Design method and system of wind farm damping controller based on synovial membrane and predictive control

Abstract

The application relates to the technical field of damping controller design, in particular to a wind farm damping controller design method and system based on a sliding mode and predictive control. The method obtains input and output measurement data of wind farm operation, constructs a partial format dynamic linearization data model, and realizes adaptive updating of the model by using a pseudo-gradient updating mechanism; on this basis, a discrete sliding mode control item is calculated, the future operation state of the wind farm is predicted in combination with the control item, an optimal predictive control item is obtained by optimization solving; further, the discrete sliding mode control item and the optimal predictive control item are fused and optimized, a damping control output is generated, and the damping control output is injected into a reactive power modulation channel of a wind turbine generator to improve the operation stability and damping performance of the wind farm. The application realizes effective suppression of power oscillation in the grid-connected operation of the wind farm, and significantly improves the damping performance, operation stability and grid friendliness of the system.

Description

Technical Field

[0001] This application relates to the field of damping controller design technology, and in particular to a design method and system for a wind farm damping controller based on sliding diaphragm and predictive control. Background Technology

[0002] In related technologies, with the rapid development of new energy power generation technologies and the continuous advancement of the "dual carbon" target, the installed capacity and grid connection scale of new energy sources such as wind power are constantly expanding, and new energy power generation has become an important part of the new power system. Wind farms are usually connected to the grid through power electronic conversion devices, and their operating characteristics are significantly different from those of traditional synchronous generators. Under the condition of high proportion of new energy access, the equivalent inertia and damping level of the power system continue to decrease, and the system is more prone to low-frequency oscillations when subjected to disturbances, thus posing a potential threat to the safe and stable operation of the power grid.

[0003] However, existing wind farm damping control methods typically rely on precise system mathematical models. Wind farms are characterized by strong nonlinearity, strong coupling, and frequent changes in operating conditions, making it difficult to obtain model parameters accurately and in a timely manner. This results in insufficient adaptability of the controller to parameter uncertainties and changes in operating conditions during actual operation, making it difficult to maintain stable damping effects over the long term. Furthermore, while existing damping strategies using sliding mode control possess some robustness, they are prone to control chattering, affecting the smoothness of reactive power modulation. Methods based solely on predictive control are highly sensitive to model accuracy and disturbances, making it difficult to balance stability and control performance in complex grid environments. Simultaneously, most existing technologies do not adequately consider the prediction of future system operating states and the effective integration of sliding mode control and predictive control, failing to fully leverage their complementary advantages.

[0004] In summary, the technical problems existing in the relevant technologies need to be improved. Summary of the Invention

[0005] The main objective of this application is to propose a design method and system for a wind farm damping controller based on sliding diaphragm and predictive control, so as to effectively suppress power oscillations during grid-connected operation of wind farms and significantly improve the damping performance, operational stability and grid friendliness of the system.

[0006] To achieve the above objectives, one aspect of this application proposes a design method for a wind farm damping controller based on sliding diaphragm and predictive control, the method comprising the following steps: Acquire input and output measurement data for wind farm operation; Based on the input and output measurement data, a partial scheme dynamic linearized data model of the wind farm is established. A pseudo-gradient update mechanism is used to adaptively update the partially formatted dynamic linearized data model. Based on the updated partial scheme dynamic linearized data model, the discrete sliding mode control term is calculated; Based on the discrete sliding mode control term, the future operating state of the wind farm is predicted, and the optimal predictive control term is obtained through optimization calculation. The discrete sliding mode control term and the optimal predictive control term are fused and optimized to form a damped control output; The damping control output is injected into the reactive power modulation channel of the wind turbine.

[0007] In some embodiments, establishing a partial-format dynamic linearized data model of the wind farm based on the input measurement data and the output measurement data includes: Based on the input and output measurement data of the wind farm, a discrete input-output data sequence of the wind farm is constructed; The discrete input-output data sequence is subjected to partial scheme dynamic linearization to obtain an equivalent linear data model represented by a pseudo gradient vector. The pseudo gradient vector is introduced into the equivalent linear data model to form a partial scheme dynamic linearized data model of the wind farm.

[0008] In some embodiments, the formula for the partial-format dynamic linearized data model is as follows: ; Where k represents the current sampling time; Indicates at time Relative to time The change in output; This represents a pseudo-gradient vector; Indicates the increment of the control input; This represents the generalized disturbance increment term.

[0009] In some embodiments, the calculation formula for the pseudo-gradient update mechanism is as follows: ; in, Indicates time The pseudo gradient vector estimate; This represents the initial value of the pseudo-gradient vector; Indicates time The pseudo gradient vector estimate; This represents the change in output between adjacent sampling times; Indicates the increment of the control input; Indicates the threshold constant; This represents the step size factor, which controls the adjustment range of the pseudo-gradient update; Indicates the weighting factor; This indicates the output incremental prediction error term; This represents the transpose of the control input increment.

[0010] In some embodiments, calculating the discrete sliding mode control term based on the updated partial scheme dynamic linearized data model includes: Based on the partial scheme dynamic linearized data model, a discrete sliding mode surface function for wind farm damping control is constructed. Based on the discrete sliding surface function, design a discrete sliding mode reaching law; The discrete sliding mode approach law and the partial scheme dynamic linearized data model are used to calculate the corresponding discrete sliding mode control law. The discrete sliding mode control law is processed into a continuous form to obtain the discrete sliding mode control term.

[0011] In some embodiments, the calculation formula for the discrete sliding mode control term is as follows: ; in, Indicates time Discrete sliding mode control term; Indicates time The pseudo gradient estimate; Represents a small positive constant; Indicates time The actual output feedback signal; Indicates time Discrete sliding mode states; Indicates the constant parameters of the sliding surface; Represents the discrete sliding mode approaching term; Indicates the weighting coefficient of the sliding surface; Indicates time Tracking error; Represents the equivalent disturbance estimate; This indicates the output reference value at the next sampling time. Represents the sliding mode reaching law coefficient; This represents the output value of the symbolic function.

[0012] In some embodiments, the step of predicting the future operating state of the wind farm based on the discrete sliding mode control term and obtaining the optimal predictive control term through optimization calculation includes: Based on the discrete sliding mode control term, the sliding mode state corresponding to the wind farm is predicted one step forward, and the multi-step sliding mode state in the prediction time domain is obtained by recursion. The multi-step sliding mode states in the prediction time domain are constructed into a sliding mode state prediction vector; Based on the sliding mode state prediction vector, a performance index function is constructed; The performance index function is optimized to obtain the optimal predictive control term.

[0013] In some embodiments, the optimal predictive control term is calculated using the following formula: ; in, Indicates time The optimal predictive control term; This represents the optimal predictive control sequence vector within the current and future prediction time domains; This represents the matrix used to select the control vector; and Represents a time-varying matrix; Representation matrix The transpose of the matrix; Represents a positive definite weighted matrix; Indicates the penalty coefficient; Represents the identity matrix; and Represents a constant matrix; This represents the known historical best predictive control term vector; This represents the discrete sliding mode control term vector.

[0014] In some embodiments, the step of fusing and optimizing the discrete sliding mode control term and the optimal predictive control term to form a damped control output includes: Based on preset fusion weight coefficients, the discrete sliding mode control term and the optimal predictive control term are weighted and combined to construct the overall control law expression; Based on the wind farm's operating status and control constraints, the overall control law expression is subjected to amplitude and rate of change constraints. The overall control law expression after constraint processing is smoothed and optimized to obtain the final damped control output.

[0015] To achieve the above objectives, another aspect of this application proposes a wind farm damping control system based on sliding diaphragm and predictive control, the system comprising: The data acquisition module is used to acquire input and output measurement data for wind farm operation. The model building module is used to establish a partial-format dynamic linearized data model of the wind farm based on the input measurement data and the output measurement data. The model update module is used to adaptively update the partial-format dynamic linearized data model using a pseudo-gradient update mechanism. The discrete sliding mode control module is used to calculate the discrete sliding mode control term based on the updated partial scheme dynamic linearized data model. The optimal predictive control module is used to predict the future operating state of the wind farm based on the discrete sliding mode control term, and to obtain the optimal predictive control term through optimization calculation. The fusion optimization module is used to fuse and optimize the discrete sliding mode control term and the optimal predictive control term to form a damped control output; The output control module is used to inject the damping control output into the reactive power modulation channel of the wind turbine.

[0016] To achieve the above objectives, another aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method.

[0017] To achieve the above objectives, another aspect of the embodiments of this application proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0018] To achieve the above objectives, another aspect of this application provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0019] The embodiments of this application include at least the following beneficial effects: This application provides a design method and system for a wind farm damping controller based on sliding mode and predictive control. This scheme constructs a partially formatted dynamic linearized data model through wind farm input and output measurement data, realizing an online equivalent description of the nonlinear and strongly coupled dynamic characteristics of the wind farm. This reduces the dependence on precise physical models and fixed parameters, and improves the adaptability and practicality of the control model under complex operating conditions. By introducing a pseudo-gradient update mechanism to adaptively correct the model, the control system can track wind speed fluctuations, grid disturbances, and changes in operating status in real time, avoiding the degradation of control performance caused by model mismatch. In terms of control strategy, this invention combines discrete sliding mode control with predictive control. It utilizes the robust suppression capability of sliding mode control against uncertainties and external disturbances to ensure rapid system convergence and stable operation; at the same time, it utilizes the predictive and optimization capability of predictive control for future operating states to obtain the optimal control input under the premise of satisfying constraints, thereby balancing control response speed and control smoothness, and reducing the chattering problem that is prone to occur in traditional sliding mode control. By fusing and optimizing discrete sliding mode control terms and optimal predictive control terms, and injecting the resulting damping control output into the reactive power modulation channel of the wind turbine, the power oscillations and low-frequency oscillations generated during the grid-connected operation of the wind farm can be effectively suppressed, thereby improving the damping support capability and grid-connected stability of the wind farm and enhancing the overall safety and grid friendliness of the wind farm operation. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating a wind farm damping controller design method based on sliding diaphragm and predictive control provided in an embodiment of this application. Figure 2 This is a schematic diagram of the additional damping control framework based on a wind farm provided in an embodiment of this application; Figure 3 This application provides a schematic diagram of the internal structure of a wind farm damping controller according to an embodiment. Figure 4 This is a structural diagram of the New England 10-unit 39-bus system containing a doubly fed wind farm provided in the embodiments of this application; Figure 5 The simulation results of rotor power angle difference, speed difference and tie line power under scenario 1 provided in the embodiments of this application are shown in the figure. Figure 6 The simulation results of rotor power angle difference, speed difference and tie line power under scenario 2 provided in the embodiments of this application are shown in the figure. Figure 7 This is a schematic diagram of a wind farm damping control system based on sliding diaphragm and predictive control provided in an embodiment of this application. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit it. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with those of this application; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this application as detailed in the appended claims.

[0022] It is understood that the terms “first,” “second,” etc., used in this application may be used herein to describe various concepts, but unless otherwise stated, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the words “if,” “when,” or “in response to a determination” as used herein may be interpreted as “when…” or “when…” or “in response to a determination.”

[0023] As used in this application, the terms "at least one", "multiple", "each", "any", etc., "at least one" includes one, two or more, "multiple" includes two or more, "each" refers to each of the corresponding multiples, and "any" refers to any one of the multiples.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0025] This application provides a design method and system for a wind farm damping controller based on sliding mode and predictive control. This scheme constructs a partially formatted dynamic linearized data model using input and output measurement data from the wind farm, achieving an online equivalent description of the nonlinear and strongly coupled dynamic characteristics of the wind farm. This reduces the dependence on precise physical models and fixed parameters, improving the adaptability and practicality of the control model under complex operating conditions. By introducing a pseudo-gradient update mechanism to adaptively correct the model, the control system can track wind speed fluctuations, grid disturbances, and changes in operating status in real time, avoiding performance degradation caused by model mismatch. Regarding the control strategy, this invention combines discrete sliding mode control with predictive control. Sliding mode control's robust suppression of uncertainties and external disturbances ensures rapid system convergence and stable operation; simultaneously, predictive control's ability to predict and optimize future operating states yields optimal control input while meeting constraints, thus balancing control response speed and control smoothness, and reducing chattering problems easily generated in traditional sliding mode control. By fusing and optimizing discrete sliding mode control terms and optimal predictive control terms, and injecting the resulting damping control output into the reactive power modulation channel of the wind turbine, the power oscillations and low-frequency oscillations generated during the grid-connected operation of the wind farm can be effectively suppressed, thereby improving the damping support capability and grid-connected stability of the wind farm and enhancing the overall safety and grid friendliness of the wind farm operation.

[0026] The wind farm damping controller design method based on sliding diaphragm and predictive control provided in this application relates to the field of damping controller design technology. This wind farm damping controller design method based on sliding diaphragm and predictive control can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or vehicle terminal, but is not limited to these. The server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network. The software can be an application implementing the wind farm damping controller design method based on sliding diaphragm and predictive control, but is not limited to the above forms.

[0027] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0028] Figure 1 This is an optional flowchart of a wind farm damping controller design method based on sliding diaphragm and predictive control provided in an embodiment of this application. Figure 1 The method may include, but is not limited to, steps S1 to S7: S1: Acquire input and output measurement data for wind farm operation; In this embodiment, the operating signals of the wind farm and its grid-connected power system are first measured online under the grid-connected operation status of the wind farm. The input measurement data mainly includes control input quantities that can be directly obtained or indirectly calculated by the rotor-side controller of the wind turbine, such as the rotor-side reactive power modulation reference signal and the additional damping control signal; the output measurement data includes key response signals that can characterize the low-frequency oscillation characteristics of the system, such as synchronous generator speed deviation, relative speed difference, power angle difference, or tie-line active power deviation.

[0029] In practical implementation, to improve the observability of the target low-frequency oscillation modes, the dominant mode ratio method is preferred. This method selects the output signal that contributes significantly to the mode of interest from multiple measurable signals as the system output feedback quantity. For example, in a multi-regional power system containing a doubly-fed wind farm, the relative speeds between specific synchronous generators can be selected as the output measurement data. All input and output measurement data are discretized according to a unified sampling period to form the time-series input / output data sequence required for subsequent data modeling and control calculations.

[0030] refer to Figure 2 As shown, Figure 2 This is a schematic diagram of a wind farm damping control structure based on sliding diaphragm and predictive control. The overall structure includes a doubly-fed induction generator (DFIG), a rotor-side controller (RSC), a grid-side controller (GSC), a DC bus, a power system network, and an additional damping controller. The wind turbine is connected to the power system via back-to-back converters, which include a rotor-side converter and a grid-side converter, connected by a DC bus capacitor. Energy coupling is performed to achieve independent regulation of active and reactive power.

[0031] At the wind turbine control level, the rotor-side controller adopts a current dual-closed-loop control structure based on a synchronous rotating coordinate system, converting the three-phase current signals to [the appropriate coordinate system] via a coordinate transformation module. The coordinate system is used to generate rotor-side voltage commands through current regulators, which are then used to drive the rotor-side converter via pulse width modulation (PWM) modules. The active power control channel of the rotor-side controller is used to regulate electromechanical torque to achieve power point tracking, while the reactive power control channel is used to regulate the reactive power output on the stator side, thereby providing voltage support and reactive power regulation capabilities for the power system.

[0032] The speed signals of multiple synchronous generators in the power system are collected and constructed into a system oscillation characterization quantity, forming the output feedback signal. This signal serves as the input measurement data for the damping controller. The damping controller is based on the online-acquired input / output data and constructs a partial scheme dynamic linearized data model. It adaptively tracks the system dynamics through a pseudo-gradient update mechanism and further combines discrete sliding mode control and optimal predictive control to calculate the additional damping control output.

[0033] Damping control output with additional reactive power modulation signal The reactive power is injected into the rotor-side controller's reactive power control channel in the form of a variable frequency drive (VFD), and superimposed on the existing reactive power reference command. By adjusting the reactive power output of the wind turbine, the wind farm provides additional damping to the grid when low-frequency oscillations occur in the system, thereby suppressing low-frequency oscillations in the power system. This structure does not require changes to the original control framework of the wind turbine, fully utilizes the existing reactive power modulation channel to achieve data-driven damping control, and has good engineering feasibility, robustness, and adaptability to changes in operating conditions.

[0034] S2: Based on the input and output measurement data, establish a partial scheme dynamic linearized data model for the wind farm; Among them, a partial-format dynamic linearized data model of the wind farm is established based on input and output measurement data, including: Based on the input and output measurement data of the wind farm, a discrete input-output data sequence of the wind farm is constructed; The discrete input-output data sequence is subjected to partial scheme dynamic linearization to obtain an equivalent linear data model represented by a pseudo gradient vector; By introducing pseudo-gradient vectors into the equivalent linear data model, a partial scheme dynamic linearized data model of the wind farm is formed.

[0035] In this embodiment, after acquiring the input and output measurement data during the wind farm's operation, the data is first discretized according to a uniform sampling period to construct a discrete input-output data sequence for the wind farm. The input measurement data consists of the control input signals for damping regulation in the wind farm, while the output measurement data is a dynamic response signal reflecting the low-frequency oscillation characteristics of the power system. By organizing the input and output values ​​at multiple consecutive sampling times in a time sequence, a discrete input-output data sequence characterizing the dynamic changes of the system is formed.

[0036] After obtaining the discrete input-output data sequence, a partial scheme dynamic linearization process is performed on it. This process does not rely on the physical structure and parameter information of the wind turbine and power system, but rather on the mapping relationship between changes in input data and changes in output data to provide an equivalent linear description of the system's nonlinear dynamic behavior. Through this partial scheme dynamic linearization process, the originally complex nonlinear system is transformed into an equivalent linear data model that uses pseudo-gradient vectors to characterize the system's dynamic characteristics, enabling the system's dynamic behavior under current operating conditions to be characterized in data form.

[0037] Furthermore, pseudo-gradient vectors are introduced into the equivalent linear data model to form a partial scheme dynamic linearized data model for the wind farm. This data model can reflect changes in the dynamic characteristics of the system by adjusting the pseudo-gradient parameters as operating conditions and disturbance conditions change. This provides a unified and reliable data model foundation for subsequent pseudo-gradient adaptive updates, discrete sliding mode control, and predictive control calculations, and is suitable for grid-connected wind farm scenarios where system parameters are unknown and operating states change frequently.

[0038] Specifically, the formula for the partially formatted dynamic linearized data model is as follows: ; Where k represents the current sampling time; Indicates at time Relative to time The change in output; This represents a pseudo-gradient vector; Indicates the increment of the control input; This represents the generalized disturbance increment term.

[0039] S3: Adaptive updates of the partially formatted dynamic linearized data model are performed using a pseudo-gradient update mechanism. After establishing a partially skewed dynamic linearized data model, the dynamic characteristics of the wind farm grid-connected system exhibit significant time-varying and nonlinear features due to various uncertainties such as load fluctuations, wind speed changes, network topology switching, and fault disturbances during operation. To avoid the accumulation of modeling errors caused by using a fixed parameter model, this embodiment introduces a pseudo-gradient update mechanism based on online input and output measurement data to adaptively correct the pseudo-gradient vector in the model in real time, enabling the data model to continuously reflect the actual operating status of the system.

[0040] Specifically, within each sampling period, a pseudo-gradient update criterion function is constructed based on the input and output increments between the current and previous times. By introducing a step size factor and a weighting factor, the pseudo-gradient parameters are recursively updated. Simultaneously, bounded constraints are applied to the range of pseudo-gradient changes during the update process to prevent parameter divergence due to measurement noise or transient disturbances, thereby ensuring the numerical stability and convergence of the pseudo-gradient estimation process. Through this update mechanism, the pseudo-gradient can adaptively adjust itself relying solely on real-time input / output data, without requiring any system physical parameter information.

[0041] By continuously implementing a pseudo-gradient update mechanism, the partially scheme dynamic linearized data model can automatically correct its parameters when the system's operating conditions change, ensuring that the model always maintains its ability to effectively characterize the system's dynamics. This adaptive update process provides a reliable data foundation for the subsequent calculation of discrete sliding mode control terms and optimal predictive control terms, thereby ensuring that the overall damping controller has good robustness and adaptability under different operating points and complex disturbance conditions.

[0042] The calculation formula for the pseudo-gradient update mechanism is as follows: ; in, Indicates time The pseudo gradient vector estimate; This represents the initial value of the pseudo-gradient vector; Indicates time The pseudo gradient vector estimate; This represents the change in output between adjacent sampling times; Indicates the increment of the control input; This represents the threshold constant, which is a sufficiently small positive constant. This represents the step size factor, which controls the adjustment range of the pseudo-gradient update; Indicates the weighting factor; This indicates the output incremental prediction error term; This represents the transpose of the control input increment.

[0043] S4: Calculate the discrete sliding mode control term based on the updated partial scheme dynamic linearized data model; Among them, based on the updated partial scheme dynamic linearized data model, the discrete sliding mode control term is calculated, including: Based on the partial scheme dynamic linearized data model, a discrete sliding mode surface function for wind farm damping control is constructed. Design a discrete sliding mode reaching law based on the discrete sliding mode surface function; The discrete sliding mode reaching law and the partial scheme dynamic linearized data model are used to calculate the corresponding discrete sliding mode control law. The discrete sliding mode control law is processed into a continuous form to obtain the discrete sliding mode control term.

[0044] In this embodiment, after obtaining the updated partial scheme dynamic linearized data model, to enhance the wind farm's ability to participate in grid oscillation suppression under disturbance and parameter uncertainty conditions, discrete sliding mode control is designed based on the data model to calculate the discrete sliding mode control term required for wind farm damping control. This discrete sliding mode control term serves as the basic control quantity for subsequent predictive control and fusion control, primarily used to improve the robust damping performance of the system under different operating conditions and complex disturbances.

[0045] In practical implementation, based on the updated partial scheme dynamic linearized data model, the output feedback signal that reflects the low-frequency oscillation characteristics of the power system is selected as the control object, and a discrete sliding surface function for wind farm damping control is constructed. The discrete sliding surface function focuses on the deviation between the output feedback signal and its reference value. By discretely describing the trend of deviation changes, the system state has a clear convergence target near the sliding surface. Simultaneously, the construction process of the sliding surface matches the data model parameters under the current operating conditions, avoiding a decrease in control performance due to dynamic changes in the system.

[0046] Specifically, the formula for the discrete sliding surface function is as follows: ; ; In the formula, Indicates time Discrete sliding surface function values; Indicates time The sliding mode state; Indicates time The sliding mode state; This represents a constant to ensure convergence; Indicates the output reference value; Indicates time The output tracking error; Indicates time Historical tracking error; Indicates time The actual output feedback signal.

[0047] After obtaining the discrete sliding mode surface function, a discrete sliding mode reaching law is designed based on the function to constrain the evolution direction and convergence rate of the sliding mode state at discrete sampling times. By reasonably setting the reaching law parameters, the system state can maintain its stable convergence to the sliding surface even under external disturbances and modeling errors, thereby accelerating the oscillation decay rate and enhancing the damping effect. Subsequently, the corresponding discrete sliding mode control law is calculated by combining the discrete sliding mode reaching law with the partial scheme dynamic linearized data model.

[0048] Specifically, the formula for calculating the discrete sliding mode reaching law is as follows: ; In the formula, Indicates time The discrete sliding surface function value; q represents the sliding mode convergence factor. This represents a linear convergence term, used to ensure that the sliding surface function decays exponentially under undisturbed conditions; This represents the reaching law gain coefficient; The sign function, used to indicate the sign of the sliding surface function, has its output defined as: when hour, ; when hour, ; when hour, .

[0049] This symbolic function is used to implement the switching characteristics in sliding mode control.

[0050] Furthermore, considering that switching characteristics in traditional sliding mode control may cause control chattering, and that the reactive power modulation channel of wind turbines has engineering requirements for the smoothness of the control signal, this embodiment performs continuous processing on the discrete sliding mode control law to obtain the final discrete sliding mode control term. The control term after continuous processing maintains the robustness of sliding mode control while effectively reducing the abrupt amplitude and high-frequency jitter of the control signal, providing a stable and feasible basic control input for subsequent predictive control calculations and control quantity fusion optimization.

[0051] Specifically, the calculation formula for the discrete sliding mode control term is as follows: ; in, Indicates time Discrete sliding mode control term; Indicates time The pseudo gradient estimate; Represents a small positive constant; Indicates time The actual output feedback signal; Indicates time Discrete sliding mode states; Indicates the constant parameters of the sliding surface; Represents the discrete sliding mode approaching term; Indicates the weighting coefficient of the sliding surface; Indicates time Tracking error; Represents the equivalent disturbance estimate; This indicates the output reference value at the next sampling time. Represents the sliding mode reaching law coefficient; This represents the output value of the symbolic function.

[0052] S5: Based on discrete sliding mode control terms, predict the future operating state of the wind farm and obtain the optimal predictive control terms through optimization calculation; Among them, based on discrete sliding mode control terms, the future operating state of the wind farm is predicted, and the optimal predictive control terms are obtained through optimization calculations, including: Based on discrete sliding mode control terms, the sliding mode state corresponding to the wind farm is predicted one step ahead, and the multi-step sliding mode state in the prediction time domain is obtained by recursion. The multi-step sliding mode states in the prediction time domain are constructed into sliding mode state prediction vectors; A performance index function is constructed based on the sliding mode state prediction vector; The optimal predictive control term is obtained by optimizing the performance index function.

[0053] In this embodiment, after obtaining the discrete sliding mode control term, to further reduce the chattering phenomenon that may be caused by sliding mode control, and to improve the smoothness of the control signal and engineering feasibility while ensuring the damping effect, optimal predictive control is introduced to predict and optimize the future operating state of the wind farm. This step is based on the sliding mode state formed by discrete sliding mode control, and by predicting the state evolution trend in the time domain, the control action is proactively adjusted, thereby improving the overall dynamic performance of the system.

[0054] In practice, based on the sliding mode state at the current sampling moment and the calculated discrete sliding mode control term, combined with the updated partial scheme dynamic linearized data model, the sliding mode state corresponding to the wind farm is predicted one step forward. Subsequently, following the same prediction logic, recursive calculations are performed within a preset prediction time domain to obtain the sliding mode states corresponding to multiple future sampling moments, thus forming a multi-step sliding mode state sequence that reflects the future low-frequency oscillation trend. In this way, the controller can predict the direction of the system's dynamic response over a future period of time at the current moment.

[0055] After obtaining the multi-step sliding mode states in the prediction time domain, these states are organized chronologically to construct a sliding mode state prediction vector. Based on this vector, a performance index function is further constructed to comprehensively measure the oscillation suppression effect and control adjustment cost within the prediction time domain. This performance index function reflects the overall change in sliding mode state deviation and constrains the control adjustment amplitude or rate of change to avoid overly drastic control actions, thereby meeting the engineering requirements for control smoothness in the reactive power modulation channel of the wind turbine.

[0056] After constructing the performance index function, the function is optimized to obtain the predictive control result that achieves the optimal performance index. The optimal predictive control term corresponding to the current sampling time is then extracted from this result. This optimal predictive control term is updated in real time according to the system operating conditions in each sampling period. Together with the discrete sliding mode control term, it constitutes the input for subsequent fusion control. This effectively suppresses chattering and improves the overall control performance of wind farms participating in the suppression of low-frequency oscillations in the power grid while maintaining robust damping capability.

[0057] Specifically, in order to design the optimal predictive control term ,make To predict the total number of steps, the sliding mode state is further derived. The forward prediction equation is as follows: ; Define the following variables: ; ; ; ; In the formula, For the present and the future A vector consisting of sliding mode variables at each time step. This is a known vector composed of the optimal predictive control terms at past time points. Let be the vector to be determined, consisting of the optimal predictive control terms for the current and future times. The known vectors of the discrete sliding mode control term group layer at the current and past time points; This represents the pseudo-gradient estimate at the end of the prediction time domain, used to describe the impact of future control inputs on the sliding mode state; This represents the pseudo-gradient increment term; This indicates the pseudo-gradient order.

[0058] Furthermore, the calculation yields: ; In the formula, the matrix , , , , The expression is as follows: ; ; ; ; Further derivation yields: ; In the formula, For the future A vector consisting of the sliding mode states at each time step. and It is a constant matrix. and It is a time-varying matrix, matrix The expression is as follows: ; ; ; ; ; Next, by solving The optimal value is then used to obtain the optimal predictive control term. Therefore, a performance function that comprehensively considers tracking error and control cost is introduced as follows: ; In the formula, This represents the optimal predictive control performance function value; This represents the predicted sliding mode state vector; Represents the predicted sliding mode state vector The transpose of ; This represents the optimal predictive control input vector to be determined. Represents the predictive control input vector The transpose of ; This is the penalty coefficient.

[0059] minimize The result is as follows: ; In the formula, It is an identity matrix.

[0060] Finally, the formula for calculating the optimal predictive control term is as follows: ; in, Indicates time The optimal predictive control term; This represents the optimal predictive control sequence vector within the current and future prediction time domains; This represents the matrix used to select the control vector; and Represents a time-varying matrix; Representation matrix The transpose of the matrix; Represents a positive definite weighted matrix; Indicates the penalty coefficient; Represents the identity matrix; and Represents a constant matrix; This represents the known historical best predictive control term vector; This represents the discrete sliding mode control term vector.

[0061] S6: Fusion optimization of discrete sliding mode control term and optimal predictive control term to form damped control output; The discrete sliding mode control term and the optimal predictive control term are fused and optimized to form the damped control output, including: Based on the preset fusion weight coefficients, the discrete sliding mode control term and the optimal predictive control term are weighted and combined to construct the overall control law expression; Based on the wind farm's operating status and control constraints, the overall control law expression is subjected to amplitude and rate of change constraints. The overall control law expression after constraint processing is smoothed and optimized to obtain the final damped control output.

[0062] In this embodiment, after obtaining the discrete sliding mode control term and the optimal predictive control term, to simultaneously leverage the robustness of sliding mode control and the smooth optimization characteristics of predictive control, the two types of control terms are fused and optimized to form an overall control output. Specifically, based on preset fusion weight coefficients, the discrete sliding mode control term and the optimal predictive control term are weighted and combined to construct the overall control law expression. The fusion weight coefficients can be set according to the wind farm's operating conditions and the system oscillation intensity, so that the overall control emphasizes rapid vibration suppression during periods of severe oscillation and focuses on control smoothness and energy efficiency during periods of oscillation decay.

[0063] After obtaining the overall control law expression, engineering constraints are applied to the overall control law based on the real-time operating status of the wind farm and control constraints. Specifically, this includes limiting the amplitude of the control output to ensure that the control command does not exceed the allowable range of the wind turbine's reactive power modulation channel, avoiding reactive power saturation or adverse effects on the original voltage control; at the same time, the rate of change of the control output is constrained to limit the abrupt change in the control quantity at adjacent sampling times, in order to prevent excessively rapid excitation of the rotor-side controller and reduce the risk of control jitter and execution instability.

[0064] After constraining the amplitude and rate of change, the overall control law after constraint processing is smoothed and optimized to obtain the final damping control output. Through smoothing optimization, the control output maintains effective damping while possessing good continuity and noise immunity, thereby better matching the dynamic characteristics of the reactive power modulation outer loop of the wind turbine. This provides a stable and implementable control signal basis for subsequently injecting the damping control output into the reactive power modulation channel.

[0065] S7: Injects the damping control output into the reactive power modulation channel of the wind turbine.

[0066] In this embodiment, after obtaining the fused damping control output, the control signal is introduced as an additional damping control quantity into the reactive power modulation outer loop of the wind turbine rotor-side controller. Specifically, the damping control output is superimposed on the original reactive power control command in the form of an additional reactive power reference signal, enabling the wind turbine to actively participate in damping regulation when the system experiences low-frequency oscillations while maintaining its original voltage regulation and reactive power control functions.

[0067] At the engineering implementation level, since the rotor-side controller of a doubly-fed wind turbine typically adopts a hierarchical control structure, the reactive power modulation outer loop has good dynamic response characteristics and adjustment margin. Therefore, injecting the damping control output into this channel will not destroy the stability of the original control system. The damping control output changes in real time with the system oscillation state. By adjusting the reactive power injection or absorption characteristics of the wind turbine, an additional damping torque matching the oscillation mode is formed in the power grid, thereby suppressing the continuous growth of the low-frequency oscillation amplitude of the system.

[0068] By employing the aforementioned injection method, wind farms can effectively suppress low-frequency oscillations in the power system without altering their original main control structure and parameter configuration. Simulation results demonstrate that, under conditions of system topology changes, operating condition switching, and multiple superimposed disturbances, this reactive power modulation method can continuously provide a stable damping effect, significantly shortening the oscillation decay time and improving the overall stability and operational safety of the renewable energy grid-connected power system.

[0069] refer to Figure 3 As shown, Figure 3 This diagram illustrates the internal structure of a wind farm damping controller based on sliding mode and predictive control. The controller is located between the wind farm and the power system. Its input is an oscillation-related feedback signal from the power system, and its output is an additional damping control signal acting on the rotor-side controller of the doubly-fed induction generator (DFIG). The dashed box in the diagram represents the main structure of the damping controller, while the area outside the dashed box represents the controlled wind farm grid-connected power system.

[0070] In the main structure of the damping controller, a pseudo-gradient updater is first set up. This updater updates the pseudo-gradient parameters in the partial scheme dynamic linearized data model online based on real-time acquired input and output measurement data. This is used to characterize the dynamic characteristics of the wind farm grid-connected system under the current operating conditions. The pseudo-gradient updater provides real-time updated data model parameters for subsequent control modules, enabling the controller to track system dynamic changes without relying on the system's physical model.

[0071] Based on the pseudo-gradient updater, the controller internally incorporates a discrete sliding mode control module and an optimal predictive control module in parallel. The discrete sliding mode control module constructs a discrete sliding surface and generates discrete sliding mode control variables based on the updated data model, providing fast and robust damping when the system experiences disturbances and parameter uncertainties. The optimal predictive control module optimizes the system state in the prediction time domain based on the future evolution trend of the sliding mode state, obtaining a smooth and optimized predictive control variable to mitigate chattering phenomena that may be caused by sliding mode control. Figure 2 The delay unit shown ( It is used to realize discrete-time recursion and multi-step prediction calculation of sliding mode state and control quantity.

[0072] The control quantities output by the discrete sliding mode control module and the optimal predictive control module are weighted, superimposed, and differentially processed (diff) to form an overall control law, which serves as the final damping control output. This damping control output is fed into the reactive power modulation channel of the wind turbine rotor-side controller. By adjusting the reactive power output of the DFIG, the wind farm provides additional damping to the grid when low-frequency oscillations occur in the power system, thereby effectively suppressing system oscillations. This control structure is entirely data-driven and can maintain good damping performance and engineering feasibility under conditions of unknown system parameters, changing operating conditions, and complex disturbances.

[0073] This embodiment is... Figure 4 The three-region New England 10-unit 39-node system shown was validated, including one doubly fed wind farm.

[0074] Figure 4 In this system, the entire power system is divided into multiple interconnected regions, including Region 1, Region 2, and Region 3. These regions are connected by interconnecting lines, forming a typical multi-region interconnected power system structure. Each region has several synchronous generator nodes to characterize the distribution of conventional power sources in different regions. Additionally, some regions are connected to wind farms to reflect the impact of renewable energy grid connection on the system's dynamic characteristics.

[0075] The circular markers in the diagram indicate the locations of power generation units or key nodes in the system. Synchronous generators located within the region provide basic active and reactive power support to the system. Wind farm nodes located in specific areas are connected to the system via grid interfaces and participate in system oscillation suppression as additional damping resources in the method of this invention. The arrows in the diagram indicate the direction of disturbances or the location of power disturbances experienced by the system during operation. Disturbances can be caused by factors such as sudden load changes, line faults, short-circuit events, or topology changes.

[0076] In this multi-regional system structure, when a disturbance occurs in one region, it propagates to other regions through interconnecting lines, inducing low-frequency oscillation modes across regions. By introducing a sliding mode and predictive control-based damping controller in regions containing wind farms, the wind farms can generate additional damping control outputs based on system output feedback signals and participate in system regulation through reactive power modulation channels, thereby effectively suppressing low-frequency oscillations within and between regions. This figure illustrates a typical scenario of the application of the method of this invention in a multi-regional interconnected system, providing a system structure reference for subsequent simulation analysis and effect verification.

[0077] The two initial operating conditions are shown in Table 1 below: Table 1: Initial working conditions Wind farm output Area 1 (SG 8-10) Area 2 (SG 1-3) Area 3 (SG 4-7) 1 480 MW 1620 MW 2485 MW 2350 MW 2 360 MW 1620 MW 2285 MW 2350 MW Using the principal mode ratio method, the signal with high observability for the mode of interest, namely the relative speeds of synchronizers 2 and 9, is selected. As the feedback signal, the control signal is selected as the additional reactive power reference signal on the RSC reactive power modulation loop. A damping controller was designed based on the aforementioned method. By comparing the oscillation waveforms of the system before and after adding the proposed data-driven adaptive damping controller under different operating scenarios 1-2, and by comparing them with traditional power oscillation damping methods and model-free adaptive control methods, the method proposed in this invention can be verified.

[0078] Scenario 1: The system is running in initial condition 1. t At time 1 second, a three-phase ground fault occurs between lines 4 and 14 in region 2. t At 1.1s, the circuit breaker of line 4-14 trips but does not close to eliminate the fault, at which point the system topology changes. t At 20 seconds, a three-phase ground fault occurred at busbar 12 in region 2, and simultaneously, the load at node 15 in region 3 increased by 200MW. t The short circuit fault cleared itself after 20.1 seconds.

[0079] Scenario 2: The system is running in initial condition 2. t At 1 second, a three-phase ground fault occurred between lines 7 and 8 in region 2.t =1.1s, the circuit breaker of line 7-8 tripped and did not close; t At 20 seconds, a three-phase ground fault occurred at busbar 12 in region 2, and simultaneously, the load at node 27 in region 1 decreased by 200MW. t The short circuit fault cleared itself after 20.1 seconds.

[0080] The simulation calculations of the embodiments were performed using the method of the present invention, and the results are as follows: Figure 5 Simulation results for rotor power angle difference, speed difference, and tie-line power under scenario 1 are presented. The results show that, in both disturbances, without a damping controller, the power system continuously oscillates after the disturbance. With the damping controller added, by adjusting the reactive power output of the wind turbines, the low-frequency oscillations subsided after 13.2s and 10.1s respectively, indicating that the controller can still effectively suppress oscillations even after the system topology has changed.

[0081] Figure 6 Simulation results for rotor power angle difference, speed difference, and tie-line power under scenario 2 are presented. The results show that, in both disturbances, without a damping controller, the power system experiences continuous oscillations after the disturbance. With the damping controller, the oscillations subsided after 13s and 12s respectively by adjusting the reactive power output of the wind turbine. The results indicate that even when operating conditions change within a certain range, the proposed controller can effectively suppress oscillations under complex disturbances, demonstrating excellent damping performance and robustness. Therefore, the controller designed in this invention can effectively suppress oscillations using the wind farm when system parameters are unknown, the system exhibits a weakly damped oscillation mode, system operating conditions change, and the system is subjected to complex disturbances.

[0082] Please see Figure 7 This application also provides a wind farm damping control system based on sliding diaphragm and predictive control, the system comprising: The data acquisition module is used to acquire input and output measurement data for wind farm operation. The model building module is used to build a partial-format dynamic linearized data model of the wind farm based on the input and output measurement data. The model update module is used to adaptively update the partially formatted dynamically linearized data model using a pseudo-gradient update mechanism. The discrete sliding mode control module is used to calculate the discrete sliding mode control term based on the updated partial scheme dynamic linearized data model. The optimal predictive control module is used to predict the future operating state of the wind farm based on discrete sliding mode control terms, and to obtain the optimal predictive control terms through optimization calculations. The fusion optimization module is used to fuse and optimize the discrete sliding mode control term and the optimal predictive control term to form a damped control output; The output control module is used to inject the damping control output into the reactive power modulation channel of the wind turbine.

[0083] It is understood that the content of the above method embodiments is applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0084] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0085] It is understood that the content of the above method embodiments is applicable to this device embodiment. The specific functions implemented by this device embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0086] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0087] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0088] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0089] It is understood that the content of the above method embodiments is applicable to the embodiments of this program product. The specific functions implemented by the embodiments of this program product are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0090] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0091] This application provides a design method and system for a wind farm damping controller based on sliding mode and predictive control. This scheme constructs a partially formatted dynamic linearized data model using input and output measurement data from the wind farm, achieving an online equivalent description of the nonlinear and strongly coupled dynamic characteristics of the wind farm. This reduces the dependence on precise physical models and fixed parameters, improving the adaptability and practicality of the control model under complex operating conditions. By introducing a pseudo-gradient update mechanism to adaptively correct the model, the control system can track wind speed fluctuations, grid disturbances, and changes in operating status in real time, avoiding performance degradation caused by model mismatch. Regarding the control strategy, this invention combines discrete sliding mode control with predictive control. Sliding mode control's robust suppression of uncertainties and external disturbances ensures rapid system convergence and stable operation; simultaneously, predictive control's ability to predict and optimize future operating states yields optimal control input while meeting constraints, thus balancing control response speed and control smoothness, and reducing chattering problems easily generated in traditional sliding mode control. By fusing and optimizing discrete sliding mode control terms and optimal predictive control terms, and injecting the resulting damping control output into the reactive power modulation channel of the wind turbine, the power oscillations and low-frequency oscillations generated during the grid-connected operation of the wind farm can be effectively suppressed, thereby improving the damping support capability and grid-connected stability of the wind farm and enhancing the overall safety and grid friendliness of the wind farm operation.

[0092] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0093] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0094] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0095] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0096] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A design method for a wind farm damping controller based on sliding diaphragm and predictive control, characterized in that, The method includes the following steps: Acquire input and output measurement data for wind farm operation; Based on the input and output measurement data, a partial scheme dynamic linearized data model of the wind farm is established. A pseudo-gradient update mechanism is used to adaptively update the partially formatted dynamic linearized data model. Based on the updated partial scheme dynamic linearized data model, the discrete sliding mode control term is calculated; Based on the discrete sliding mode control term, the future operating state of the wind farm is predicted, and the optimal predictive control term is obtained through optimization calculation. The discrete sliding mode control term and the optimal predictive control term are fused and optimized to form a damped control output; The damping control output is injected into the reactive power modulation channel of the wind turbine.

2. The method according to claim 1, characterized in that, The step of establishing a partial-format dynamic linearized data model for the wind farm based on the input and output measurement data includes: Based on the input and output measurement data of the wind farm, a discrete input-output data sequence of the wind farm is constructed; The discrete input-output data sequence is subjected to partial scheme dynamic linearization to obtain an equivalent linear data model represented by a pseudo gradient vector. The pseudo gradient vector is introduced into the equivalent linear data model to form a partial scheme dynamic linearized data model of the wind farm.

3. The method according to claim 2, characterized in that, The formula for the partially formatted dynamic linearized data model is as follows: ； Where k represents the current sampling time; Indicates at time Relative to time The change in output; This represents a pseudo-gradient vector; Indicates the increment of the control input; This represents the generalized disturbance increment term.

4. The method according to claim 1, characterized in that, The calculation formula for the pseudo-gradient update mechanism is as follows: ； in, Indicates time The pseudo gradient vector estimate; This represents the initial value of the pseudo-gradient vector; Indicates time The pseudo gradient vector estimate; This represents the change in output between adjacent sampling times; Indicates the increment of the control input; Indicates the threshold constant; This represents the step size factor, which controls the adjustment range of the pseudo-gradient update; Indicates the weighting factor; This indicates the output incremental prediction error term; This represents the transpose of the control input increment.

5. The method according to claim 1, characterized in that, The calculation of the discrete sliding mode control term based on the updated partial scheme dynamic linearized data model includes: Based on the partial scheme dynamic linearized data model, a discrete sliding surface function for wind farm damping control is constructed. Based on the discrete sliding surface function, design a discrete sliding mode reaching law; The discrete sliding mode approach law and the partial scheme dynamic linearization data model are used to calculate the corresponding discrete sliding mode control law. The discrete sliding mode control law is processed into a continuous form to obtain the discrete sliding mode control term.

6. The method according to claim 5, characterized in that, The calculation formula for the discrete sliding mode control term is as follows: ； in, Indicates time Discrete sliding mode control term; Indicates time The pseudo gradient estimate; Represents a small positive constant; Indicates time The actual output feedback signal; Indicates time Discrete sliding mode states; Indicates the constant parameters of the sliding surface; Represents the discrete sliding mode approaching term; Indicates the weighting coefficient of the sliding surface; Indicates time Tracking error; Represents the equivalent disturbance estimate; This indicates the output reference value at the next sampling time. Represents the sliding mode reaching law coefficient; This represents the output value of the symbolic function.

7. The method according to claim 1, characterized in that, The process of predicting the future operating state of a wind farm based on the discrete sliding mode control term, and obtaining the optimal predictive control term through optimization calculation, includes: Based on the discrete sliding mode control term, the sliding mode state corresponding to the wind farm is predicted one step forward, and the multi-step sliding mode state in the prediction time domain is obtained by recursion. The multi-step sliding mode states in the prediction time domain are constructed into a sliding mode state prediction vector; Based on the sliding mode state prediction vector, a performance index function is constructed; The performance index function is optimized to obtain the optimal predictive control term.

8. The method according to claim 7, characterized in that, The formula for calculating the optimal predictive control term is as follows: ； in, Indicates time The optimal predictive control term; This represents the optimal predictive control sequence vector within the current and future prediction time domains; This represents the matrix used to select the control vector; and Represents a time-varying matrix; Representation matrix The transpose of the matrix; Represents a positive definite weighted matrix; Indicates the penalty coefficient; Represents the identity matrix; and Represents a constant matrix; This represents the known historical best predictive control term vector; This represents the vector of discrete sliding mode control terms.

9. The method according to claim 1, characterized in that, The process of fusing and optimizing the discrete sliding mode control term and the optimal predictive control term to form a damped control output includes: Based on preset fusion weight coefficients, the discrete sliding mode control term and the optimal predictive control term are weighted and combined to construct the overall control law expression; Based on the wind farm's operating status and control constraints, the overall control law expression is subjected to amplitude and rate of change constraints. The overall control law expression after constraint processing is smoothed and optimized to obtain the final damped control output.

10. A wind farm damping control system based on sliding diaphragm and predictive control, characterized in that, The system includes: The data acquisition module is used to acquire input and output measurement data for wind farm operation. The model building module is used to establish a partial-format dynamic linearized data model of the wind farm based on the input measurement data and the output measurement data. The model update module is used to adaptively update the partial-format dynamic linearized data model using a pseudo-gradient update mechanism. The discrete sliding mode control module is used to calculate the discrete sliding mode control term based on the updated partial scheme dynamic linearized data model. The optimal predictive control module is used to predict the future operating state of the wind farm based on the discrete sliding mode control term, and to obtain the optimal predictive control term through optimization calculation. The fusion optimization module is used to fuse and optimize the discrete sliding mode control term and the optimal predictive control term to form a damped control output; The output control module is used to inject the damping control output into the reactive power modulation channel of the wind turbine.

Images