Wind turbine variable pitch active disturbance rejection controller design method, system, medium, and computing device

Abstract

The application discloses a fan variable-pitch active disturbance rejection controller design method, system, medium and computing device, and comprises the following steps: acquiring the variable-pitch actuator dynamic characteristics of a fan and the input and output of a controlled system, and using the same to construct an extended state observer; compensating the input quantity of the extended state observer based on the variable-pitch actuator dynamic characteristics, after compensation, the system dynamic characteristics observed in the extended state observer are accelerated, a suitable control law is redesigned to accelerate the dynamic response speed of the closed-loop system; the bandwidth method is used to simplify the parameters, that is, the controller bandwidth ω c and the observer bandwidth ω o are used to calculate the corresponding parameters; gain scheduling is completed by adjusting the size of ω c and b0; the active disturbance rejection control algorithm is modified after considering the additional control quantity, that is, the active disturbance rejection controller design is completed. The application increases the variable-pitch actuator dynamic characteristic compensation on the basis of the active disturbance rejection control algorithm, and can improve the hysteresis of the variable-pitch system and improve the control precision of the fan operating parameters.

Description

Technical Field

[0001] This invention relates to the technical field of wind turbine pitch control, and in particular to a design method, system, storage medium, and computing device for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation of the pitch actuator. It adds dynamic characteristic compensation of the pitch actuator to the active disturbance rejection control algorithm to improve the hysteresis of the pitch system, thereby reducing the thrust on the impeller and improving the control accuracy of the wind turbine operating parameters. Background Technology

[0002] As wind turbines continue to grow larger, the diameter and weight of the impellers are constantly increasing. The system lag caused by the large moment of inertia increases the difficulty of system control. For closed-loop feedback control, a slow system response leads to slow actuator action, resulting in the inability to act in time when wind speed, wind direction, and wind shear change drastically. This causes drastic changes in aerodynamic thrust, which in turn leads to significant changes in the load on key wind turbine components, affecting the safe and stable operation of the wind turbine.

[0003] For pitch control loops, the dynamic characteristics of the actuator are one of the main sources of hysteresis. The PID control algorithm used in mainstream wind turbine pitch control systems is a classic feedback control algorithm, which cannot compensate for known inertia and hysteresis components within the control loop, thus its control performance needs further improvement. To address these issues, some scholars have attempted to use predictive control to solve pitch problems. However, due to the complex modeling of wind turbine systems, strong model uncertainty, and the highly nonlinear characteristics of dynamics changing with operating conditions, predictive control design is complex and real-time calculations consume significant computational resources. Overall, the implementation and parameter tuning of most advanced control algorithms are complex, requiring numerous parameters that are difficult for field engineers to master, and long-term maintenance is challenging, making it difficult to change the current situation where feedback controllers still primarily rely on PID. To resolve these contradictions, introducing control technologies more suitable for engineering practice into wind turbine pitch control system design is of great significance. Such control technologies should have the following characteristics: simple structure, easy on-site implementation, convenient for engineers to learn and master, and applicable to wind turbine pitch control systems. This is where active disturbance rejection control (ADRC) comes into focus.

[0004] Active Disturbance Rejection Control (ADRC) is most importantly structured by the Extended State Observer (ESO). Its core idea is to reconstruct the system's state by observing the system's input and output using the ESO, and then treating both internal and external disturbances as an extended state, designing a control law to compensate for them. Therefore, ADRC possesses excellent disturbance rejection capabilities. Since ADRC is developed based on PID control, its structure is simple. It directly observes the system's input and output, using error feedback for compensation, without requiring a mathematical model of the system. Therefore, it has a wide range of applications, and practical experience shows that it has demonstrated great application potential in control systems across various fields.

[0005] In summary, ADRC has three major advantages: 1) It is based on the classic PID controller, and its structure is similar to that of PID, making the system simple. In particular, linear ADRC has few tuning parameters, making it easy for engineers to master; 2) It has strong anti-interference capabilities to cope with the constantly changing external conditions of the wind turbine; 3) When reconstructing the system state using input and output, ADRC can compensate for known parts of the model, such as the dynamic characteristics of the pitch actuator, thereby making the observed system a faster system, accelerating the closed-loop response speed, improving the control accuracy of wind turbine operating parameters, and reducing wind turbine load fluctuations. This is of great significance for the safe and stable operation of wind turbines and for reducing design costs. Summary of the Invention

[0006] The primary objective of this invention is to overcome the shortcomings and deficiencies of existing technologies and provide a design method for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation. This method compensates for the dynamic characteristics of the pitch actuator by applying an advanced active disturbance rejection control algorithm with good disturbance rejection performance that is easy for engineers to implement, thereby enhancing the closed-loop response of the system, improving the control accuracy of wind turbine operating parameters, and reducing load fluctuations in key wind turbine components. This is of great significance for the safe and stable operation of wind turbines and for cost reduction in design.

[0007] The second objective of this invention is to provide a design system for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation.

[0008] A third objective of this invention is to provide a storage medium.

[0009] A fourth objective of this invention is to provide a computing device.

[0010] The first objective of this invention is achieved through the following technical solution: a design method for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation, comprising the following steps:

[0011] 1) Obtain the dynamic characteristics of the pitch actuator of the wind turbine under design, and obtain the input and output of the controlled system to construct an extended state observer;

[0012] 2) Based on the obtained dynamic characteristics of the pitch actuator, the input of the extended state observer is compensated. After compensation, the dynamic characteristics of the system observed in the extended state observer are accelerated. Then, a suitable linear feedback control law is designed to accelerate the dynamic response speed of the closed-loop system.

[0013] 3) The parameters are simplified using the bandwidth method, that is, by utilizing the controller bandwidth ω. c and observer bandwidth ω o Calculate the corresponding parameters;

[0014] 4) By adjusting ω c Gain scheduling is accomplished by adjusting the size of b0;

[0015] 5) After considering the additional control variables, the active disturbance rejection control algorithm is modified, that is, the design of the active disturbance rejection controller is completed, and then it can be used for the pitch system control of the wind turbine to improve the lag of the pitch system.

[0016] Furthermore, in step 1), the dynamic characteristics of the wind turbine's pitch actuator are represented by the following second-order transfer function:

[0017]

[0018] In the formula, s represents the Laplace operator, and G pitch (s) represents the second-order transfer function of the dynamic characteristics of the pitch actuator, ω represents the center frequency of the characteristic polynomial in the denominator of the second-order transfer function, and ξ represents the damping ratio.

[0019] Furthermore, in step 1), there are two methods to obtain the dynamic characteristics of the actuator. The first method is based on workshop step test data. The specific test method is as follows: determine the software speed limit of the pitch actuator according to the wind turbine model, including the pitch recovery rate limit v. up and the limit value of the propeller opening rate v down Then, a ramp reference input test is performed, and the host computer provides a rate of v. up The paddle retraction command or the rate is v down Record the oar-starting command signal r. test and pitch angle feedback value θ test Using r test and θ test The first method can identify the dynamic characteristics of the pitch actuator; the second method is to export the pitch angle command value r from the recorded data of the wind turbine based on the on-site operating data of the wind turbine. operation and pitch angle feedback value θ operation The recorded data is at least 2 minutes long, using r operationand θ operation This allows for the identification of the dynamic characteristics of the pitch actuator.

[0020] Furthermore, in step 1), for the pitch control loop of the wind turbine, the input quantity of its controlled system, i.e., the control quantity, is the pitch angle command θ. demand The system's output, i.e., the controlled variable, is the generator speed ω of the wind turbine. g Based on the above input and output quantities, the various states of the system can be reconstructed using an extended state observer; if the designed active disturbance rejection controller has an order of n, then the system states are represented by z1 to z2. n The term z represents the internal and external disturbances of the system, as well as the extended state of the unmodeled dynamics. n+1 It means that z1~z n+1 The calculation method is as follows:

[0021]

[0022] In the formula, Represents the first derivative of the corresponding state variable, β1~β n+1 For the internal parameters of the extended state observer, y = ω g Where b0 is the control law parameter, representing the control gain; the above formula is for a continuous system. In a discrete system, we consider using the Euler method for discretization, and the discretized form of the formula is:

[0023]

[0024] In the formula, k represents the number of control period sequences, z n (k) represents the variable z n The value of the k-th periodic sequence, z n (k+1) represents the variable z. n The value of the (k+1)th periodic sequence, z n+1 (k) represents the variable z n+1 The value of the k-th periodic sequence, y(k+1) represents the value of the variable y in the (k+1)-th periodic sequence, θ demand (k) represents the variable θ demand The value of the k-th period sequence, dt represents the time interval of the control cycle, and the above formula enables the discrete calculation of the extended state observer in electronic devices.

[0025] Furthermore, in step 2), the input of the extended state observer is compensated based on the obtained dynamic characteristics of the pitch actuator, i.e., equation (3) becomes:

[0026]

[0027] In the formula, θ compensation θ represents the compensated pitch control quantity.compensation (k) represents the variable θ compensation The value of the k-th cycle sequence has the following relationship with the pitch command:

[0028] θ compensation (s)=G pitch (s)θ demand (s) (5)

[0029] In the formula, θ compensation (s) represents the variable θ compensation The value obtained after performing a Laplace transform on the s-domain, θ demand (s) represents the variable θ demand The value obtained after performing the Laplace transformation on the s domain, when substituted into equation (5) after equation (1), yields:

[0030]

[0031] Transforming the above equation into discrete form, based on the backward difference method, we get:

[0032]

[0033] In the formula, variable A = 1 / ω 2 , variable B = 2ξ / ω, θ compensation (k-1) represents the variable θ compensation The value of the (k-1)th periodic sequence, θ compensation (k-2) represents the variable θ compensation The value of the (k-2)th period sequence, θ demand (k-1) represents the variable θ demand The value of the (k-1)th period sequence.

[0034] Furthermore, in step 2), the control law is designed as follows:

[0035]

[0036] In the formula, Let i be the i-th parameter of the controller, i = 1 to n+1. The discrete form of the above equation is:

[0037]

[0038] This completes the design of the control law.

[0039] Furthermore, in step 3), the extended state observer parameters include β1 to β2. n+1 There are a total of n+1 parameters, and the controller parameters include... There are a total of n+1 parameters, which is not conducive to engineering use. The parameters are simplified by using the bandwidth method, that is, by utilizing the controller bandwidth ω. cand observer bandwidth ω o The corresponding parameters are calculated according to the following rules:

[0040] Regarding the controller parameters, we have:

[0041]

[0042] In the formula, Represents ω c The n+1-i power;

[0043] For the extended state observer parameters, we have:

[0044]

[0045] In the formula, Represents ω o power of i;

[0046] In this way, the parameters of the entire active disturbance rejection controller are simplified to three, namely ω c ω o And b0, in addition, ω c and ω o It can also be further simplified based on the correspondence; due to ω c This characterizes the speed of the dynamic response of the closed-loop system, while ω o The extended state observer's tracking speed for each state of the closed-loop system is characterized by its own response speed. Therefore, to ensure the system's robustness and good dynamic performance, ω... o =3~10ω c Thus, the parameters of the active disturbance rejection controller are simplified to two, namely ω. c b0 and b0 are the same number of parameters that need to be tuned for a conventional PI controller.

[0047] Furthermore, in step 4), let the aerodynamic torque experienced by the fan impeller be T. a The actual pitch angle of the wind turbine during operation is θ, due to the rated wind speed above... The gain of the pitch controller increases with increasing wind speed, therefore the gain of the designed pitch controller decreases with increasing wind speed. This is reflected in a conventional PI controller, where the proportional gain decreases with increasing wind speed. However, to ensure the dynamic performance of the system, the integral gain increases. For an active disturbance rejection controller, this can be achieved by adjusting ω. c Gain scheduling is accomplished using the value of b0, and the specific implementation method is as follows:

[0048] Let the rated wind speed be v 额 The cut-out wind speed is v 切 Then in v 额 and v 切Select m operating points for parameter tuning, corresponding to wind speed points v1 to v2. m For each wind speed point, parameter tuning is performed to obtain the active disturbance rejection control parameter set. These are the active disturbance rejection controller parameters obtained by tuning at the operating point with a wind speed of v1. These are the active disturbance rejection controller parameters obtained by tuning at the operating point with a wind speed of v2. That is, when the wind speed is v m The active disturbance rejection controller parameters were obtained by tuning at the operating point. Since the on-site wind speed measurement is not precise enough, directly using the measured wind speed for gain adjustment can easily introduce additional errors. Therefore, it is considered to utilize the parameters v1~v under steady-state conditions. m The corresponding propeller pitch angles θ1~θ m Gain scheduling is performed, and the specific scheduling method is as follows:

[0049]

[0050] This completes the design of gain scheduling.

[0051] Furthermore, in step 5), during the actual operation of the wind turbine, the pitch angle command is not only calculated by the feedback control algorithm, but also includes additional pitch angle control quantities introduced by feedforward or other control strategies. If the extended state observer introduces a pitch angle control command that includes additional pitch angle control quantities, it will cause steady-state errors in the active disturbance rejection control closed-loop calculation process. Therefore, it is necessary to consider removing the other additional commands from the pitch angle control command introduced into the extended state observer and only retain the control command calculated by the feedback control algorithm itself.

[0052] Let the additional pitch angle command be θ delta Then equation (5) is modified to:

[0053] θ compensation (s)=G pitch (s)(θ demand (s)-θ delta (s)) (13)

[0054] Where, θ delta (s) represents the additional pitch angle command in addition to the feedback control algorithm;

[0055] Thus, the design of the active disturbance rejection controller based on dynamic characteristic compensation of the pitch actuator was completed.

[0056] The second objective of this invention is achieved through the following technical solution: a wind turbine pitch active disturbance rejection controller design system based on dynamic characteristic compensation, used to implement the aforementioned wind turbine pitch active disturbance rejection controller design method based on dynamic characteristic compensation, comprising:

[0057] The data acquisition module is used to acquire the dynamic characteristics of the pitch actuator of the wind turbine under design, as well as the input and output of the controlled system, in order to construct an extended state observer.

[0058] The compensation module compensates the input of the extended state observer based on the obtained dynamic characteristics of the pitch actuator. After compensation, the dynamic characteristics of the system observed in the extended state observer are accelerated. Then, a suitable control law is designed to accelerate the dynamic response speed of the closed-loop system.

[0059] The simplified module uses the bandwidth method to simplify the parameters, that is, it utilizes the controller bandwidth ω. c and observer bandwidth ω o Calculate the corresponding parameters;

[0060] The gain scheduling module adjusts ω c Gain scheduling is accomplished by adjusting the size of b0;

[0061] The correction module is used to correct the active disturbance rejection control algorithm after considering additional control variables, thus completing the design of the active disturbance rejection controller.

[0062] The third objective of this invention is achieved through the following technical solution: a storage medium storing a program, which, when executed by a processor, implements the above-mentioned design method for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation.

[0063] The fourth objective of this invention is achieved through the following technical solution: a computing device, including a processor and a memory for storing processor-executable programs, wherein when the processor executes the program stored in the memory, it implements the above-mentioned design method of wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation.

[0064] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0065] 1. This invention designs a new active disturbance rejection controller, which combines the active disturbance rejection control algorithm with the characteristics of the wind turbine pitch control system. This can both accelerate the dynamic response speed of the system and improve the system's disturbance rejection capability.

[0066] 2. This invention specifically explains the implementation method and precautions of the algorithm in engineering, which facilitates its promotion and application.

[0067] 3. This invention can reduce the load fluctuation of the wind turbine caused by wind speed changes, thereby ensuring the safe and stable operation of the wind turbine, and is of great significance for reducing the cost of wind turbine design. Attached Figure Description

[0068] Figure 1 This is a diagram of the active disturbance rejection controller architecture based on dynamic characteristic compensation of the pitch actuator.

[0069] Figure 2 Flowchart for designing an active disturbance rejection controller.

[0070] Figure 3 This is a schematic diagram illustrating the control effect of the active disturbance rejection controller.

[0071] Figure 4 To design the system architecture diagram. Detailed Implementation

[0072] The present invention will be further described in detail below with reference to the embodiments and accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0073] Example 1

[0074] like Figures 1 to 3 As shown in the figure, this embodiment discloses a design method for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation, the specific details of which are as follows:

[0075] 1) Obtain the dynamic characteristics of the pitch actuator of the wind turbine under design, and obtain the input and output of the controlled system to construct an extended state observer.

[0076] Generally speaking, the dynamic characteristics of a wind turbine pitch actuator can be represented by the following second-order transfer function:

[0077]

[0078] In the formula, s represents the Laplace operator, and G pitch (s) represents the second-order transfer function of the dynamic characteristics of the pitch actuator, ω represents the center frequency of the characteristic polynomial in the denominator of the second-order transfer function, and ξ represents the damping ratio.

[0079] There are two methods to obtain the dynamic characteristics of the actuator. The first method is based on workshop step test data. The specific test method is as follows: determine the software speed limit of the pitch actuator according to the wind turbine model, including the pitch recovery rate limit v. up and the limit value of the propeller opening rate v down Then, a ramp reference input test is performed, and the host computer provides a rate of v. up The paddle retraction command or the rate is v down Record the oar-starting command signal r. test and pitch angle feedback value θ test Using r test and θ test The first method can identify the dynamic characteristics of the pitch actuator; the second method is to export the pitch angle command value r from the wind turbine's recorded data (sampling frequency above 50Hz) based on the wind turbine's on-site operating data. operation and pitch angle feedback value θ operationThe recorded data is at least 2 minutes long, using r operation and θ operation This allows for the identification of the dynamic characteristics of the pitch actuator.

[0080] For the pitch control loop of a wind turbine, the input of the controlled system, i.e., the control quantity, is the pitch angle command θ. demand The system's output, i.e., the controlled variable, is the generator speed ω of the wind turbine. g Based on the above input and output quantities, the various states of the system can be reconstructed using an extended state observer; if the designed active disturbance rejection controller has an order of n, then the system states are represented by z1 to z2. n The term z represents the internal and external disturbances of the system, as well as the extended state of the unmodeled dynamics. n+1 It means that z1~z n+1 The calculation method is as follows:

[0081]

[0082] In the formula, Represents the first derivative of the corresponding state variable, β1~β n+1 For the internal parameters of the extended state observer, y = ω g Where b0 is the control law parameter, representing the control gain; the above formula is for a continuous system. In a discrete system, we consider using the Euler method for discretization, and the discretized form of the formula is:

[0083]

[0084] In the formula, k represents the number of control period sequences, z n (k) represents the variable z n The value of the k-th periodic sequence, z n (k+1) represents the variable z. n The value of the (k+1)th periodic sequence, z n+1 (k) represents the variable z n+1 The value of the k-th periodic sequence, y(k+1) represents the value of the variable y in the (k+1)-th periodic sequence, θ demand (k) represents the variable θ demand The value of the k-th period sequence, dt represents the time interval of the control cycle, and the above formula enables the discrete calculation of the extended state observer in electronic devices.

[0085] 2) Based on the obtained dynamic characteristics of the pitch actuator, the input of the extended state observer is compensated, that is, equation (3) becomes:

[0086]

[0087] In the formula, θ compensationθ represents the compensated pitch control quantity. compensation (k) represents the variable θ compensation The value of the k-th cycle sequence has the following relationship with the pitch command:

[0088] θ compensation (s)=G pitch (s)θ demand (s) (5)

[0089] In the formula, θ compensation (s) represents the variable θ compensation The value obtained after performing a Laplace transform on the s-domain, θ demand (s) represents the variable θ demand The value obtained after performing the Laplace transformation on the s domain, when substituted into equation (5) after equation (1), yields:

[0090]

[0091] Transforming the above equation into discrete form, based on the backward difference method, we get:

[0092]

[0093] In the formula, variable A = 1 / ω 2 , variable B = 2ξ / ω, θ compensation (k-1) represents the variable θ compensation The value of the (k-1)th periodic sequence, θ compensation (k-2) represents the variable θ compensation The value of the (k-2)th period sequence, θ demand (k-1) represents the variable θ demand The value of the (k-1)th period sequence.

[0094] After compensation, the dynamic characteristics of the system observed in the extended state observer are accelerated, and a suitable linear feedback control law is designed to accelerate the dynamic response speed of the closed-loop system.

[0095] The control law is designed as follows:

[0096]

[0097] In the formula, Let i be the i-th parameter of the controller, i = 1 to n+1. The discrete form of the above equation is:

[0098]

[0099] This completes the design of the control law.

[0100] 3) Currently, there are many parameters to be tuned for the controller and the extended state observer. The parameters of the extended state observer include β1 to β2.n+1 There are a total of n+1 parameters, and the controller parameters include... There are a total of n+1 parameters, which is not conducive to engineering use. Therefore, we consider using the bandwidth method to simplify the parameters, that is, utilizing the controller bandwidth ω. c and observer bandwidth ω o The corresponding parameters are calculated according to the following rules:

[0101] Regarding the controller parameters, we have:

[0102]

[0103] In the formula, Represents ω c The n+1-i power;

[0104] For the extended state observer parameters, we have:

[0105]

[0106] In the formula, Represents ω o power of i;

[0107] In this way, the parameters of the entire active disturbance rejection controller are simplified to three, namely ω c ω o And b0, in addition, ω c and ω o It can also be further simplified based on the correspondence; due to ω c This characterizes the speed of the dynamic response of the closed-loop system, while ω o The value ω represents the tracking speed of the extended state observer for each state of the closed-loop system, and is related to its own response speed. Therefore, to ensure the robustness and good dynamic performance of the system, ω can generally be taken as ω. o =3~10ω c Thus, the parameters of the active disturbance rejection controller are simplified to two, namely ω. c b0 and b0 are the same number of parameters that need to be tuned for a conventional PI controller.

[0108] 4) Let the aerodynamic torque on the fan impeller be T. a The actual pitch angle of the wind turbine during operation is θ, due to the rated wind speed above... The gain of a pitch controller increases with increasing wind speed, therefore, the gain of such a controller often decreases with increasing wind speed. This is reflected in conventional PI controllers, where the proportional gain typically decreases with increasing wind speed. However, to ensure the system's dynamic performance, the integral gain increases. For active disturbance rejection controllers, this can be achieved by adjusting ω. c Gain scheduling is accomplished using the value of b0, and the specific implementation method is as follows:

[0109] Let the rated wind speed be v 额 The cut-out wind speed is v 切 Then in v 额 and v 切 Select m operating points for parameter tuning, corresponding to wind speed points v1 to v2. m For each wind speed point, parameter tuning is performed to obtain the active disturbance rejection control parameter set. These are the active disturbance rejection controller parameters obtained by tuning at the operating point with a wind speed of v1. These are the active disturbance rejection controller parameters obtained by tuning at the operating point with a wind speed of v2. That is, when the wind speed is v m The active disturbance rejection controller parameters were obtained by tuning at the operating point. Since the on-site wind speed measurement is not precise enough, directly using the measured wind speed for gain adjustment can easily introduce additional errors. Therefore, it is considered to utilize the parameters v1~v under steady-state conditions. m The corresponding propeller pitch angles θ1~θ m Gain scheduling is performed, and the specific scheduling method is as follows:

[0110]

[0111] This completes the design of gain scheduling.

[0112] 5) During the actual operation of the wind turbine, the pitch angle command is not only calculated by the feedback control algorithm, but also includes additional pitch angle control quantities introduced by feedforward or other control strategies. If the extended state observer introduces a pitch angle control command that includes these additional pitch angle control quantities, it will cause steady-state errors in the active disturbance rejection control closed-loop calculation. Therefore, it is necessary to consider removing the other additional commands from the pitch angle control command introduced into the extended state observer and only retain the control command calculated by the feedback control algorithm itself.

[0113] Let the additional pitch angle command be θ delta Then equation (5) is modified to:

[0114] θ compensation (s)=G pitch (s)(θ demand (s)-θ delta (s)) (13)

[0115] Where, θ delta (s) represents the additional pitch angle command in addition to the feedback control algorithm;

[0116] Thus, the design of an active disturbance rejection controller based on dynamic characteristic compensation of the pitch actuator was completed, which can then be used for pitch system control of wind turbines to improve the hysteresis of the pitch system.

[0117] Taking a land-based aircraft as an example, we illustrate the design method of the active disturbance rejection controller based on the dynamic characteristics of the pitch actuator. To simplify the design process, the active disturbance rejection controller used in this case is of the first order form.

[0118] 1. Obtain the dynamic characteristics of the pitch actuator

[0119] We contacted the pitch system supplier to conduct workshop tests and obtain the dynamic characteristics of the pitch actuator; based on a genetic algorithm, we identified the dynamic characteristics, and the results are as follows:

[0120]

[0121] 2. The extended state observer of the first-order active disturbance rejection controller is constructed as follows:

[0122]

[0123] Transform into discrete form:

[0124]

[0125] Since the controller uses a frequency of 50Hz, dt = 0.02s in the above formula.

[0126] 3. Based on the obtained dynamic characteristics of the pitch actuator, the input of the extended state observer is compensated, that is:

[0127]

[0128] θ compensation and θ demand The following relationship exists:

[0129]

[0130] Transforming it into discrete form, based on the backward difference method, we obtain:

[0131]

[0132] Where A = 1 / 17 2 B = 2 × 0.85 / 17 = 0.1.

[0133] 4. Design feedback control law

[0134] The feedback control law design for the first-order active disturbance rejection controller is as follows:

[0135]

[0136] Its discrete form is:

[0137]

[0138] 5. Parameter simplification

[0139] Utilizing controller bandwidth ω c and observer bandwidth ω o After simplifying the parameters, we have:

[0140]

[0141] In addition, determine ω c =10ω o

[0142] 6. Gain Scheduling

[0143] The rated wind speed of this model is 11.4 m / s, and the cutoff wind speed is 25 m / s. To simplify the design process, only two wind speed points were selected for gain control, with wind speeds of 14 m / s and 25 m / s respectively. The corresponding steady-state pitch angles are θ1 = 5.56 degrees and θ2 = 19.42 degrees. The parameters of the active disturbance rejection controller were tuned for the two operating points, and the parameter set is as follows:

[0144]

[0145] Therefore, the gain scheduling in this example is as follows:

[0146]

[0147] 7. Correction of the Active Disturbance Rejection Control Algorithm after Considering Additional Control Variables such as Feedforward

[0148] θ compensation (s)=G pitch (s)(θ demand (s)-θ delta (s)) (25)

[0149] This concludes the design of the active disturbance rejection controller based on dynamic characteristic compensation of the pitch actuator in this case.

[0150] 8. Simulation verification: Taking an average wind speed of 18 m / s and turbulence intensity Iref = 0.12 as an example, simulation verification was performed in Bladed simulation software. The simulation time was 600 s, and the fluctuation of the rotational speed was as follows. Figure 3 As shown in the figure, the rotational speed is precisely controlled near the set value with minimal fluctuations, thus demonstrating the effectiveness of the invention.

[0151] Example 2

[0152] This embodiment discloses a wind turbine pitch active disturbance rejection controller design system based on dynamic characteristic compensation, used to implement the wind turbine pitch active disturbance rejection controller design method based on dynamic characteristic compensation described in Embodiment 1, such as... Figure 4 As shown, the system includes the following functional modules:

[0153] The data acquisition module is used to acquire the dynamic characteristics of the pitch actuator of the wind turbine under design, as well as the input and output of the controlled system, in order to construct an extended state observer.

[0154] The compensation module compensates the input of the extended state observer based on the obtained dynamic characteristics of the pitch actuator. After compensation, the dynamic characteristics of the system observed in the extended state observer are accelerated. Then, a suitable control law is designed to accelerate the dynamic response speed of the closed-loop system.

[0155] The simplified module uses the bandwidth method to simplify the parameters, that is, it utilizes the controller bandwidth ω. c and observer bandwidth ω o Calculate the corresponding parameters;

[0156] The gain scheduling module adjusts ω c Gain scheduling is accomplished by adjusting the size of b0;

[0157] The correction module is used to correct the active disturbance rejection control algorithm after considering additional control variables, thus completing the design of the active disturbance rejection controller.

[0158] Example 3

[0159] This embodiment discloses a storage medium storing a program. When the program is executed by a processor, it implements the wind turbine pitch active disturbance rejection controller design method based on dynamic characteristic compensation as described in Embodiment 1.

[0160] The storage medium in this embodiment can be a disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), USB flash drive, portable hard drive, etc.

[0161] Example 4

[0162] This embodiment discloses a computing device, including a processor and a memory for storing processor-executable programs. When the processor executes the program stored in the memory, it implements the wind turbine pitch active disturbance rejection controller design method based on dynamic characteristic compensation described in Embodiment 1.

[0163] The computing device described in this embodiment may be a desktop computer, laptop computer, smartphone, PDA handheld terminal, tablet computer, programmable logic controller (PLC), or other terminal device with processor function.

[0164] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A design method for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation, characterized in that, Includes the following steps: 1) Obtain the dynamic characteristics of the pitch actuator of the wind turbine under design, and obtain the input and output of the controlled system to construct an extended state observer; 2) Based on the obtained dynamic characteristics of the pitch actuator, the input of the extended state observer is compensated. After compensation, the dynamic characteristics of the system observed in the extended state observer are accelerated. Then, a suitable linear feedback control law is designed to accelerate the dynamic response speed of the closed-loop system. The compensation for the input of the extended state observer based on the obtained dynamic characteristics of the pitch actuator is expressed as follows: (4)； In the formula, Represents the number of control cycle sequences. Representative variable In the The value of a periodic sequence, Representative variable In the The value of a periodic sequence, Representative variable In the The value of a periodic sequence, Representative variable In the The value of a periodic sequence, These are the internal parameters of the extended state observer. , The generator speed represents the speed of the wind turbine. The control law parameters characterize the control gain. The time interval representing the control cycle. This represents the compensated pitch control quantity. Representative variable In the The value of each periodic sequence is related to the pitch command as follows: (5)； In the formula, Represents the Laplace operator, The second-order transfer function representing the dynamic characteristics of the pitch actuator. Representative variable exist The value obtained after performing a Laplace transform on the domain. Representative variable exist The value obtained after performing a Laplace transform on the domain is: (6)； In the formula, The center frequency of the characteristic polynomial in the denominator of the second-order transfer function. Represents the damping ratio; Transforming the above equation into discrete form, based on the backward difference method, we get: （7）； In the formula, variables ,variable , Representative variable In the The value of a periodic sequence, Representative variable In the The value of a periodic sequence, Representative variable In the The value of a periodic sequence; 3) The parameters are simplified using the bandwidth method, that is, by utilizing the controller bandwidth. and observer bandwidth Calculate the corresponding parameters; 4) By adjusting and The size of the gain is used to complete the gain scheduling; 5) After considering the additional control variables, the active disturbance rejection control algorithm is modified, that is, the design of the active disturbance rejection controller is completed, and then it can be used for the pitch system control of the wind turbine to improve the lag of the pitch system.

2. The design method for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation according to claim 1, characterized in that, In step 1), the dynamic characteristics of the wind turbine's pitch actuator are represented by the following second-order transfer function: (1)； In the formula, Represents the Laplace operator, The second-order transfer function representing the dynamic characteristics of the pitch actuator. The center frequency of the characteristic polynomial in the denominator of the second-order transfer function. This represents the damping ratio.

3. The design method for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation according to claim 2, characterized in that, In step 1), there are two methods to obtain the dynamic characteristics of the actuator. The first method is based on workshop step test data. The specific test method is as follows: determine the software speed limit of the pitch actuator according to the wind turbine model, including the pitch recovery rate limit. and propeller speed limit Then, a ramp reference input test was performed, and the host computer provided a rate of... The paddle retraction command or rate is Record the oar-starting command signal. and pitch angle feedback value ,use and The first method can identify the dynamic characteristics of the pitch actuator; the second method is to export the pitch angle command value from the recorded data of the wind turbine based on the on-site operating data of the wind turbine. and pitch angle feedback value The recorded data is at least 2 minutes long and utilizes... and This allows for the identification of the dynamic characteristics of the pitch actuator.

4. The design method for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation according to claim 3, characterized in that, In step 1), for the pitch control loop of the wind turbine, the input quantity of the controlled system, i.e., the control quantity, is the pitch angle command. The system's output, i.e., the controlled variable, is the generator speed of the wind turbine. Based on the above input and output quantities, the various states of the system can be reconstructed using an extended state observer; if the designed active disturbance rejection controller has an order of... The state of the system is then used The term is used to represent the internal and external disturbances of the system, as well as the extended state of the unmodeled dynamics. express, The calculation method is as follows: (2)； In the formula, This represents the first derivative of the corresponding state variable. These are the internal parameters of the extended state observer. ,and Here, represents the control law parameter, characterizing the control gain; the above formula is for continuous systems. In discrete systems, we consider using the Euler method for discretization, and the discretized form of the formula is: (3)； In the formula, Represents the number of control cycle sequences. Representative variable In the The value of a periodic sequence, Representative variable In the The value of a periodic sequence, Representative variable In the The value of a periodic sequence, Representative variable In the The value of a periodic sequence, Representative variable In the The value of a periodic sequence, The time interval representing the control cycle can be used to perform discrete calculations of the extended state observer in electronic devices using the above formula.

5. The design method for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation according to claim 4, characterized in that, In step 2), the control law is designed as follows: (8)； In the formula, For the controller One parameter, The discrete form of the above equation is: (9)； This completes the design of the control law.

6. The design method for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation according to claim 5, characterized in that, In step 3), the extended state observer parameters include common The controller parameters include [number of parameters]. common The single parameter is not conducive to engineering use. Therefore, the parameter is simplified by using the bandwidth method, that is, by utilizing the controller bandwidth. and observer bandwidth The corresponding parameters are calculated according to the following rules: Regarding the controller parameters, we have: (10)； In the formula, represent of Power; For the extended state observer parameters, we have: (11)； In the formula, represent of Power; In this way, the parameters of the entire active disturbance rejection controller are simplified to three, namely... , and ,also, and It can also be further simplified based on the corresponding relationship; because This characterizes the speed of the dynamic response of the closed-loop system, while The extended state observer's tracking speed for each state of the closed-loop system is characterized by its own response speed. Therefore, to ensure the system's robustness and good dynamic performance, This simplifies the parameters of the active disturbance rejection controller to two, namely... and The number of parameters required for tuning is consistent with that of a conventional PI controller.

7. The design method for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation according to claim 6, characterized in that, In step 4), let the aerodynamic torque on the fan impeller be... The actual pitch angle of the wind turbine during operation is Due to the rated wind speed above The gain of the pitch controller increases with increasing wind speed, therefore the gain of the designed pitch controller decreases with increasing wind speed. This is reflected in a conventional PI controller, where the proportional gain decreases with increasing wind speed. However, to ensure the dynamic performance of the system, the integral gain increases. For an active disturbance rejection controller, this can be achieved by adjusting... and Gain scheduling is accomplished by adjusting the size of the variable, and the specific implementation method is as follows: Let the rated wind speed be Cut-out wind speed is Then in and Choose between Parameters were tuned at each working point, corresponding to the wind speed points. For each wind speed point, parameter tuning is performed to obtain the active disturbance rejection control parameter set. , That is, when the wind speed is The parameters of the active disturbance rejection controller obtained by tuning at the operating point. That is, when the wind speed is The parameters of the active disturbance rejection controller obtained by tuning at the operating point. That is, when the wind speed is The active disturbance rejection controller parameters were obtained by tuning at the operating point. Since the on-site wind speed measurement is not precise enough, directly using the measured wind speed for gain adjustment can easily introduce additional errors. Therefore, it is considered to utilize the parameters under steady-state conditions. Corresponding pitch angle Gain scheduling is performed, and the specific scheduling method is as follows: (12)； This completes the design of gain scheduling.

8. The design method for a wind turbine pitch active disturbance rejection controller based on dynamic characteristic compensation according to claim 7, characterized in that, In step 5), during the actual operation of the wind turbine, the pitch angle command is not only calculated by the feedback control algorithm, but also includes additional pitch angle control quantities introduced by feedforward or other control strategies. If the extended state observer introduces a pitch angle control command that includes additional pitch angle control quantities, it will cause steady-state errors in the active disturbance rejection control closed-loop calculation process. Therefore, it is necessary to consider removing the other additional commands from the pitch angle control command introduced into the extended state observer and only retain the control command calculated by the feedback control algorithm itself. Assume the additional pitch angle command is Then equation (5) is modified to: (13)； in, This is an additional pitch angle command besides the feedback control algorithm; Thus, the design of the active disturbance rejection controller based on dynamic characteristic compensation of the pitch actuator was completed.

9. A wind turbine pitch active disturbance rejection controller design system based on dynamic characteristic compensation, characterized in that, The method for implementing the wind turbine pitch active disturbance rejection controller design based on dynamic characteristic compensation as described in any one of claims 1 to 8 includes: The data acquisition module is used to acquire the dynamic characteristics of the pitch actuator of the wind turbine under design, as well as the input and output of the controlled system, in order to construct an extended state observer. The compensation module compensates the input of the extended state observer based on the obtained dynamic characteristics of the pitch actuator. After compensation, the dynamic characteristics of the system observed in the extended state observer are accelerated. Then, a suitable control law is designed to accelerate the dynamic response speed of the closed-loop system. The simplified module uses a bandwidth method to simplify the parameters, that is, it utilizes the controller bandwidth. and observer bandwidth Calculate the corresponding parameters; The gain scheduling module adjusts... and The size of the gain is used to complete the gain scheduling; The correction module is used to correct the active disturbance rejection control algorithm after considering additional control variables, thus completing the design of the active disturbance rejection controller.

Images