CN113217275B - Wind turbine generator set variable pitch control method based on strategy iteration - Google Patents

Abstract

The invention relates to a control method of a wind turbine generator variable pitch based on strategy iteration _r And the actual generator power P _g Obtaining the optimal pitch angle beta through a strategy iterative algorithm according to the difference value and the actual pitch angle beta of the wind wheel; when the difference value epsilon (k) between the value V (k) and the value V (k-1) reaches the preset algorithm precision epsilon, obtaining the optimal pitch angle beta; the variable pitch mechanism realizes variable pitch control according to beta; under the condition of the variable pitch control, wind energy captured by the rotation of the wind wheel is output by the generator and is the actual power P of the generator _g And the actual generator power P is converted into _g The signal is transmitted to a strategic iterative pitch controller. The invention can effectively stabilize the output power fluctuation caused by the wind speed change, improve the output power of the wind turbine generator, and reduce the extra fatigue load caused by the redundant actions of the variable pitch actuating mechanism.

Description

A strategy iteration-based pitch control method for wind turbines

technical field

The invention belongs to the technical field of wind power generation, and relates to a method for controlling the pitch of wind turbines based on strategy iteration, which belongs to the technical field of operation control of wind turbines.

Background technique

Wind turbines account for a large proportion of the world's electricity production. At the same time, higher requirements are put forward for the reliability of the wind turbine control system to ensure power generation and reduce operation and maintenance costs. However, in practical operation, the fluctuation of wind speed will lead to frequent pitch action, and the turbulent wind speed will cause additional fatigue load and output power fluctuation, which will have a significant negative impact on the mechanical structure of wind turbines and the stability of the power grid.

Due to the strong nonlinearity of wind turbines, the traditional control technology mainly focuses on the control design based on multiple operating points, usually using proportional-integral control for one or more operating points. But when the working point deviates from the working point, the control effect will decrease. In addition, the randomness of the wind speed leads to frequent switching of the operating point of the wind turbine, which brings further difficulties to the control design that satisfies the above-mentioned pitch control strategy. Modern control theory has been widely used in wind turbine control design, such as linear variable parameter control, model predictive control, nonlinear feedback control and so on. However, these methods still have many deficiencies in practical applications, such as the solving of the linear variable parameter control controller is very complicated, and the switching distance is far from the actual randomness.

SUMMARY OF THE INVENTION

Purpose of the invention: The present invention provides a method and system for pitch control of wind turbines based on strategy iteration, the purpose is to solve the problem that the fluctuation of wind speed causes additional fatigue load and output power fluctuation, which affects the service life and output stability of wind turbines.

Technical solutions:

A method for controlling the pitch of wind turbines based on strategy iteration. The strategy iteration pitch controller of the control system is based on the difference between the input rated motor power P _r and the actual generator power P _g , as well as the actual pitch of the wind rotor. Pitch angle β, the optimal pitch angle β* is obtained through the strategy iteration algorithm; when the difference ε(k) between the value of V(k) and V(k-1) reaches the preset algorithm accuracy ε, the optimal pitch angle is obtained angle β*; the pitch mechanism realizes the pitch control according to β*; under the condition of the pitch control, the wind energy captured by the rotation of the wind rotor is output by the generator, which is the actual generator power P _g , and the actual power is generated. The engine power _Pg signal is sent to the strategy iterative pitch controller.

Further, the steps of the policy iteration algorithm are as follows:

1) According to the rated power P _r , the actual generator power P _g and the pitch angle β, establish the performance index function J(k);

2) According to the performance index function J(k) in step 1) and the adjustable parameters Q, R and discount factor γ given by the designer, the Bellman equation of the pitch tracking control is obtained;

3) The Bellman equation in step 2) is based on the Bellman optimality principle to obtain the Bellman optimal equation, which includes the optimal value function V* and the optimal pitch angle β*:

4) According to the Bellman optimal equation in step 3), and by making the difference between the actual generator power P _g and the rated power P _r , a two-step iterative equation is obtained;

5) Set the control system performance index function;

6) In combination with the form of the performance index function in step 1), a quadratic value function V ¹ (k) is given, an initial allowable control β(1) and a preset algorithm precision ε are given;

7) From the iteration index k=1,...,n, execute the two-step iterative equation in step 4);

When V(k) in the Bellman equation of the tracking control problem in step 2) is equal to the control system performance index function J(k) assumed in step 4), the strategy evaluation formula is obtained. According to the strategy evaluation formula, solve P( k) matrix;

8) Substitute the P(k) matrix obtained in step 7) into the strategy improvement formula to solve the pitch angle β(k);

9) Through the iterative solution of the value function V(k) and the pitch angle β(k), when the difference between the value of V(k) and V(k-1) calculated by the strategy iteration equation in step 8) is ε(k ) when the preset algorithm precision ε is reached, the optimal pitch angle β* is obtained, and the strategy iteration algorithm stops.

The formula of the performance index function J(k) in step 1) is,

In the formula: P _r is the rated generating power, P _g is the actual generator power, β is the pitch angle, R and Q are the preset parameter matrix given by the designer, and γ is the discount factor.

The Bellman equation of the tracking control problem in step 2) is,

In the formula: P _r is the rated generating power, P _g is the actual generator power, β is the pitch angle, R and Q are the preset parameter matrix given by the designer, and γ is the discount factor.

The optimal value function in step 3) is

In the formula: utility function U(K)=(P _r (k)-P _g (k)) ^T Q(P _r (k)-P _g (k))+β(k) ^T Rβ(k), min Represents the minimum value that can be achieved in the set of all β(k);

The optimal pitch angle is

In the formula: β* is the optimal pitch angle, V*(k+1) is the optimal value function of the k+1th step, and argmin means when U(k)+γV ^* (k+1) achieves the minimum value β(k).

Two-step iterative equations are obtained in step 4);

Strategy Evaluation:

Policy Improvements:

In step 5), the control system value function is:

V(x)=x(k) ^T Px(k)

In the formula: x(k) is the system state.

The strategy evaluation formula in step 7):

The strategy improvement formula in step 8):

Beneficial effects: Through the iterative learning process of the strategy iterative pitch controller, the pitch controller can fully learn the internal relationship between the wind speed change and the pitch angle, and give a reasonable pitch angle control signal. Therefore, the output power fluctuation caused by the change of wind speed can be effectively suppressed, the output power of the wind turbine can be improved, and the additional fatigue load caused by the redundant action of the pitch actuator can be reduced.

Description of drawings

Fig. 1 is the overall structure diagram of the pitch control system of the wind turbine of the present invention;

Fig. 2 is the flow chart of the pitch control method of wind turbine based on strategy iteration;

Figure 3 is the incoming wind speed of the wind turbine;

Fig. 4 is the pitch angle of the pitch system based on strategy iteration;

Figure 5 shows the output power of the wind turbine.

Detailed ways

In order to describe in detail the technical content, achieved objects and effects of the present invention, further detailed descriptions are given below in conjunction with the accompanying drawings and specific embodiments.

As shown in Figure 1, it is the overall structure diagram of the wind turbine pitch control method, a control method for wind turbine pitch control based on strategy iteration. The strategy iteration pitch controller of the control system is based on the input rated motor power. The difference between P _r and the actual generator power P _g , the actual pitch angle β of the wind rotor, the optimal pitch angle β* is obtained through the strategy iterative algorithm; when the difference between the value of V(k) and the value of V(k-1) When ε(k) reaches the preset algorithm accuracy ε, the optimal pitch angle β* is obtained; the pitch actuator realizes the pitch control according to β*; under the condition of the pitch control, the rotation of the wind rotor captures the The output power of the wind power generator is the actual generator power P _g , and the actual generator power P _g signal is transmitted to the strategy iterative pitch controller to form a cyclic control system.

As shown in Figure 2, the steps of the policy iteration algorithm are as follows:

1) Detect the actual generator power P _g of the generator and the actual pitch angle β of the wind turbine. According to the model of the target wind turbine, obtain the rated power P _r given by the wind turbine. According to the rated power P _r , the actual generator power P _g and pitch angle β, establish the performance index function J(k);

In the formula: P _r is the rated generating power, P _g is the actual generator power, β is the pitch angle, R and Q are the preset parameter matrix given by the designer, and γ is the discount factor.

2) According to the performance index function J(k) in step 1) and the adjustable parameters Q, R and discount factor γ given by the designer, the Bellman equation of the tracking control problem can be obtained:

where V(k) is the value function, P _r is the rated power generation, P _g is the actual generator power, β(k) is the pitch angle, R and Q are the preset parameter matrix given by the designer, γ is the discount factor.

3) The Bellman equation in step 2) According to the Bellman optimality principle, the Bellman optimal equation and the Bellman optimal equation are obtained, including the optimal value function V* and the optimal pitch angle β*:

The optimal value function is

In the formula: utility function U(K)=(P _r (k)-P _g (k)) ^T Q(P _r (k)-P _g (k))+β(k) ^T Rβ(k), min Represents the minimum value that can be achieved in the set of all β(k);

The optimal pitch angle is

In the formula: β* is the optimal pitch angle, V*(k+1) is the optimal value function of the k+1th step, and argmin means when U(k)+γV ^* (k+1) achieves the minimum value β(k).

4) The input of the strategy iterative controller is divided into: the measured actual generator power P _g and the rated power P _r , and the current actual pitch angle β, and the output is the optimal pitch angle β*. The goal of the strategy iterative control is to make the actual generator power P _g track the rated power P _r , that is, the generator power tracking error e tends to zero. According to the Bellman optimal equation in step 3), and by making the difference between the actual generator power Pg and the rated generated power Pr, the generated power tracking error e is obtained, that is, _e = _Pg - _Pr , and a two-step iterative equation is obtained;

5) Set the control system value function;

V(k)=x(k)P ⁱ⁺¹ x(k)

where x(k) is the system state.

6) In combination with the quadratic value function V ¹ (k) given in step 1), the initial allowable pitch control β(1) and the preset algorithm precision ε are given;

7) From the iteration index k=1,...,n, execute the two-step iterative equation in step 4);

When the value function V(k) of the tracking control problem in step 2) is equal to the control system performance index function J(k) assumed in step 4), the strategy evaluation formula is obtained, and the P matrix is solved according to the strategy evaluation formula,

Strategy evaluation formula:

8) Substitute the P matrix obtained in step 7) into the strategy improvement formula to solve the pitch angle β(k);

Strategy improvement formula:

9) Through the iterative solution of the value function V(k) and the pitch angle β(k), when the difference between the β(k) value calculated by the strategy iteration equation in step 8) and the β(k-1) value ε(k ) when the preset algorithm accuracy ε is reached, the optimal pitch angle β* is obtained, that is, the pitch angle β corresponding to ε(k) at this time is β*, and the strategy iteration algorithm stops.

For the variable pitch control of the 1.2MW wind turbine, the incoming wind speed of the wind turbine is shown in Figure 3. From Figure 3, it can be seen that the wind speed changes greatly and frequently. The pitch angle control signal of the pitch system is shown in Figure 4. From 4, it can be seen that the pitch angle of the control method of the strategy iterative pitch pitch changes smoothly. The corresponding wind turbine output power is shown in Figure 5. It can be seen from Figure 5 that the control method of wind turbine pitch control based on strategy iteration proposed in this patent can effectively suppress wind disturbance, reduce fatigue load, and make the wind turbine The output power is stable around the rated power.

To sum up, the method of the present invention can effectively improve the control performance of the variable pitch system of the wind turbine, stabilize the output power near the rated power, adapt to a wide range of changing operating points, suppress wind disturbance, reduce fatigue load, and has a wide range of applications Prospects are worth promoting.

The above-mentioned embodiments are only preferred embodiments of the present invention, and are not intended to limit the scope of implementation of the present invention. Therefore, any changes made according to the methods and principles of the present invention should be included within the protection scope of the present invention.

Claims (5)

1. A control method for wind turbine generator pitch variation based on strategy iteration is characterized in that: the strategy iteration variable pitch controller of the control method is used for controlling the variable pitch according to the input rated power P _r And the actual generator power P _g Obtaining the optimal pitch angle beta through a strategy iterative algorithm according to the difference value and the actual pitch angle beta of the wind wheel; when the difference value epsilon (k) of the value function V (k) and the value function V (k-1) reaches the preset algorithm precision epsilon, obtaining the optimal pitchAngle β; the variable pitch mechanism realizes variable pitch control according to beta; under the condition of the variable pitch control, wind energy captured by the rotation of the wind wheel is output by the generator and is the actual generator power P _g And the actual generator power P is converted into _g Transmitting the signal to a strategy iteration variable pitch controller;

the strategy iterative algorithm comprises the following steps:

1) according to rated power P _r Actual generator power P _g Establishing a performance index function J (k) according to the pitch angle beta;

2) obtaining a Bellman equation for variable pitch tracking control according to the performance index function J (k) in the step 1), the adjustable parameter Q, R given by a designer and the discount factor gamma;

3) the Bellman equation in the step 2) obtains a Bellman optimal equation according to a Bellman optimality principle, wherein the equation comprises an optimal value function V and an optimal pitch angle beta:

4) according to the Bellman optimal equation in the step 3) and through the actual generator power P _g Rated power P _r Making a difference to obtain a two-step iteration equation;

5) setting a control system performance value function;

6) combining the form of the performance index function in the step 1), giving a quadratic value function V ¹ (k) Setting an initial allowable control beta (1) and a preset algorithm precision epsilon;

7) performing the two-step iteration equation in step 4) from the iteration index k being 1, …, n;

when V (k) in the Bellman equation of the tracking control problem in the step 2) is equal to the performance index function J (k) in the step 1), obtaining a strategy evaluation formula, and solving a P (k) matrix according to the strategy evaluation formula;

8) substituting the matrix P (k) obtained in the step 7) into a strategy improvement formula to solve the pitch angle beta (k);

9) and (3) obtaining the optimal pitch angle beta by iterative solution of the value function V (k) and the pitch angle beta (k) when the difference epsilon (k) between the value V (k) and the value V (k-1) calculated by the strategy iterative equation in the step 8) reaches the preset algorithm precision epsilon, so that the strategy iterative algorithm stops.

2. The strategy iteration based control method for wind turbine generator pitch according to claim 1, characterized in that: in the step 1), the performance index function J (k) has the formula,

in the formula: p is _r For rated power generation, P _g For the actual generator power, β is the pitch angle, R and Q are the preset parameter matrices given by the designer, and γ is the discount factor.

3. The strategy iteration based control method for wind turbine pitch control according to claim 1, characterized by: the bellman equation for the tracking control problem in step 2) is,

in the formula: p _r For rated power generation, P _g For the actual generator power, β is the pitch angle, R and Q are the preset parameter matrices given by the designer, and γ is the discount factor.

4. The strategy iteration based control method for wind turbine generator pitch according to claim 1, characterized in that: the optimal value function in the step 3) is

In the formula: utility function u (k) ═ (P) _r (k)-P _g (k)) ^T Q(P _r (k)-P _g (k))+β(k) ^T R β (k), min represents the minimum value that can be obtained in the set of all β (k);

the optimum pitch angle is

In the formula: β is the optimum pitch angle, V (k +1) is the optimum function of step k +1, and argmin represents β (k) when u (k) + γ V (k +1) takes a minimum.

5. Method for controlling the pitch of a wind turbine according to claim 4, characterised in that: obtaining a two-step iteration equation in the step 4);

and (3) policy evaluation:

strategy improvement:

in the formula: u (K) is a utility function.

Images