _d corresponding to the maximum average wind energy capture efficiency; Step 1-5: Fit the optimal torque gain coefficient with the traversal results as sample data

and the average wind energy capture efficiency under the PSF method

functional relationship

(2) Construction of virtual wind turbines Step 2-1: Embed the FAST model of the wind turbine to be applied into its actual main control PLC to construct a virtual unit that can run synchronously with the actual unit; Step 2-2: Embed the PSF method code for the MPPT control of the virtual wind turbine in the main control PLC; (3) Online operation Step 3-1: Set the torque gain coefficient optimization period as _Tw , initialize the torque gain coefficient K _d of the actual wind turbine operation, initialize the virtual wind turbine and the actual unit starting speed ω _bgn as the minimum speed in the MPPT stage, set Set the current period as n=1; Step 3-2: Read the current measured wind speed value, the actual wind turbine adopts the torque gain reduction method optimized by the torque gain coefficient for MPPT control, and the virtual wind turbine in the PLC adopts the PSF method for MPPT control, and the two run synchronously; Step 3-3: Record the operation data of the virtual wind turbine, including rotor speed ω _r , rotor acceleration

generator electromagnetic torque _Te ; Step 3-4: Determine whether the operation of the nth _Tw period is over; if so, calculate the average wind energy capture efficiency corresponding to the PSF method of the virtual wind turbine in the current period according to the recorded operation data

And substitute the function relationship built offline

The optimal torque gain coefficient is estimated from

Otherwise, return to step 3-2; Step 3-5: Apply the optimal torque gain coefficient given in Step 3-4

Set as the torque gain coefficient of the n+1th period of the actual wind turbine; Step 3-6: n=n+1, jump to step 3-2, and enter the next operation cycle. 2. The wind turbine maximum power point tracking control method based on torque gain coefficient optimization according to claim 1, wherein, in steps 1-3,

The variation range is 4~9m/s, the step size is 1m/s, the turbulence intensity TI changes according to the A, B, C turbulence levels, the integral scale L changes in the range of 100~500m, and the step size is 50m; a total of 162 types can be obtained. The parameter sets the combined wind conditions, and generates 10 wind speed sequences corresponding to each wind condition. 3. The maximum power point tracking control method for wind turbines based on torque gain coefficient optimization according to claim 2, wherein in steps 1-5, 1620 groups of results obtained by traversal are used as sample data to fit the optimal Torque gain factor

and the average wind energy capture efficiency under the PSF method

functional relationship

4. The maximum power point tracking control method for wind turbines based on torque gain coefficient optimization according to claim 1, characterized in that in step 3-1, the torque gain coefficient optimization period T _w is set as 10min～1h, The torque gain coefficient K _d for initializing the actual operation of the wind turbine is 0.9K _opt ~ 0.98K _opt , where K _opt is the torque gain coefficient of the PSF method.

Images