Wind turbine power smoothing method based on hybrid weighted filtering and active disturbance rejection variable pitch
By using adaptive hybrid weighted filtering and active disturbance rejection pitch control technology, the problem of power output fluctuation of wind turbines under complex turbulent conditions was solved, achieving smooth power output and coordinated dynamic response of wind turbines, and improving the robustness and stability of the system.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HOHAI UNIV
- Filing Date
- 2026-03-13
- Publication Date
- 2026-06-26
AI Technical Summary
Existing technologies struggle to effectively smooth the output power of wind turbines under complex turbulent conditions and sudden drops in wind speed. Furthermore, filtering methods fail to achieve an adaptive balance between dynamic response and fluctuation suppression. Fixed controller parameters lack adaptability to time-varying conditions, making it difficult to achieve optimal system performance.
An adaptive hybrid weighted filtering algorithm based on the differential variance of power command and a bandwidth self-tuning active disturbance rejection control technology based on Lyapunov stability are adopted. Combined with extended Kalman filtering and recursive mean filtering, a smooth power reference command is generated. By adaptively adjusting the bandwidth of the extended state observer, efficient and smooth control of the wind turbine output power is achieved.
It significantly improves the power smoothness and dynamic response speed of wind turbines under complex turbulent conditions, enhances the robustness and stability of the system, and ensures high-quality grid connection of wind power generation.
Smart Images
Figure CN121828089B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of wind power generation control technology, and in particular to a wind turbine power smoothing method based on hybrid weighted filtering and active disturbance rejection pitch control. Background Technology
[0002] With the large-scale development and high grid connection rate of wind power generation, large wind turbine units have become a new type of power supply unit with active power support capabilities. When wind turbine units operate at or above rated wind speeds, they face complex operating conditions such as high-frequency turbulence and sudden wind speed drops, leading to significant power quality issues such as power oscillations and rapid drops in output power. Effective suppression and smoothing of these problems directly affect the service life of wind turbine units and the stable operation of the power grid, and also create an urgent need for the effectiveness and reliability of "power smoothing control technology adapted to complex turbulent conditions."
[0003] In the research of wind turbine power smoothing control, existing technologies mainly revolve around two aspects: power reference command generation and pitch control execution. In the command generation stage, model-based filtering methods such as the extended Kalman filter have good dynamic tracking performance, but they tend to follow power drops under extreme conditions and lack an effective smoothing mechanism. Time-domain filtering methods such as the recursive mean filter, while possessing strong smoothing characteristics, suffer from inherent phase lag that sacrifices the system's dynamic response speed, affecting wind energy capture efficiency. A single filtering architecture lacks a mechanism for adaptive adjustment based on operating conditions, making it difficult to achieve a dynamic balance between fast tracking and effective smoothing.
[0004] In the execution control stage, traditional linear active disturbance rejection control (ADRC) estimates and compensates for the total system disturbance by extending the state observer, exhibiting superior robustness compared to traditional PI controllers. However, the observer bandwidth of traditional ADRC is typically a fixed parameter. A fixed bandwidth parameter cannot maintain optimal performance under complex turbulent conditions: when facing large dynamic processes such as sudden changes in wind speed, a low bandwidth configuration leads to slow response and insufficient disturbance estimation; while when dealing with high-frequency turbulence, a high bandwidth configuration easily amplifies measurement noise, causing actuator oscillations.
[0005] Currently, existing technologies for power smoothing control of wind turbines in medium-to-high wind speed areas have the following shortcomings: First, a collaborative control architecture adapted to complex turbulence and sudden wind speed drops has not been established; second, the filtering method has failed to achieve an adaptive balance between dynamic response and fluctuation suppression; and third, the controller parameters are fixed and lack the ability to adapt to time-varying conditions, making it difficult to achieve optimal system-level performance.
[0006] Therefore, a power smoothing control method that can adapt to complex turbulent conditions and effectively coordinate power command smoothing and pitch control is needed to achieve global optimization is required. Summary of the Invention
[0007] This invention addresses the technical problem of power output fluctuations in wind turbines under complex turbulent conditions by providing a power smoothing method based on hybrid weighted filtering and active disturbance rejection (ADRP) pitch control. Building upon traditional single-filter and fixed-parameter control methods, this method introduces adaptive hybrid weighted filtering based on the differential variance of power commands and bandwidth self-tuning ADRP control based on Lyapunov stability, achieving efficient and smooth control of wind turbine output power. Through the organic combination of front-end adaptive filtering and back-end ADRP control, this invention not only significantly improves the smoothness and dynamic response speed of power output but also enhances the robustness and stability of the system under complex turbulent conditions, providing a solution for high-quality grid connection of wind power generation.
[0008] The inventive concept of this invention is as follows: First, an adaptive hybrid weighted filtering algorithm based on the differential variance of the power command is designed: the overshoot is obtained by calculating the instantaneous rate of change of the power command and comparing it with a preset threshold; the overshoot is smoothly mapped to the weight coefficients of the improved mean filter using an exponential function, and the outputs of the extended Kalman filter and the recursive mean filter are dynamically fused to generate a smooth power reference command; Second, an improved linear active disturbance rejection pitch controller based on Lyapunov stability is designed: a Lyapunov function containing state estimation error and bandwidth estimation error is constructed, and the bandwidth of the extended state observer is dynamically adjusted by designing an adaptive law to ensure the global stability of the system and improve the disturbance estimation accuracy.
[0009] To achieve the aforementioned objectives, the present invention employs the following technical solution: a wind turbine power smoothing method based on hybrid weighted filtering and self-disturbance rejection pitch control, comprising the following steps:
[0010] Step S1: Establish a nonlinear mathematical model of the permanent magnet direct-drive wind power generation system;
[0011] Step S2: An adaptive hybrid weighted filtering algorithm based on the differential variance of power command is adopted to generate a smooth power reference command: the instantaneous rate of change of the power command is calculated in real time, the overshoot is obtained by comparing it with a preset threshold, and the overshoot is mapped to the weight coefficient of the improved mean filter using an exponential function. The output of the extended Kalman filter and the output of the improved recursive mean filter are weighted and fused.
[0012] Step S3: Based on the smoothed power reference command obtained in step S2, and combined with the mapping relationship between rotor kinetic energy and power, calculate the corresponding smoothed reference speed command;
[0013] Step S4: A linear active disturbance rejection pitch control algorithm based on Lyapunov stability is adopted to perform pitch tracking control. The deviation between the smooth reference speed command and the actual speed is used as the controller input. A Lyapunov function containing state estimation error and bandwidth estimation error is constructed. A bandwidth adaptive law is designed to dynamically adjust the bandwidth of the extended state observer. The total system disturbance estimated by the observer is used for feedforward compensation to generate the reference pitch angle.
[0014] Step S5: Combine the generation of the smooth reference speed command described in step S3 with the linear active disturbance rejection pitch control described in step S4. Through the synergy of front-end command optimization and back-end robust control, the comprehensive power smoothing of the wind turbine under complex turbulent conditions is achieved.
[0015] Furthermore, in step S1, the mechanical power captured by the wind turbine in the permanent magnet direct-drive wind power generation system P w and mechanical torque T w It can be represented as:
[0016] (1)
[0017] in, air density, The radius of the wind turbine blades. For wind speed, The wind energy utilization coefficient, For the tip speed ratio, The pitch angle is the propeller angle. The influence coefficient of wind speed change rate. The rotor angular velocity, t For time, d It is a differential operator;
[0018] Wind energy utilization coefficient It is a key parameter reflecting the aerodynamic conversion efficiency of a wind turbine, and its value depends on the tip speed ratio. and propeller pitch angle The nonlinear coupling relationship can be expressed as:
[0019] (2)
[0020] in, As an intermediate variable, e It is the Euler number.
[0021] Furthermore, in step S2, the instantaneous rate of change of the power command at adjacent sampling times is calculated. :
[0022] (3)
[0023] In equation (3), This refers to the electromagnetic power command at the current moment. The electromagnetic power command from the previous moment. The sampling period is This is the rated power.
[0024] The instantaneous rate of change Compared with the preset power change rate threshold Compare and calculate overshoot. :
[0025] (4)
[0026] The overshoot was measured using an exponential function. Smooth mapping to the weight interval, and calculate the real-time weight coefficients of the recursive mean filter. for:
[0027] (5)
[0028] In equation (5), This is the upper limit of the weight. To control the sensitivity coefficient.
[0029] According to the real-time weighting coefficient For the extended Kalman filter output With recursive mean filter output Weighted fusion is performed to generate the final smoothed power reference value. for:
[0030] (6)
[0031] in, The weighting coefficients are as described above. .
[0032] Furthermore, in step S3, based on the smoothing power reference value... Smooth reference speed obtained by combining rotor kinetic energy control for:
[0033] (7);
[0034] in, Let be the moment of inertia.
[0035] Furthermore, in step S4, the bandwidth self-tuning mechanism based on Lyapunov stability is implemented as follows:
[0036] a) Define the state estimation error and bandwidth estimation error as:
[0037] (8);
[0038] in, This represents the change in rotational speed. This is the differential change in rotational speed. This represents the total disturbance of a permanent magnet direct-drive wind power generation system. , , They are respectively , , The observed estimates, For the observer bandwidth, For ideal bandwidth, For bandwidth estimation error, , They are respectively , , and , , The deviation between them.
[0039] b) Construct a Lyapunov function V that includes the state estimation error and the bandwidth estimation error:
[0040] (9)
[0041] In equation (9), For adaptive rate and .
[0042] c) Design a bandwidth adaptive law and differentiate it with respect to the Lyapunov function V:
[0043] (10)
[0044] In equation (10), The derivative of V, for The derivative, for The derivative, for The derivative, for The derivative of .
[0045] To meet Choose the adaptive law to cancel out the remainder:
[0046] (11)
[0047] As can be seen from the observation error, All Higher-order terms, if the observer converges, satisfy the following: At this point, you can select Thus, the observer bandwidth is obtained. for:
[0048] (12)
[0049] In equation (12), This is the initial value for the observer bandwidth.
[0050] d) Based on the control objective The control law is designed as follows:
[0051] (13)
[0052] In equation (13), for The derivative, It is a constant. For proportional gain, This is the integral gain.
[0053] Meanwhile, the present invention proposes an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed, it implements the steps of the method described in the present invention.
[0054] Furthermore, the present invention proposes a computer-readable storage medium having a computer program stored thereon, the computer program being configured to implement the steps of the method described in the present invention when invoked by a processor.
[0055] Finally, the present invention proposes a computer program product, including a computer program / instructions, characterized in that the computer program / instructions, when executed by a processor, implement the steps of the method described in the present invention.
[0056] Compared with the prior art, the beneficial effects of the present invention are as follows:
[0057] 1. The method of this invention is specifically designed to address the fluctuation problem of wind turbine output power under high wind speed turbulence and sudden wind speed drop turbulence conditions. It generates a smooth power command through an adaptive hybrid weighted filtering algorithm and achieves robust tracking control by combining it with a linear active disturbance rejection pitch controller. The method dynamically fuses the outputs of extended Kalman filtering and improved mean filtering through an adaptive weight allocation mechanism based on the differential variance of the power command to balance the contradiction between power smoothing and dynamic response. Through a bandwidth self-tuning mechanism based on Lyapunov stability, the bandwidth of the extended state observer in the active disturbance rejection controller is adjusted in real time to enhance the system's adaptability to time-varying disturbances and its global stability.
[0058] 2. The wind turbine power smoothing control method of the present invention achieves full-process optimization from power command generation to pitch execution by constructing a control architecture that coordinates the adaptive hybrid weighted filtering algorithm and the linear active disturbance rejection pitch controller.
[0059] 3. This invention effectively solves the contradiction between dynamic response and fluctuation suppression in a single filtering strategy, overcomes the insufficient adaptability of a fixed bandwidth controller under time-varying conditions, and significantly improves the power smoothing effect and operational reliability of the wind turbine under complex turbulence. Attached Figure Description
[0060] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.
[0061] Figure 1 This is a flowchart of the method in Embodiment 1 of the present invention.
[0062] Figure 2 This is a block diagram of the adaptive weighted hybrid filter rotor kinetic energy control in Embodiment 1 of the present invention.
[0063] Figure 3 This is a block diagram of the bandwidth self-tuning linear active disturbance rejection pitch control in Example 1.
[0064] Figure 4 The figure shows the simulation results of high wind speed turbulence in Embodiment 2 of the present invention.
[0065] Figure 5 The figure shows the simulation results of the turbulent working condition with sudden wind speed drop in Embodiment 3 of the present invention. Detailed Implementation
[0066] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. Of course, the specific embodiments described herein are merely illustrative and not intended to limit the invention.
[0067] Example 1: See Figure 1 This embodiment provides a technical solution: a wind turbine power smoothing method based on hybrid weighted filtering and self-disturbance rejection pitch control, comprising the following steps:
[0068] Step S1: Establish a nonlinear mathematical model of the permanent magnet direct-drive wind power generation system;
[0069] Step S2: An adaptive hybrid weighted filtering algorithm based on the differential variance of power command is adopted to generate a smooth power reference command: the instantaneous rate of change of the power command is calculated in real time, the overshoot is obtained by comparing it with a preset threshold, and the overshoot is mapped to the weight coefficient of the improved mean filter using an exponential function. The output of the extended Kalman filter and the output of the improved recursive mean filter are weighted and fused.
[0070] Step S3: Based on the smoothed power reference command obtained in step S2, and combined with the mapping relationship between rotor kinetic energy and power, calculate the corresponding smoothed reference speed command;
[0071] Step S4: A linear active disturbance rejection pitch control algorithm based on Lyapunov stability is adopted to perform pitch tracking control. The deviation between the smooth reference speed command and the actual speed is used as the controller input. A Lyapunov function containing state estimation error and bandwidth estimation error is constructed. A bandwidth adaptive law is designed to dynamically adjust the bandwidth of the extended state observer. The total system disturbance estimated by the observer is used for feedforward compensation to generate the reference pitch angle.
[0072] Step S5: Combine the generation of the smooth reference speed command described in step S3 with the linear active disturbance rejection pitch control described in step S4. Through the synergy of front-end command optimization and back-end robust control, the comprehensive power smoothing of the wind turbine under complex turbulent conditions is achieved.
[0073] Furthermore, in step S1, the mechanical power captured by the wind turbine in the permanent magnet direct-drive wind power generation system P w and mechanical torque T w It can be represented as:
[0074] (1)
[0075] in, air density, The radius of the wind turbine blades. For wind speed, The wind energy utilization coefficient, For the tip speed ratio, The pitch angle is the propeller angle. The influence coefficient of wind speed change rate. The rotor angular velocity, t For time, d It is a differential operator.
[0076] Wind energy utilization coefficient It is a key parameter reflecting the aerodynamic conversion efficiency of a wind turbine, and its value depends on the tip speed ratio. and propeller pitch angle The nonlinear coupling relationship can be expressed as:
[0077] (2)
[0078] in, As an intermediate variable, e It is the Euler number.
[0079] Furthermore, in step S2, the instantaneous rate of change of the power command at adjacent sampling times is calculated. :
[0080] (3)
[0081] in, This refers to the electromagnetic power command at the current moment. The electromagnetic power command from the previous moment. The sampling period is This is the rated power.
[0082] The instantaneous rate of change Compared with the preset power change rate threshold Compare and calculate overshoot. :
[0083] (4)
[0084] The overshoot was measured using an exponential function. Smooth mapping to the weight interval, and calculate the real-time weight coefficients of the recursive mean filter. for:
[0085] (5)
[0086] in, This is the upper limit of the weight. To control the sensitivity coefficient.
[0087] According to the real-time weighting coefficient For the output of the extended Kalman filter (EKF) With Recursive Mean Filter (RMF) output Weighted fusion is performed to generate the final smoothed power reference value. for:
[0088] (6)
[0089] in, The weighting coefficients are as described above. .
[0090] Furthermore, in step S3, based on the smoothing power reference value... Smooth reference speed obtained by combining rotor kinetic energy control for:
[0091] (7);
[0092] in, Let be the moment of inertia.
[0093] The adaptive weighted hybrid filter rotor kinetic energy control block diagram formed by steps S2 and S3 is as follows: Figure 2 As shown.
[0094] Furthermore, in step S4, the bandwidth self-tuning mechanism based on Lyapunov stability is implemented as follows:
[0095] a) Define the state estimation error and bandwidth estimation error as:
[0096] (8);
[0097] in, This represents the change in rotational speed. This is the differential change in rotational speed. This represents the total disturbance of a permanent magnet direct-drive wind power generation system. , , They are respectively , , The observed estimates, For the observer bandwidth, For ideal bandwidth, For bandwidth estimation error, , They are respectively , , and , , The deviation between them.
[0098] b) Construct a Lyapunov function V that includes the state estimation error and the bandwidth estimation error:
[0099] (9)
[0100] in, For adaptive rate and .
[0101] c) Design a bandwidth adaptive law and differentiate it with respect to the Lyapunov function V:
[0102] (10)
[0103] in, The derivative of V, for The derivative, for The derivative, for The derivative, for The derivative of .
[0104] To meet Choose the adaptive law to cancel out the remainder:
[0105] (11)
[0106] As can be seen from the observation error, All Higher-order terms, if the observer converges, satisfy the following: At this point, you can select Thus, the observer bandwidth is obtained. for:
[0107] (12)
[0108] in, This is the initial value for the observer bandwidth.
[0109] d) Based on the control objective The control law is designed as follows:
[0110] (13)
[0111] in, for The derivative, It is a constant. For proportional gain, This is the integral gain.
[0112] The control block diagram of the bandwidth self-tuning linear active disturbance rejection pitch (ILADRC) formed in step S4 is as follows: Figure 3 As shown.
[0113] Example 2
[0114] Based on the control strategy adopted in this invention, a wind turbine power output simulation is performed, using the following wind turbine simulation parameters: , , , , , Rated wind speed 10.5 m / s, rated power 6 MW. Adopting... Figure 4Simulation of the high-wind-speed turbulence shown in (a) yields the following results: Figure 4 (b) power output Figure 4 The rotor speed shown in (c) and Figure 4 (d) shows the pitch angle. Figure 4 (e) represents the adaptive weighting coefficients during the simulation. Figure 4 (f) represents the observer bandwidth during the simulation process.
[0115] Depend on Figure 4 As can be seen, under high wind speed turbulence, the present invention generates smooth power commands through adaptive hybrid weighted filtering, and combined with bandwidth self-tuning linear active disturbance rejection pitch controller, it achieves significant smoothing of power output and stable speed tracking. At the same time, the observer bandwidth and filter weights can be dynamically and adaptively adjusted according to the operating conditions. Figure 4 (b) The power output of the three methods under high wind speed turbulence is compared: the traditional PI control has significant fluctuations and poor output stability; the extended Kalman filter combined with fixed bandwidth active disturbance rejection control (EKF-ADRC) has improved the situation, but there are still some fluctuations; while the proposed method has the most stable output power and the least fluctuations.
[0116] Example 3: The simulated fan parameters are the same as those in Example 2. Figure 5 Simulation of the turbulent condition with a sudden drop in wind speed shown in (a) yields the following results: Figure 5 (b) power output Figure 5 The rotor speed shown in (c) and Figure 5 (d) shows the pitch angle. Figure 5 (e) represents the adaptive weighting coefficient under turbulent conditions with a sudden drop in wind speed. Figure 5 (f) represents the change in observer bandwidth during the simulation process.
[0117] Depend on Figure 5 (b) It can be seen that under the turbulent condition of sudden wind speed drop, the control scheme adopted by the present invention can maintain the rated power output, and the power fluctuation is significantly reduced compared with other control strategies when the wind speed drops to below the rated wind speed. Therefore, the present invention can ensure the smoothness of power output and the stable operation of the system.
[0118] Example 4: This example proposes an electronic system, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method steps of the present invention.
[0119] Example 5: This example proposes a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the steps of the method described in this invention, which will not be repeated here.
[0120] Example 6: This example proposes a computer program product, including a computer program / instructions. When the computer program / instructions are executed by a processor, they implement the steps of the method described in this invention, which will not be repeated here.
[0121] It should be noted that the processing flow of embodiments 4-6 corresponds to the specific steps of the method provided in Embodiment 1 of the present invention, and has the corresponding functions and beneficial effects of the method. Technical details not described in detail in this embodiment can be found in the method provided in Embodiment 1 of the present invention.
[0122] The program code used to implement the methods of this application may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.
[0123] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A wind turbine power smoothing method based on hybrid weighted filtering and active disturbance rejection pitch control, characterized in that, Includes the following steps: Step S1: Establish a nonlinear mathematical model of the permanent magnet direct-drive wind power generation system; Step S2: An adaptive hybrid weighted filtering algorithm based on the differential variance of power command is adopted to generate a smooth power reference command: the instantaneous rate of change of the power command is calculated in real time, the overshoot is obtained by comparing it with a preset threshold, and the overshoot is mapped to the weight coefficient of the improved mean filter using an exponential function. The output of the extended Kalman filter and the output of the improved recursive mean filter are weighted and fused. Step S3: Based on the smoothed power reference command obtained in step S2, and combined with the mapping relationship between rotor kinetic energy and power, calculate the corresponding smoothed reference speed command; Step S4: A linear active disturbance rejection pitch control algorithm based on Lyapunov stability is adopted to perform pitch tracking control. The deviation between the smooth reference speed command and the actual speed is used as the controller input. A Lyapunov function containing state estimation error and bandwidth estimation error is constructed. A bandwidth adaptive law is designed to dynamically adjust the bandwidth of the extended state observer. The total system disturbance estimated by the observer is used for feedforward compensation to generate the reference pitch angle. In step S4, the bandwidth self-tuning mechanism based on Lyapunov stability is implemented as follows: a) Define the state estimation error and bandwidth estimation error as: （8）； in, This represents the change in rotational speed. This is the differential change in rotational speed. This represents the total disturbance of a permanent magnet direct-drive wind power generation system. , , They are respectively , , The observed estimates, For the observer bandwidth, For ideal bandwidth, For bandwidth estimation error, , They are respectively , , and , , Deviation between; b) Construct a Lyapunov function V that includes the state estimation error and bandwidth estimation error: （9）； in, For adaptive rate, and ; c) Design a bandwidth adaptive law and differentiate it with respect to the Lyapunov function V: （10）； in, The derivative of V, for The derivative, for The derivative, for The derivative, for The derivative; To meet Choose the adaptive law to cancel out the remainder: （11）； From the observation error, we know that All Higher-order terms, if the observer converges, satisfy the following: Select The observer bandwidth is obtained. for: （12）； in, This is the initial value for the observer bandwidth; d) Based on the control objective The control law is designed as follows: (13)； in, In order to control the target, for The derivative, It is a constant. For proportional gain, For integral gain; Step S5: Combine the generation of the smooth reference speed command described in step S3 with the linear active disturbance rejection pitch control described in step S4. Through the synergy of front-end command optimization and back-end robust control, the comprehensive power smoothing of the wind turbine under complex turbulent conditions is achieved.
2. The wind turbine power smoothing method based on hybrid weighted filtering and active disturbance rejection pitch control according to claim 1, characterized in that: In step S1, the mechanical power captured by the wind turbine in the permanent magnet direct-drive wind power generation system P w and mechanical torque T w Represented as: (1)； in, P w The mechanical power captured by the wind turbine. T w The mechanical torque captured by the wind turbine, air density, The radius of the wind turbine blades. For wind speed, The wind energy utilization coefficient, For the tip speed ratio, The pitch angle is the propeller angle. The influence coefficient of wind speed change rate. The rotor angular velocity, t For time, d It is a differential operator; Wind energy utilization coefficient The aerodynamic efficiency of a fan is reflected in its value, which depends on the tip speed ratio. and propeller pitch angle The nonlinear coupling relationship is expressed as: (2)； in, As an intermediate variable, e It is the Euler number.
3. The wind turbine power smoothing method based on hybrid weighted filtering and active disturbance rejection pitch control according to claim 2, characterized in that: In step S2, the instantaneous rate of change of the power command at adjacent sampling times is calculated. : (3)； in, The instantaneous rate of change This refers to the electromagnetic power command at the current moment. The electromagnetic power command from the previous moment. The sampling period is Rated power; The instantaneous rate of change Compared with the preset power change rate threshold Compare and calculate the overshoot. : (4)； The overshoot was measured using an exponential function. Smooth mapping to the weight interval, and calculate the real-time weight coefficients of the improved recursive mean filter. for: (5)； in, This is the upper limit of the weight. To control the sensitivity coefficient; According to the real-time weighting coefficient For the extended Kalman filter output With improved recursive mean filter output Weighted fusion is performed to generate the final smoothed power reference value. for: （6）； in, As a smooth power reference value, The weighting coefficient is the one in formula (5). .
4. The wind turbine power smoothing method based on hybrid weighted filtering and self-disturbance rejection pitch control according to claim 3, characterized in that: In step S3, based on the smoothed power reference value Smooth reference speed obtained by combining rotor kinetic energy control for: （7）； in, Let be the moment of inertia.
5. An electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the computer program is executed, it implements the steps of the method as described in any one of claims 1 to 4.
6. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program is configured to implement the steps of the method according to any one of claims 1 to 4 when invoked by a processor.
7. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method according to any one of claims 1 to 4.