A fuzzy adaptive optimal power tracking control method for wind turbine

Abstract

The application provides a fuzzy adaptive optimal power tracking control method of a wind turbine, comprising a wind turbine real-time data acquisition and preprocessing module, an online optimization calculation module based on an improved whale optimization algorithm (IWOA), an effective wind speed estimation module based on the IWOA and a kernel based extreme learning machine (KELM) algorithm, and a fuzzy adaptive optimal power tracking controller module of a generator. The application designs an adaptive fuzzy output power tracking control method of a wind generator, so as to realize the smooth and optimal output power of the wind generator and improve the power generation of the wind generator.

Description

Technical Field

[0001] This invention relates to the field of fuzzy adaptive optimal power point tracking control methods for wind turbines in wind power generation systems, specifically to an online power optimization calculation method based on an improved whale optimization algorithm and a fuzzy adaptive optimal power point tracking control method. Background Technology

[0002] Due to the influence of factors such as turbulence, tower, wind shear difference, and surface roughness, the wind speed distribution on the entire plane of rotation of a wind turbine is different. This results in a large deviation between the optimal actual output power curve (called the dynamic power curve) and the theoretical power design curve (called the static power curve), which poses a great challenge to the operation and control of wind power generation.

[0003] The annual power generation of a wind turbine is determined by its dynamic power curve, not its static power curve. Existing wind turbine power point tracking (PPT) methods all use the static power curve as the tracking target, failing to consider the influence of wind turbulence intensity, wind shear, surface roughness, and complex terrain, which can cause significant deviations between the dynamic and static power curves, resulting in unsatisfactory control performance. Therefore, this invention constructs an online computational optimization module based on an improved IWOA algorithm and a multi-objective optimization model to obtain the dynamic power curve of the wind turbine's output power, providing a tracking target for optimal power point tracking control of the wind turbine.

[0004] Wind power generation systems are highly nonlinear systems, making it extremely difficult, if not impossible, to establish accurate mathematical models. This invention utilizes the powerful nonlinear approximation capabilities of fuzzy logic systems to approximate the unknown nonlinear time-varying functions of wind power generation systems using fuzzy basis functions, thus solving the problem of difficult modeling of wind power generation systems. Summary of the Invention

[0005] To address the challenges of mathematical modeling and online acquisition of optimal dynamic power curves in wind power generation systems, this invention provides a fuzzy adaptive optimal power tracking control method for wind turbines. It proposes an online optimization calculation method based on the improved whale optimization algorithm (IWOA) and a fuzzy adaptive optimal power tracking control design method, thus solving the two key problems mentioned above.

[0006] The present invention is achieved through the following technical solutions.

[0007] This invention provides a fuzzy adaptive optimal power point tracking control method for wind turbines, including a real-time data acquisition and preprocessing module for the wind turbine, an online optimization calculation module based on the improved whale optimization algorithm (IWOA), an effective wind speed estimation module based on IWOA and kernel extreme learning machine (KELM) algorithms, and a fuzzy adaptive optimal power point tracking controller module for the generator; and,

[0008] The wind turbine real-time data acquisition and preprocessing module acquires real-time data on wind turbine output power, generator speed, blade pitch angle, and wind speed.

[0009] The online optimization computation module based on the improved whale optimization algorithm (IWOA), i.e., the first microprocessor, is as follows: Figure 1 As shown, it includes a multi-objective optimization model and an improved Grey Wolf Optimization (IGWO) algorithm, specifically structured as follows:

[0010] a. Establish a multi-objective optimization model under full wind speed conditions, with wind turbine output power, generator speed, and blade pitch angle as optimization objectives:

[0011] minf(x)

[0012]

[0013] In the formula, f(x) is a multi-objective function, h(x) is an equality vector constraint function, and g(x) is an inequality vector constraint function;

[0014] The multi-objective function is:

[0015]

[0016]

[0017]

[0018]

[0019] In the formula T P The objective function is T, where T is the actual output power of the generator under full wind speed to track the expected predicted value. g The objective function is to minimize the control variations of generator speed and pitch angle; x is the decision variable; P gk ω gk and β k These are the actual output power, generator speed, and blade pitch angle at time k, respectively. and These are the expected predicted values ​​of generator output power, generator rotor angular velocity, and blade pitch angle at time k+1, respectively. w P and w gThese are the weighting coefficients; n is the optimization process, and P is the weighting coefficient. N P m These are the rated output power and theoretical design output power of the wind turbine, respectively.

[0020] a.1. Set the constraints for the multi-objective optimization model:

[0021] a.1.1. Aerodynamic power balance constraints

[0022]

[0023]

[0024] In the formula: ρ refers to air density (kg / m³) 3 ); v refers to wind speed, in m / s; R represents the radius of the wind turbine, in m; P m These are the theoretical design output power of the wind turbine, and β. k It is the pitch angle of the wind turbine at time k, in degrees, C. P (ω gk ,β k The wind energy utilization factor of the wind turbine is given by the following expression:

[0025]

[0026] In the formula, λ is the tip speed ratio of the fan blades, and n g It refers to the gearbox speed ratio.

[0027] a.1.2. Weighting coefficient constraints

[0028]

[0029] 0 < w p <1,

[0030] 0 < w g <1.

[0031] a.1.3. Output Power Constraint

[0032]

[0033] a.1.4. Pitch Angle and Generator Speed ​​Constraints

[0034]

[0035]

[0036] In the formula, β H and β L These are the upper and lower limits of the pitch angle, ω H and ωL These are the upper and lower limits of the generator speed.

[0037] b. Improved Whale-Optimized IWOA Algorithm:

[0038] b.1. Improved nonlinear convergence factor

[0039]

[0040] Where κ, τ, and μ are adjustment parameters, and t and t max These are the current iteration count and the maximum iteration count, respectively.

[0041] b.2. Nonlinear inertial weight ω:

[0042]

[0043] In the formula, γ is the adjustment parameter, and t and t max These are the current iteration count and the maximum iteration count, respectively.

[0044] b.3. Improved position update equation:

[0045] X(t+1)=ωX rand -A·D,|A|>1 and p<0.5

[0046] X(t+1)=ωX best (t)-A·D, |A|≤1 and p<0.5

[0047] X(t+1)=(1-ω)X best (t)+D p e bl cos(2πl), p≥0.5

[0048] In the formula, X(t+1) is the position vector of generation (t+1), X(t) is the position vector of generation t, and X... rand Let X be a position vector randomly selected from the current population. best (t) represents the optimal position vector so far, p is the probability of random selection, l is a random number in the interval [-1, 1], b represents the coefficients used to describe the spiral shape, and A, D, and D p It is a coefficient vector, and its expression is as follows:

[0049] A = a(2r1 - 1)

[0050] C = 2r²

[0051] D = |C·X rand -X(t)|

[0052] D p =|X best(t)-X(t)|

[0053] In the formula, r1 and r2 are random numbers in the interval [0,1].

[0054] The second microprocessor, such as Figure 1 As shown, the effective value of wind speed is used to estimate the wind speed and provides input variables for the fuzzy adaptive optimal power tracking controller (i.e., the third microprocessor) of the wind turbine.

[0055] The third microprocessor, such as Figure 1 As shown, its internal structure is as follows:

[0056]

[0057] In the formula, x = [x1, x2, ..., x n ] T ∈R n T is a fuzzy variable. em Here, e is the generator torque, and e is the generator output power tracking error, e = P. gopt -P gk K t J t These are the generator damping coefficient and the equivalent inertial constant, respectively. P1 is an unknown constant, and ε1 and ε2 are sufficiently small positive constants. The parameters θ = P1 are estimates, and k1, k2, k3, and k4 are strictly positive constants. It is an estimate of the unknown constant k1. It is an estimate of the vector parameter Φ, φ(x)=[φ1(x),φ2(x),…,φ m (x)] T It is a function vector, φ i (x) is a fuzzy basis function, and its expression is:

[0058]

[0059] In the formula, x represents j Belongs to fuzzy set The membership function is given by m, where m is the number of fuzzy rules, and in this invention, m = 2.

[0060] The originality of this invention lies in the following: An online optimization model is constructed with the generator output power, generator speed, and pitch angle as optimization objectives, aiming to achieve smooth and optimal generator output power. The linear convergence factor, inertia weight, and adaptive position update equation of the standard whale optimization algorithm are improved. This improved whale optimization algorithm is used to solve the online optimization model, obtaining the optimal predicted values ​​of generator output power, generator speed, and pitch angle. A fuzzy adaptive optimal power tracking controller for the wind turbine is designed. This controller uses the fuzzy basis functions of a fuzzy logic system to approximate the unknown nonlinear time-varying function of the wind power generation system, solving the problem of establishing an accurate mathematical model of the wind power generation system, improving the controller's control effect, exhibiting strong robustness and effectiveness, low cost, and ease of implementation. Attached Figure Description

[0061] Figure 1 This is a structural diagram of the fuzzy adaptive optimal power tracking control strategy for wind turbines according to the present invention. Detailed Implementation

[0062] The technical solution of the present invention is further described below, but the scope of protection is not limited to what is described.

[0063] like Figure 1 As shown, this invention provides a fuzzy adaptive optimal power point tracking control method for wind turbines, including a real-time data acquisition and preprocessing module for wind turbines, an online optimization calculation module based on the improved whale optimization algorithm (IWOA), an effective wind speed estimation module based on IWOA and kernel extreme learning machine (KELM) algorithms, and a fuzzy adaptive optimal power point tracking controller module for generators; and,

[0064] The wind turbine real-time data acquisition and preprocessing module acquires real-time data on wind turbine output power, generator speed, blade pitch angle, and wind speed.

[0065] The first microprocessor includes a multi-objective optimization model and an improved Grey Wolf Optimization (IGWO) algorithm, specifically configured as follows:

[0066] a. Establish a multi-objective optimization model under full wind speed conditions, with wind turbine output power, generator speed, and blade pitch angle as optimization objectives:

[0067] minf(x)

[0068]

[0069] In the formula, f(x) is a multi-objective function, h(x) is an equality vector constraint function, and g(x) is an inequality vector constraint function;

[0070] The multi-objective function is:

[0071]

[0072]

[0073]

[0074]

[0075] In the formula T P The objective function is T, where T is the actual output power of the generator under full wind speed to track the expected predicted value. g The objective function is to minimize the control variations of generator speed and pitch angle; x is the decision variable; P gk ω gk and β k These are the actual output power, generator speed, and blade pitch angle at time k, respectively. and These are the expected predicted values ​​of generator output power, generator rotor angular velocity, and blade pitch angle at time k+1, respectively. w P and w g These are the weighting coefficients; n is the optimization process, and P is the weighting coefficient. N P m These are the rated output power and theoretical design output power of the wind turbine, respectively.

[0076] a.1. Set the constraints for the multi-objective optimization model:

[0077] a.1.1. Aerodynamic power balance constraints

[0078]

[0079]

[0080] In the formula: ρ refers to air density (kg / m³) 3 ); v refers to wind speed, in m / s; R represents the radius of the wind turbine, in m; P m These are the theoretical design output power of the wind turbine, and β. k It is the pitch angle of the wind turbine at time k, in degrees, C. P (ω gk ,β k The wind energy utilization factor of the wind turbine is given by the following expression:

[0081]

[0082] In the formula, λ is the tip speed ratio of the fan blades, and n g It refers to the gearbox speed ratio.

[0083] a.1.2. Weighting coefficient constraints

[0084]

[0085] 0 < w p <1,

[0086] 0 < w g <1.

[0087] a.1.3. Output Power Constraint

[0088]

[0089] a.1.4. Pitch Angle and Generator Speed ​​Constraints

[0090]

[0091]

[0092] In the formula, β H and β L These are the upper and lower limits of the pitch angle, ω H and ω L These are the upper and lower limits of the generator speed.

[0093] b. Improved Whale-Optimized IWOA Algorithm:

[0094] b.1. Improved nonlinear convergence factor

[0095]

[0096] Where κ, τ, and μ are adjustment parameters, and t and t max These are the current iteration count and the maximum iteration count, respectively.

[0097] b.2. Nonlinear inertial weight ω:

[0098]

[0099] In the formula, γ is the adjustment parameter, and t and t max These are the current iteration count and the maximum iteration count, respectively.

[0100] b.3. Improved position update equation:

[0101] X(t+1)=ωX rand -A·D,|A|>1 and p<0.5

[0102] X(t+1)=ωX best (t)-A·D, |A|≤1 and p<0.5

[0103] X(t+1)=(1-ω)Xbest (t)+D p e bl cos(2πl), p≥0.5

[0104] In the formula, X(t+1) is the position vector of generation (t+1), X(t) is the position vector of generation t, and X... rand Let X be a position vector randomly selected from the current population. best (t) represents the optimal position vector so far, p is the probability of random selection, l is a random number in the interval [-1, 1], b represents the coefficients used to describe the spiral shape, and A, D, and D p It is a coefficient vector, and its expression is as follows:

[0105] A = a(2r1 - 1)

[0106] C = 2r²

[0107] D = |C·X rand -X(t)|

[0108] D p =|X best (t)-X(t)|

[0109] In the formula, r1 and r2 are random numbers in the interval [0,1].

[0110] The second microprocessor is used to estimate the effective value of the wind speed, providing input variables for the fuzzy adaptive optimal power tracking controller of the wind turbine (i.e., the third microprocessor);

[0111] The third microprocessor has the following internal structure:

[0112]

[0113] In the formula, x = [x1, x2, ..., x n ] T ∈R n T is a fuzzy variable. em Here, e is the generator torque, and e is the generator output power tracking error, e = P. gopt -P gk K t J t These are the generator damping coefficient and the equivalent inertial constant, respectively. P1 is an unknown constant, and ε1 and ε2 are sufficiently small positive constants. The parameters θ = P1 are estimates, and k1, k2, k3, and k4 are strictly positive constants. It is an estimate of the unknown constant k1. It is an estimate of the vector parameter Φ, φ(x)=[φ1(x),φ2(x),…,φ m (x)] T It is a function vector, φ i (x) is a fuzzy basis function, and its expression is:

[0114]

[0115] In the formula, x represents j Belongs to fuzzy set The membership function is given by m, where m is the number of fuzzy rules, and in this invention, m = 2.

Claims

1. A fuzzy adaptive optimal power point tracking control method for a wind turbine, comprising a real-time data acquisition and preprocessing module for the wind turbine, an online optimization calculation module based on the IWOA algorithm, an effective wind speed estimation module based on the IWOA algorithm and KELM, and a fuzzy adaptive optimal power point tracking controller module for the generator; The wind turbine real-time data acquisition and preprocessing module collects real-time data on generator output power, pitch angle, and generator speed, removes abnormal data, and provides input variables for the online optimization calculation module. The online optimization calculation module solves the multi-objective optimization model to obtain the optimal predicted values ​​of wind turbine generator output power, pitch angle and generator speed, providing input variables for the effective wind speed estimation module; The effective wind speed estimation module outputs an estimated value of the effective wind speed, which provides an input variable for the generator's fuzzy adaptive optimal power tracking controller. The generator’s fuzzy adaptive optimal power point tracking controller performs optimal power point tracking control on the wind turbine. The online optimization calculation module uses the preprocessing module to perform online optimization calculations on the real-time data collected from the wind turbine and the input variables provided by the preprocessing module to obtain the optimal predicted values ​​of generator output power, generator speed, and blade pitch angle. The online optimization calculation module also includes an improved IWOA algorithm and a multi-objective optimization model; a. Establish a multi-objective optimization model under full wind speed conditions, with wind turbine output power, generator speed, and blade pitch angle as optimization objectives: In the formula It is a multi-objective function. It is an equality vector constraint function. It is an inequality vector constraint function; The multi-objective function is: In the formula: The objective function is to track the actual output power of the generator to the expected predicted value under all wind speeds. The objective function is to minimize the control variations of generator speed and pitch angle. It is a decision variable; , and They are k Real-time output power, generator speed, and blade pitch angle; , and These are the generator output power, generator rotor angular velocity, and blade pitch angle. k Expected prediction value at time +1; , and These are weighting coefficients; n It's an optimization process. , These are the rated output power and theoretical design output power of the wind turbine, respectively. a.

1. Set constraints for the multi-objective optimization model: a.1.

1. Aerodynamic power balance constraints In the formula: This refers to air density (kg / m³) 3 ); v This refers to wind speed, measured in m / s. R Represents the radius of the wind turbine, in meters (m). These are the theoretical design output power of the wind turbine. It is a wind turbine in k The pitch angle at a given moment, in degrees. This refers to the wind energy utilization coefficient of the wind turbine generator. The relationship between the above parameters is expressed as follows: In the formula, It is the tip speed ratio of the fan blades. It refers to the gearbox speed ratio; a.1.

2. Weighting coefficient constraints a.1.

3. Output Power Constraint a.1.

4. Pitch Angle and Generator Speed ​​Constraints In the formula, and These are the upper and lower limits of the pitch angle. and These are the upper and lower limits of the generator speed; b. The improved IWOA algorithm is as follows: b.

1. Improved nonlinear convergence factor in , and It's about adjusting parameters. t and These are the current iteration count and the maximum iteration count, respectively. b.

2. Nonlinear inertial weight : In the formula, To adjust the parameters, t and These are the current iteration count and the maximum iteration count, respectively. b.

3. Improved position update equation: In the formula, For the first t +1 generation position vector, For the first t The position vector of the substitute, Let the position vector be randomly selected from the current population. This is the best position vector to date. p The probability of random selection. l It is a random number in the interval [-1, 1], and b represents the coefficient used to describe the spiral shape. A , D and It is a coefficient vector, and its expression is as follows: In the formula, , It is a random number in the interval [0, 1]. The input variable for the effective wind speed estimation module is the optimal predicted power. Optimal predicted rotational speed and optimal predicted pitch angle Its output variable is the estimated effective wind speed. .

2. The fuzzy adaptive optimal power point tracking control method for a wind turbine as described in claim 1, characterized in that: The control method of the fuzzy adaptive optimal power point tracking controller of the wind turbine is as follows: In the formula, For fuzzy variables, It is the generator torque. It is the generator output power tracking error. , , These are the generator damping coefficient and the equivalent inertial constant, respectively. , For an unknown constant, and They are sufficiently small positive numbers. , , It is a parameter The estimate k 1. k 2. k 3 and k 4 is a strictly positive number. yes k The estimate of 1 Vector parameters The estimate, It is a function vector. Let be a fuzzy basis function, and its expression is: In the formula, express Belongs to fuzzy set membership function, m For fuzzy rule numbers, m =2.

Images