CN-121980725-A - Proportional-integral state estimation method of nonlinear wind turbine system based on dynamic summation type event triggering

Abstract

The invention discloses a proportional-integral state estimation method of a nonlinear wind turbine system based on dynamic summation type event triggering. The method comprises the steps of 1, describing nonlinear dynamic characteristics of a wind turbine system based on a T-S fuzzy model method, establishing a state space model of the wind turbine system, 2, establishing an event trigger mechanism based on dynamic summation for saving network resources, reducing redundant trigger, 3, designing a proportional integral estimator to estimate the system state due to the fact that all system states are difficult to measure in actual engineering, 4, establishing a closed loop estimation error system by combining the T-S fuzzy wind turbine system and the proportional integral estimator, designing a state estimator gain matrix, and carrying out state estimation on the wind turbine system. Compared with the traditional event trigger mechanism based on the current information, the invention provides the event trigger mechanism based on the historical sampling data, introduces internal variables into trigger conditions, further expands trigger intervals, saves more network resources, and designs a dynamic trigger threshold which can be adaptively regulated along with the evolution of a system, thereby improving the flexibility of the trigger mechanism. In addition, compared with the traditional proportional estimator, the invention introduces an integral term of the state observation error to improve the precision of system state estimation, and has certain engineering application value.

Inventors

• YAN SHEN
• Dou Zehao
• YANG FAN

Assignees

• 南京林业大学

Dates

Publication Date

20260505

Application Date

20250701

Claims (1)

1. 1. The proportional-integral state estimation method of the nonlinear wind turbine system triggered by the dynamic summation type event is characterized by comprising the following steps of: Step 1, based on the characteristics of a T-S fuzzy wind turbine system, establishing a state space model of the system, wherein the method specifically comprises the following steps: Defining system states And the kinetic equation of a nonlinear wind turbine system can be expressed as Wherein v t (t) and v g (t) represent the angular speed of the wind turbine and the angular speed of the generator, respectively, Representing the spring section torque on the shaft, u (t) is the generator torque, i.e. the torque control input, ω (t) is the external disturbance, Indicating the compliance of the shaft and, And Representing the inertia of the wind turbine and the inertia of the generator respectively, Representing the damping of the shaft, f 1 、f 2 and f 3 are matrices of known parameters with appropriate dimensions, Representing the length of the blade, ρ represents the density of air, Representing the power coefficient of the wind turbine, S representing the rotor disk area. Assuming x 1 (t)∈[M 1 ,M 2 , using T-S fuzzy model to handle the nonlinearity of the wind turbine system, there is a rule i that if v t (T) belongs to Then Wherein, the Representing the fuzzy set associated with the ith rule. m u ＝1,...,s 1 ,i＝1,...,s 2 ,s 1 ＝1,s 2 =2 represents the number of preconditions and the number of fuzzy rules, respectively, and the system matrix is as follows: The state feedback controller u (T) =kx (y) is selected to calm the system, and then the closed loop T-S fuzzy system is: Wherein, the And Representing the system output and the performance output, respectively, v i (t) is a normalized membership function satisfying the following condition Step 2, in order to reduce redundant data transmission, a dynamic summation type event triggering mechanism based on historical data is designed, and the method specifically comprises the following steps: Wherein, the The error between the current time and the last time may be expressed as epsilon n (t)＝y(t k h-nh)-y(l k h-nh), n=0, 1. Phi (t) is the internal variable satisfies H is the sampling period, l k h＝t k h+lh.Φ n >0 is the weight matrix, λ and ρ are both positive scaling, δ n (lh)∈[δ nm ,δ nM ](0<δ nm ≤δ nM < 1) is the dynamic trigger threshold is met D n (t)＝t-t k h-lh+nh is defined, which satisfies nh=d nm ≤d n (t)≤(n+1)h＝d nM . Step 3, in order to estimate the system state and improve the estimation accuracy, consider the following T-S fuzzy proportional integral estimator as a state estimation scheme Wherein, the Is the state vector of the estimator and, Is the output of the estimator, and L j and Q j are the estimator gains to be designed. For ease of description, μ i (t) and μ j (t k h are abbreviated as μ i and μ j (t k h in the following derivation Step 4, combining the wind turbine system model and the proportional integral estimator to obtain a closed loop estimation error system, and designing the estimation error system by the estimator, wherein the method specifically comprises the following steps of: 1) Closed loop estimation error system obtained by combining fuzzy wind turbine system and proportional integral estimator Wherein, the And is also provided with 2) To facilitate solving the estimator parameters, n=2 is first selected and the following lyapunov function is selected: Wherein the method comprises the steps of 3) The derivative of V (t) is calculated as follows: Wherein the method comprises the steps of 4) According to the Jensen inequality, the following inequality is satisfied for n=0, 1, 2: Wherein, the 5) Combining with the dynamic summation type event triggering mechanism, the following can be obtained: 6) In combination with the above formula, the following inequality can be obtained: 7) Further, according to the definition of ψ n (t), we get: 8) The new variables are defined as follows: 9) For an estimation error system, a relaxation matrix is constructed And then obtain: Wherein, the 10 According to the above formula, can be obtained Wherein, the 11 According to the following Introducing a free matrix Γ i satisfying the following relation: 12 Combining the above formula, it can be derived that: 13 By taking into account And xi ij -Γ i <0, can be obtained: 14 According to the above conditions, the following inequality can be obtained: This ensures that when ω (t) =0, The constant holds. 15 Formula (19) is set at [ 0], + -infinity), then there are: V(∞)-V(0)<∫ 0 ∞ (γ 2 ω T (t)ω(y)-z T (t)z(t))dt。 (20) 16 For the case where ω (t) +.0 and the initial state is zero, according to equation (20), it is possible to obtain Thus, the filter error system is progressively stable while meeting the H ∞ performance index γ. The equivalent conditions are as follows: Ξ ij -Γ i <0, ρ i Ξ ii -ρ i Γ i +Γ i <0(i=j), ρ j Ξ ij +ρ i Ξ ji -ρ j Γ i -ρ i Γ j +Γ i +Γ j ＜0(i＜j). (21) 17 A matrix is constructed as follows: 18 At this time, the matrix inside the condition (21) can be rewritten as: Wherein the method comprises the steps of 19 Defining two new variables X Mj ＝M 2 M j and X Lj ＝M 2 L j , substituting the formula (23) into the condition (21) can be obtained: Wherein the method comprises the steps of 20 Through the Yalmip toolbox of MATLAB), a proportional-integral estimator that ensures system stability and meets the given H ∞ performance can be obtained, and a state estimation is performed on the wind turbine system:

Description

Proportional-integral state estimation method of nonlinear wind turbine system based on dynamic summation type event triggering Technical Field The invention relates to a state estimation method based on a dynamic summation type event triggering mechanism, in particular to a proportional-integral state estimation method for a nonlinear wind turbine system. Background In recent years, the global energy pattern is being transformed into sustainable and renewable energy, and this trend has profoundly affected the development direction of the energy industry. Wind energy plays a vital role in coping with climate change and reducing greenhouse gas emissions as an important representative of renewable energy sources. The accelerated transformation of fossil fuel to clean energy promotes the deep research of wind turbine systems in the aspects of optimizing design, improving efficiency, reducing cost and the like. With the continuous improvement of the networking degree of the wind turbine system, the introduction of a communication network also brings new challenges to the field. Along with the continuous expansion of the scale of the wind turbine system, the channel load is obviously increased due to the competition of a large number of devices to limited communication resources, so that the problems of network congestion and the like are caused. In order to solve the problem, researchers put forward a communication strategy based on event triggering, effectively reduce the use of communication network resources, greatly relieve the communication network pressure and obtain a series of important research results. Compared with the traditional event triggering mechanism based on the current information, the invention creatively provides a dynamic summation type event triggering mechanism based on historical sampling data. Meanwhile, internal variables are introduced into the triggering conditions, so that the triggering interval is further prolonged, and the utilization efficiency of network resources is remarkably improved. In addition, the invention designs a dynamic trigger threshold, which can adaptively adjust the trigger frequency according to the dynamic evolution of the system, and compared with a trigger mechanism which only depends on a static threshold, the flexibility of the trigger mechanism is greatly improved. In addition, in actual engineering, the performance of a control system is affected by a plurality of system states which are difficult to directly measure, and the proportional integral estimator is adopted by the invention, so that the accuracy of state estimation is remarkably enhanced by introducing integral terms compared with the traditional proportional estimator. Disclosure of Invention Based on the analysis, the invention designs a proportional-integral state estimation method based on dynamic summation type event triggering for effectively solving the problem that redundant data transmission and system state parts are not measurable. The specific technical scheme of the invention is as follows, a proportional integral state estimation method of a nonlinear wind turbine system based on dynamic summation type event triggering comprises the following steps: Based on the characteristics of the wind turbine system, establishing a state space model of the wind turbine system; The dynamic summation-based event triggering mechanism is designed to reduce redundant triggering of signals in data transmission and save network resources; Designing a proportional integral estimator, combining a wind turbine system model to obtain a closed-loop estimation error system, and analyzing the stability of the closed-loop estimation error system; And solving parameters of an estimator by means of the Liapunov theory and a linear matrix inequality method, and carrying out state estimation by adopting the estimator and carrying out state estimation on a wind turbine system. Technical proposal A proportional-integral state estimation method of a wind turbine system based on dynamic summation type event triggering is realized by the following steps in sequence: the method for estimating the proportional-integral state of a nonlinear wind turbine system based on dynamic summation type event triggering as set forth in claim 1, wherein the established closed loop T-S fuzzy system is as follows: Wherein, the AndRepresenting the system output and the performance output, respectively, v i (t) is a normalized membership function satisfying the following condition The method for estimating the proportional-integral state of a nonlinear wind turbine system based on dynamic summation type event triggering as set forth in claim 1, wherein the designed dynamic summation type event triggering is as follows: Wherein, the The error between the current time and the last time may be expressed as epsilon n(t)＝y(tkh-nh)-y(lk h-nh), n=0, 1. Phi (t) is the internal variable satisfiesH is the sampling period, l kh＝tkh+lh.Φn >0