A data-driven event-triggered sliding mode control method and system for wind turbines

Abstract

The present application relates to the technical field of wind power generation control, in particular to an event-triggered wind turbine sliding mode control method and system based on data driving, which calculates the error by real-time acquisition of actual and expected aerodynamic torque of the wind turbine, quantifies the error by a quantizer and restricts it to a preset range by a prescribed performance function, constructs a sliding surface function and a torque controller based on the equivalent error to generate a control signal, screens the control signal by an event-triggered mechanism, and iteratively optimizes until the error converges or the preset time is reached to maximize the power generation efficiency. This method has strong robustness under wind speed gradual change / mutation conditions, reduces communication volume through quantization and event-triggered mechanism, and achieves the collaborative optimization of power generation efficiency and communication energy consumption.

Description

Technical Field

[0001] This invention relates to the field of wind power generation control technology, and in particular to a data-driven, event-triggered sliding mode control method and system for wind turbines. Background Technology

[0002] With the ever-increasing global demand for clean energy, wind turbines, as the core equipment of wind power generation, have become an important technological means to alleviate the traditional energy crisis and reduce environmental pressure due to their advanced aerodynamic design and efficient energy conversion capabilities. With significant advantages such as optimizing the energy structure and reducing carbon emissions, wind turbines continue to occupy a leading position in the field of new energy research. However, their deployment in remote mountainous areas has given rise to a series of technical challenges, especially the limited communication resources, which increasingly highlight the contradictions with traditional control technologies.

[0003] At the communication level, traditional control methods do not fully consider the optimal allocation of communication resources, resulting in an excessive burden on signal transmission and making data congestion and packet loss very easy, which seriously threatens the stability of system operation. At the same time, unoptimized control strategies will exacerbate energy loss, which is contrary to the low power consumption and high reliability requirements of equipment in remote areas, thus limiting the practical application efficiency of wind power generation systems.

[0004] In the field of control technology, existing sliding mode control technology has significant shortcomings:

[0005] One is the contradiction between model dependence and environmental adaptability. Traditional sliding mode control based on an initial model often leads to control chattering and decreased accuracy due to the mismatch between the model and the actual working conditions. On the other hand, a single model-free sliding mode control strategy is difficult to balance strong anti-interference capability and precise control in complex and ever-changing environments.

[0006] Secondly, it is difficult to balance robustness and control accuracy. Although sliding mode control has a natural advantage in anti-interference, noise and measurement errors will weaken its performance under a single architecture. Especially when the state of the wind turbine changes rapidly, it is difficult to ensure dynamic response accuracy and system stability at the same time.

[0007] Thirdly, there is a lack of coordination between communication resource optimization and control performance. Existing sliding mode control technology rarely incorporates communication bandwidth limitations for strategy optimization and lacks coordinated design between control algorithms and signal transmission. This results in either low data transmission efficiency affecting control real-time performance or excessive energy consumption in remote areas when used in conjunction with communication quality.

[0008] In summary, the shortcomings of existing technologies in terms of model accuracy, control robustness, and communication resource adaptability urgently need to be overcome through novel sliding mode control strategies that integrate adaptive control, intelligent algorithms, and communication optimization technologies. Summary of the Invention

[0009] Therefore, the technical problem to be solved by the present invention is to overcome the low power generation efficiency caused by the difficulty in obtaining accurate dynamic models of wind turbines in the prior art, and at the same time solve the problem of data transmission burden in scenarios with limited communication resources.

[0010] To address the aforementioned technical problems, this invention provides a data-driven, event-triggered sliding mode control method and system for wind turbines, comprising the following steps:

[0011] Step S1: Obtain the actual aerodynamic torque of the wind turbine at each moment in real time, and calculate the desired aerodynamic torque. Based on the actual aerodynamic torque and the desired aerodynamic torque, obtain the aerodynamic torque error.

[0012] Step S2: Construct a quantizer function based on quantization, quantize the aerodynamic torque error through the quantizer function to obtain the quantized aerodynamic torque error; and establish a specified performance function, based on the specified performance function, and introduce a smoothing function to converge the quantized aerodynamic torque error to a preset range to obtain the equivalent transformed aerodynamic torque error.

[0013] Step S3: Based on the equivalent transformed aerodynamic torque error, construct a sliding mode surface function and construct a torque controller for the wind turbine. Input the quantized aerodynamic torque error at the current moment into the torque controller and output the control signal at the current moment.

[0014] Step S4: Design an event triggering mechanism to optimize the target control signal: Determine whether the control signal at the current moment satisfies the event triggering mechanism.

[0015] If the conditions are met, the current time is set as the current input trigger time, and the control signal at the current time is sent to the actuator as the target control signal at the current input trigger time.

[0016] Otherwise, the target control signal at the previous input trigger moment will be sent to the actuator as the target control signal at the current input trigger moment;

[0017] Step S5: Repeat steps S1 to S4 until the preset running time is reached or the error converges to the threshold, and maximize power generation efficiency through iterative optimization.

[0018] In one embodiment of the present invention, the aerodynamic torque error e(k) is calculated using the following formula:

[0019] e(k) = T opt (k)-T a (k)=y * (k)-t(k)

[0020] Among them, T opt (k)=y * (k) represents the desired aerodynamic torque at the current time k, T a (k) = y(k) represents the actual aerodynamic torque at the current time k. It is the optimal tip speed ratio. It is the maximum power coefficient, ω r (k) is the angular velocity of the wind turbine at time k, ρ is the air density, and R is the blade radius.

[0021] In one embodiment of the present invention, step S2, the method for obtaining the quantified aerodynamic torque error includes:

[0022] Define the quantizer function U based on the quantization as follows:

[0023] U={±z i :z i =θ i z0,i=0,±1,±2,…}∪{0}

[0024] Where θ∈(0,1) is the quantization factor, which determines the exponential distribution characteristics of the quantization interval; i represents the quantization level, corresponding to signal intervals of different amplitudes, and i takes integer values; z0 is the reference value of the initial quantization interval, z0>0, which determines the minimum quantization unit; z i These are the quantized non-zero values;

[0025] Construct the mapping function q e (x), through the mapping function q e (x) Map the aerodynamic torque error e(k) to the quantizer function U to obtain the quantized aerodynamic torque error q. e (e(k)); where the mapping function q e (x) has the property of being symmetric and time-invariant, and is defined as follows:

[0026]

[0027] Where δ represents the parameter used to determine the width boundary of the quantization interval.

[0028] In one embodiment of the present invention, step S2, the method for obtaining the equivalent converted aerodynamic torque error τ(k) includes:

[0029] Construct the specified performance function as follows:

[0030] ρ(k+1)=(1-θ1)ρ(k)+θ1ρ ∞

[0031] Where ρ(k) is the error boundary at time k, ρ(k+1) is the error boundary at time k+1, and ρ(0)>ρ ∞ >0, ρ(0) is the initial error boundary, ρ ∞ Let θ1 be the steady-state error boundary and θ1 be the convergence rate parameter. If θ1∈(0,1), then we have: |e(k)|<ρ(k);

[0032] Based on the specified performance function, the quantized aerodynamic torque error q is smoothed by introducing a smoothing function ln(·). e (e(k)) is mapped to the equivalent transformation error form without boundary constraints through coordinate mapping:

[0033]

[0034] In one embodiment of the present invention, in step S3, based on the equivalent transformed aerodynamic torque error, a sliding mode surface function is constructed as follows:

[0035] s(k+1)=s(k)+τ(k+1)+ατ(k),

[0036] Where s(k) represents the sliding surface state value at time k, s(k+1) represents the sliding surface state value at the next time step of s(k), α is the weighting parameter, τ(k) is the equivalent transformed aerodynamic torque error at time k, and τ(j+1) is the equivalent transformed aerodynamic torque error at time k+1; the following convergence rule is established:

[0037] Δs(j+1)=s(j+1)-s(k).

[0038] In one embodiment of the present invention, step S4, the method for setting the event triggering mechanism includes:

[0039] Define k i To satisfy the input triggering time of the event triggering mechanism, where i represents the number of successful triggers of the event triggering mechanism, design an input triggering time k that satisfies the event triggering mechanism. i as follows:

[0040] or

[0041] Where e(k) is the aerodynamic torque error, To preset the error threshold, Indicates the event triggering error. e(k i ) represents the input trigger time k i aerodynamic torque error, q e (e(k i )) represents e(k iThe quantized error value; e(k) represents the aerodynamic torque error at time k, q e (e(k)) represents the error value after quantization of e(k); M(k) represents the dynamic threshold parameter. ρ1 represents the step size constant. Indicates spurious bias quantity The first element, λ is an estimate of Φ(k); λ represents the weighting constant. Δy * (k+1) represents the expected increment at time k+1, n u It is the linearization length constant of the controlled system, j = 2, 3, ..., n u N(k) represents the estimated value of the noise or disturbance term, and Δu(k-j+1) represents the historical control increment.

[0042] In one embodiment of the present invention, step S4, the method for optimizing the target control signal through the event triggering mechanism includes:

[0043] Define k i To satisfy the input triggering time of the event triggering mechanism, i represents the number of successful triggers of the event triggering mechanism; determine whether the control signal u(k) at the current time k satisfies the event triggering mechanism:

[0044] If satisfied, set the current time k as the current input trigger time k. i That is, k = k i The control signal u(k) at time k is used as the current input trigger time k. i The target control signal is sent to the actuator, wherein the control signal at the current moment is u(k) = u(k) i-1 )+Δu(k i ), Δu(k i ) represents the control increment term, u(k) i-1 ) is k i The control signal at the last trigger moment;

[0045] Otherwise, the target control signal u(k) from the previous input trigger time will be used. i-1 (k) is the current input trigger time. i The target control signal is sent to the actuator, i.e., u(k) = u(k) i-1 ).

[0046] In one embodiment of the present invention, the control increment term Δu(k) i The calculation methods for ) include:

[0047] Estimated by pseudopartial derivative Calculate the control increment term Δu(k) i ),as follows:

[0048] Δu(k i )=M(k i )q e (e(j i ))-N(k i )=u E (k i )+u F (k i ),

[0049] Where M(k) i ) represents the input trigger time k i The dynamic threshold parameter, e(k) i ) represents the input trigger time k i aerodynamic torque error, q e (e(k i )) represents e(k i ) Quantized error value; u E (k i ) represents the error-driven control term. Δu(k-i+1) is the control increment at time k-i+1, where i = 2, 3, ..., n u ,ρ(k i +1) is the trigger time k i The specified performance function value at the next moment, n u λ is the linearization length constant of the controlled system, and λ represents the weighting constant.

[0050] u F (k i ) represents the sliding mode switching control item. It is a step size vector. It is the trigger time k i Estimates of pseudo-partial derivatives;

[0051] N(k i ) is a compensation item. Γ s ζ(k) is a positive definite matrix. i () indicates the time-varying damping ratio;

[0052] The pseudo-partial derivative estimate Update rate:

[0053]

[0054] Where, σ∈R +ζ is the weighting factor, ζ∈(0,2) is the step size constant, Δy(k) represents the state change at time k, and q Δ (Δy(k)) represents the quantization function of the state change, which quantizes the state change Δy(k); ΔU(k-1) is the historical control increment vector, ΔU(k-1) = [Δu(k-1), Δu(k-2), ..., Δu(kn)]. u )] T .

[0055] Based on the same inventive concept, this invention also provides a data-driven, event-triggered wind turbine sliding mode control system, comprising: a data acquisition and error calculation module, a quantization and error constraint module, a sliding mode surface construction and controller generation module, an event trigger judgment and signal update module, and an efficiency optimization module; wherein...

[0056] The data acquisition and error calculation module is configured to: acquire the actual aerodynamic torque of the wind turbine at each moment in real time, calculate the expected aerodynamic torque, and obtain the aerodynamic torque error based on the actual aerodynamic torque and the expected aerodynamic torque;

[0057] The quantization processing and error constraint module is configured to: construct a quantizer function based on quantization, quantize the aerodynamic torque error through the quantizer function to obtain the quantized aerodynamic torque error; and establish a specified performance function, based on the specified performance function, and introduce a smoothing function to converge the quantized aerodynamic torque error to a preset range to obtain the equivalent transformed aerodynamic torque error.

[0058] The sliding surface construction and controller generation module is configured to: construct a sliding surface function based on the equivalent transformed aerodynamic torque error, construct a torque controller for the wind turbine, input the quantized aerodynamic torque error at the current moment into the torque controller, and output the control signal at the current moment;

[0059] The event triggering judgment and signal update module is configured to: design an event triggering mechanism, optimize the target control signal through the event triggering mechanism; determine whether the control signal at the current moment satisfies the event triggering mechanism; if it does, set the current moment as the current input triggering moment, and send the control signal at the current moment as the target control signal at the current input triggering moment to the actuator; otherwise, send the target control signal at the previous input triggering moment as the target control signal at the current input triggering moment to the actuator.

[0060] The efficiency optimization module is configured to: determine whether the preset running time or error converges to a threshold, and maximize power generation efficiency through iterative optimization.

[0061] The present invention also provides an electronic device comprising a processor, a memory, and a bus system, wherein the processor and the memory are connected via the bus system, the memory is used to store instructions, and the processor is used to execute the instructions stored in the memory to implement the data-driven event-triggered wind turbine sliding mode control method described above.

[0062] The technical solution of the present invention has the following advantages compared with the prior art:

[0063] This invention, based on a data-driven and event-triggered mechanism, integrates quantization, predefined performance control, and sliding mode control techniques, offering multi-dimensional advantages: It overcomes dependence on precise models through data-driven adaptive updates of pseudo-partial derivatives, adapting to the strong nonlinearity and time-varying characteristics of wind turbines; it optimizes communication resources through both the quantizer and the event-triggered mechanism, reducing data transmission volume and frequency, making it suitable for limited bandwidth scenarios in remote mountainous areas; and it dynamically constrains errors within a preset range using a predefined performance function combined with sliding mode surface design, ensuring control accuracy and convergence speed. Experimental verification shows that this method exhibits strong robustness under both gradual and sudden wind speed conditions, with small tracking errors and adaptive triggering intervals, achieving synergistic optimization of maximizing power generation efficiency and minimizing communication energy consumption. Attached Figure Description

[0064] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings, wherein...

[0065] Figure 1 This is a flowchart illustrating a data-driven, event-triggered sliding mode control method for wind turbines provided in an embodiment of the present invention.

[0066] Figure 2 This is a schematic diagram of the event triggering system principle provided in the embodiments of the present invention;

[0067] Figure 3 These are the effect diagrams before and after error signal quantization provided in the embodiments of the present invention;

[0068] Figure 4 These are diagrams showing the effect of output incremental signal quantization before and after in the embodiments of the present invention;

[0069] Figure 5 This is the tracking performance of the system for slow wind speed changes provided in the embodiments of the present invention;

[0070] Figure 6 This refers to the specified performance control provided in the embodiments of the present invention;

[0071] Figure 7 This is an event triggering interval diagram provided in the embodiments of the present invention;

[0072] Figure 8 The wind shear system tracking performance provided in the embodiments of the present invention;

[0073] Figure 9 This is a diagram showing the wind speed shear event triggering interval provided in an embodiment of the present invention;

[0074] Figure 10 This is a schematic diagram of a data-driven, event-triggered wind turbine sliding mode control system provided in an embodiment of the present invention.

[0075] Explanation of reference numerals in the accompanying drawings: 100, Data acquisition and error calculation module; 200, Quantization processing and error constraint module; 300, Sliding surface construction and controller generation module; 400, Event trigger judgment and signal update module; 500, Efficiency optimization module. Detailed Implementation

[0076] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention.

[0077] Example 1:

[0078] See Figure 1 and Figure 2 This invention provides a data-driven, event-triggered sliding mode control method and system for wind turbines, the method comprising the following steps:

[0079] Step S1: Obtain the actual aerodynamic torque of the wind turbine at each moment in real time, and calculate the desired aerodynamic torque. Based on the actual aerodynamic torque and the desired aerodynamic torque, obtain the aerodynamic torque error.

[0080] Step S2: Construct a quantizer function based on quantization, quantize the aerodynamic torque error through the quantizer function to obtain the quantized aerodynamic torque error; and establish a specified performance function, based on the specified performance function, and introduce a smoothing function to converge the quantized aerodynamic torque error to a preset range to obtain the equivalent transformed aerodynamic torque error.

[0081] Step S3: Based on the equivalent transformed aerodynamic torque error, construct a sliding mode surface function and construct a torque controller for the wind turbine. Input the quantized aerodynamic torque error at the current moment into the torque controller and output the control signal at the current moment.

[0082] Step S4: Design an event triggering mechanism to optimize the target control signal: Determine whether the control signal at the current moment satisfies the event triggering mechanism.

[0083] If the conditions are met, the current time is set as the current input trigger time, and the control signal at the current time is sent to the actuator as the target control signal at the current input trigger time.

[0084] Otherwise, the target control signal at the previous input trigger moment will be sent to the actuator as the target control signal at the current input trigger moment;

[0085] Step S5: Repeat steps S1 to S4 until the preset running time is reached or the error converges to the threshold, and maximize power generation efficiency through iterative optimization.

[0086] Furthermore, in step S1, the actual aerodynamic torque T of the wind turbine at the current time k is collected in real time. a (k), based on the optimal tip speed ratio Calculate the expected aerodynamic torque T at the current time k using the maximum power coefficient. opt (k): Where, ω r (k) is the angular velocity of the wind turbine at time k, ρ is the air density, and R is the blade radius.

[0087] Based on the actual aerodynamic torque T a (k) and the desired aerodynamic torque T opt (k), calculate the aerodynamic torque error e(k):

[0088] e(k) = T opt (k)-T a (k)=y * (k)-y(k)

[0089] Among them, T opt (k)=y * (k) represents the desired aerodynamic torque at the current time k, T a (k)=y(k) represents the actual aerodynamic torque at the current time k.

[0090] Further, in this embodiment, step S2, the method for constructing a quantizer function based on quantization, and quantizing the aerodynamic torque error using the quantizer function to obtain the quantized aerodynamic torque error, includes:

[0091] Define the quantizer function U based on the quantization as follows:

[0092] U={±z i :z i =θi z0,i=0,±1,±2,…}∪{0}

[0093] Where θ∈(0,1) is the quantization factor, which determines the exponential distribution characteristics of the quantization interval; i represents the quantization level, corresponding to signal intervals of different amplitudes, and i takes integer values; z0 is the reference value of the initial quantization interval, z0>0, which determines the minimum quantization unit; z i These are the quantized non-zero values;

[0094] Construct the mapping function q e (x), through the mapping function q e (x) Map the aerodynamic torque error e(k) to the quantizer function U to obtain the quantized aerodynamic torque error q. e (e(k)); where the mapping function q e (x) has the property of being symmetric and time-invariant, and is defined as follows:

[0095]

[0096] Where δ represents the parameter used to determine the width boundary of the quantization interval.

[0097] Furthermore, in this embodiment, the method for obtaining the equivalent converted aerodynamic torque error τ(k) in step S2 includes:

[0098] Construct the specified performance function as follows:

[0099] ρ(k+1)=(1-θ1)ρ(k)+θ1ρ ∞

[0100] Where ρ(k) is the error boundary at time k, ρ(k+1) is the error boundary at time k+1, and ρ(0)>ρ ∞ >0, ρ(0) is the initial error boundary, ρ ∞ Let θ1 be the steady-state error boundary and θ1 be the convergence rate parameter. If θ1∈(0,1), then we have: |e(k)|<ρ(k);

[0101] Based on the specified performance function, the quantized aerodynamic torque error q is smoothed by introducing a smoothing function ln(·). e (e(k)) is mapped to the equivalent transformation error form without boundary constraints through coordinate mapping:

[0102]

[0103] Furthermore, in this embodiment, in step S3, based on the equivalent transformed aerodynamic torque error, a sliding mode surface function is constructed as follows:

[0104] s(k+1)=s(k)+τ(k+1)+ατ(k),

[0105] Where s(k) represents the sliding surface state value at time k, s(k+1) represents the sliding surface state value at the next time step of s(k), α is the weighting parameter, τ(k) is the equivalent transformed aerodynamic torque error at time k, and τ(k+1) is the equivalent transformed aerodynamic torque error at time k+1; the following convergence rule is established:

[0106] Δs(k+1)=s(k+1)-s(k).

[0107] Furthermore, in this embodiment, the method for setting the event triggering mechanism in step S4 includes:

[0108] Define k i To satisfy the input triggering time of the event triggering mechanism, where i represents the number of successful triggers of the event triggering mechanism, design an input triggering time k that satisfies the event triggering mechanism. i as follows:

[0109] or

[0110] Where e(k) is the aerodynamic torque error, To preset the error threshold, Indicates the event triggering error. e(k i ) represents the input trigger time k i aerodynamic torque error, q e (e(k i )) represents e(k i The quantized error value; e(k) represents the aerodynamic torque error at time k, q e (e(k)) represents the error value after quantization of e(k); M(k) represents the dynamic threshold parameter. ρ1 represents the step size constant. Indicates spurious bias quantity The first element, λ is an estimate of Φ(k); λ represents the weighting constant.

[0111] Δy * (k+1) represents the expected increment at time k+1, n u It is the linearization length constant of the controlled system, j = 2, 3, ..., n u N(k) represents the estimated value of the noise or disturbance term, and Δu(k-j+1) represents the historical control increment.

[0112] Furthermore, in this embodiment, step S4, the method for optimizing the target control signal through the event triggering mechanism includes:

[0113] Define k i To meet the input triggering time of the event triggering mechanism, i represents the number of successful triggers of the event triggering mechanism; the formula for updating the target control signal according to the event triggering mechanism is as follows:

[0114]

[0115] Specifically, it determines whether the control signal u(k) at time k satisfies the event triggering mechanism:

[0116] If satisfied, set the current time k as the current input trigger time k. i That is, k = k i The control signal u(k) at time k is used as the current input trigger time k. i The target control signal is sent to the actuator, wherein the control signal at the current moment is u(k) = u(k) i-1 )+Δu(k i ), Δu(k i ) represents the control increment term, u(k) i-1 ) is k i The control signal at the last trigger moment;

[0117] Otherwise, the target control signal u(k) from the previous input trigger time will be used. i-1 (k) is the current input trigger time. i The target control signal is sent to the actuator, i.e., u(k) = u(k) i-1 ).

[0118] In one embodiment of the present invention, the control increment term Δu(k) i The calculation methods for ) include:

[0119] Estimated by pseudopartial derivative Calculate the control increment term Δu(k) i ),as follows:

[0120] Δu(k i )=M(k i )q e (e(k i ))-N(k i )=u E (k i )+u F (k i ),

[0121] Where M(k) i) represents the input trigger time k i The dynamic threshold parameter, e(k) i ) represents the input trigger time k i aerodynamic torque error, q e (e(k i )) represents e(k i ) Quantized error value; u E (k i ) represents the error-driven control term. Δu(k-i+1) is the control increment at time k-i+1, where i = 2, 3, ..., n u ,ρ(k i +1) is the trigger time k i The specified performance function value at the next moment, n u λ is the linearization length constant of the controlled system, and λ represents the weighting constant.

[0122] u F (k i ) represents the sliding mode switching control item. It is a step size vector. It is the trigger time k i Estimates of pseudo-partial derivatives;

[0123] N(k i ) is a compensation item. Γ s ζ(k) is a positive definite matrix. i () indicates the time-varying damping ratio;

[0124] The pseudo-partial derivative estimate Update rate:

[0125]

[0126] Where, σ∈R + As a weighting factor, ζ∈(0,2) is a step size constant, Δy(k) represents the state change at time k, and q Δ (Δy(k)) represents the quantization function of the state change, which quantizes the increment of the state change Δy(k); ΔU(k-1) is the historical control increment vector, ΔU(k-1) = [Δu(k-1), Δu(k-2), ..., Δu(kn)]. u )] T To improve the accuracy of the estimate in the above equation, the reset law is designed as follows:

[0127] when or hour, yes The initial value, and This is the initial error threshold.

[0128] To verify the effectiveness of the method described in this invention, a wind power generation system operation model was used as the research object. The parameters of the wind turbine model were defined as follows: air density ρ = 1.225 kg / m³. 3 Blade radius R = 63m, blade pitch angle β = 0, optimal tip speed ratio Maximum power coefficient Moment of inertia J = 2.6 × 10 6 kg.

[0129] The torque controller parameters are set as follows: damping ratio ζ = 1.6, gain coefficient... Initial error threshold Adaptive weighting parameter α = 0.2, smoothing factor λ = 10 4 Positive definite matrix Γ s =10 5 I, Initial estimate of pseudopartial derivative The performance function parameters are specified as follows: quantization factor θ = 0.1, initial error boundary ρ(0) = 3 × 10⁻⁶. 6 Steady-state error boundary ρ ∞ =5×10 4 Under gradually varying wind speed conditions, the design wind speed model is V(t) = V r +0.6sin(0.1t), where the average wind speed V r =10m / s, select quantization density θ for the quantizer Δ =θ e =0.9, reference quantization interval z0=10 6 .

[0130] Figure 3 and Figure 4 The comparison of the signal e(k) and Δy(k) before and after quantization is shown respectively. The results indicate that the quantized signal exhibits significant multi-level characteristics, and the non-uniform mapping mechanism of the quantizer effectively reduces the amount of communication data transmission. (Refer to...) Figure 5 Simulation results show that, compared to existing control methods, the method described in this invention exhibits stronger robustness under wind speed fluctuation conditions and achieves high-precision tracking of the desired torque. The simulation data also demonstrates that, when an appropriate quantization density parameter is selected, the quantization control algorithm reduces the communication bandwidth burden while having a negligible impact on the system's tracking performance. Figure 6 Experimental results show that the tracking error is strictly constrained within the predefined boundary range of the performance function. Figure 7The event triggering interval diagram shown indicates that this method achieves optimized scheduling of communication resources through a dynamic triggering mechanism, and has significant advantages such as fast convergence speed, strong anti-interference ability, and small tracking error.

[0131] Based on the above experimental conditions, and considering the rapidly changing wind speed, a wind speed model was designed as V(t) = V r +0.8sin(0.1t), where the average wind speed V r The speed jumps from 8 m / s to 9 m / s at 500 s. The quantizer is then adjusted to a quantization density θ. Δ =θ e =0.8, reference quantization interval z0=10 6 .

[0132] Reference Figure 8 This method still achieves rapid tracking of the optimal torque under wind speed shear conditions; Figure 9 The event trigger interval diagram shows that the system can still maintain high-precision torque control under strong interference conditions, which verifies the excellent handling capability and control accuracy of the method of the present invention for unknown interference.

[0133] Example 2:

[0134] Based on the same inventive concept as Embodiment 1, this invention also provides a data-driven event-triggered wind turbine sliding mode control system, used to implement the steps of the data-driven event-triggered wind turbine sliding mode control method described in Embodiment 1. Figure 10 As shown, the data-driven, event-triggered wind turbine sliding mode control system includes: a data acquisition and error calculation module 100, a quantization processing and error constraint module 200, a sliding mode surface construction and controller generation module 300, an event trigger judgment and signal update module 400, and an efficiency optimization module 500; wherein,

[0135] The data acquisition and error calculation module 100 is configured to: acquire the actual aerodynamic torque of the wind turbine at each moment in real time, calculate the expected aerodynamic torque, and obtain the aerodynamic torque error based on the actual aerodynamic torque and the expected aerodynamic torque.

[0136] The quantization processing and error constraint module 200 is configured to: construct a quantizer function based on quantization, quantize the aerodynamic torque error through the quantizer function to obtain the quantized aerodynamic torque error; and establish a specified performance function, based on the specified performance function, and introduce a smoothing function to converge the quantized aerodynamic torque error to a preset range to obtain the equivalent transformed aerodynamic torque error.

[0137] The sliding surface construction and controller generation module 300 is configured to: construct a sliding surface function based on the equivalent transformed aerodynamic torque error, construct a torque controller for the wind turbine, input the quantized aerodynamic torque error at the current moment into the torque controller, and output the control signal at the current moment;

[0138] The event triggering judgment and signal update module 400 is configured to: design an event triggering mechanism, optimize the target control signal through the event triggering mechanism; determine whether the control signal at the current moment satisfies the event triggering mechanism; if it does, set the current moment as the current input triggering moment, and send the control signal at the current moment as the target control signal at the current input triggering moment to the actuator; otherwise, send the target control signal at the previous input triggering moment as the target control signal at the current input triggering moment to the actuator.

[0139] The efficiency optimization module 500 is configured to: determine whether the preset running time or error converges to a threshold, and maximize power generation efficiency through iterative optimization.

[0140] This embodiment proposes a data-driven event-triggered wind turbine sliding mode control system to implement the aforementioned data-driven event-triggered wind turbine sliding mode control method. Therefore, the specific implementation of the data-driven event-triggered wind turbine sliding mode control system can be found in the embodiment section of the aforementioned data-driven event-triggered wind turbine sliding mode control method. For example, the data acquisition and error calculation module 100, the quantization processing and error constraint module 200, the sliding mode surface construction and controller generation module 300, the event trigger judgment and signal update module 400, and the efficiency optimization module 500 are respectively used to implement steps S1, S2, S3, S4, and S5 in the data-driven event-triggered wind turbine sliding mode control method described in Embodiment 1. Therefore, its specific implementation can be referred to the description of the corresponding embodiments. To avoid redundancy, it will not be repeated here.

[0141] Example 3:

[0142] The present invention also provides an electronic device, which includes a processor, a memory, and a bus system. The processor and the memory are connected through the bus system. The memory is used to store instructions, and the processor is used to execute the instructions stored in the memory to implement the data-driven event-triggered wind turbine sliding mode control method described in Embodiment 1.

[0143] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0144] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0145] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0146] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0147] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A data-driven, event-triggered sliding mode control method for wind turbines, characterized in that, Includes the following steps: Step S1: Obtain the actual aerodynamic torque of the wind turbine at each moment in real time, and calculate the desired aerodynamic torque. Based on the actual aerodynamic torque and the desired aerodynamic torque, obtain the aerodynamic torque error. Step S2: Construct a quantizer function based on quantization, and quantize the aerodynamic torque error through the quantizer function to obtain the quantized aerodynamic torque error; A specified performance function is established, and based on the specified performance function, a smoothing function is introduced to converge the quantized aerodynamic torque error to a preset range, thereby obtaining the equivalent converted aerodynamic torque error. Step S3: Based on the equivalent transformed aerodynamic torque error, construct a sliding mode surface function and construct a torque controller for the wind turbine. Input the quantized aerodynamic torque error at the current moment into the torque controller and output the control signal at the current moment. Step S4: Design an event triggering mechanism to optimize the target control signal: Determine whether the control signal at the current moment satisfies the event triggering mechanism. If the conditions are met, the current time is set as the current input trigger time, and the control signal at the current time is sent to the actuator as the target control signal at the current input trigger time. Otherwise, the target control signal at the previous input trigger moment will be sent to the actuator as the target control signal at the current input trigger moment; The method for setting the event triggering mechanism includes: definition To meet the input triggering time of the event triggering mechanism, Indicate the number of successful triggers of the event triggering mechanism, and design the input trigger time that satisfies the event triggering mechanism. as follows: ； in, This is for aerodynamic torque error. To preset the error threshold, Indicates the event triggering error. , Indicates the input trigger time aerodynamic torque error. express Quantized error value; Indicates time aerodynamic torque error. express Quantized error value; Indicates the dynamic threshold parameter. , Represents the step size constant. Indicates spurious bias quantity The first element, , yes The estimated value; Represents the weighting constant; , This represents the expected increment at time k+1. It is the linearized length constant of the controlled system. , This represents the estimated value of the noise or interference term. Indicates the historical control increment; Step S5: Repeat steps S1 to S4 until the preset running time is reached or the error converges to the threshold, and maximize power generation efficiency through iterative optimization.

2. The data-driven, event-triggered sliding mode control method for wind turbines according to claim 1, characterized in that, The aerodynamic torque error The calculation formula is as follows: ； in, Indicates the current time The desired aerodynamic torque Indicates the current time The actual aerodynamic torque, , It is the optimal tip speed ratio. It is the maximum power coefficient. yes The angular velocity of the wind turbine at any given moment. air density, Where is the blade radius.

3. The data-driven, event-triggered sliding mode control method for wind turbines according to claim 1, characterized in that, In step S2, the method for obtaining the quantified aerodynamic torque error includes: Define a quantizer function based on quantization. ,as follows: ； in, As a quantization factor, it determines the exponential distribution characteristics of the quantization interval; This indicates the quantization level, corresponding to different signal amplitude ranges. Take the integer part; This serves as the baseline value for the initial quantization interval. This determines the smallest unit of quantization; These are the quantized non-zero values; Construct mapping function Through the mapping function The aerodynamic torque error Mapped to the quantizer function In the process, the quantified aerodynamic torque error is obtained. ; where, mapping function It possesses symmetric and time-invariant properties, and is defined as follows: ； in, This represents the parameter used to determine the width boundary of the quantization interval. .

4. The data-driven, event-triggered sliding mode control method for wind turbines according to claim 1, characterized in that, In step S2, the equivalent transformed aerodynamic torque error is obtained. The methods include: Construct the specified performance function as follows: ； in, Let k be the error boundary. The error boundary at time k+1, , As the initial error boundary, For steady-state error boundary, For the convergence rate parameter, Then we have: ; Based on the aforementioned performance function, a smoothing function is introduced. The quantified aerodynamic torque error The coordinate mapping is transformed into an equivalent transformation error form without boundary constraints: 。 5. The data-driven, event-triggered sliding mode control method for wind turbines according to claim 4, characterized in that, In step S3, based on the equivalent transformed aerodynamic torque error, the sliding mode surface function is constructed as follows: ， in, This represents the sliding surface state value at time k. express The sliding surface state value at the next moment. For weight parameters, Let be the equivalent transformed aerodynamic torque error at time k. Let the equivalent aerodynamic torque error at time k+1 be the value; establish the following convergence rule: 。 6. The data-driven, event-triggered sliding mode control method for wind turbines according to claim 1, characterized in that, In step S4, the method for optimizing the target control signal through the event triggering mechanism includes: definition To meet the input triggering time of the event triggering mechanism, Indicates the number of successful triggers of the event triggering mechanism; determines the control signal at the current time k. Does the event triggering mechanism meet the requirements? If satisfied, set the current time k as the current input trigger time. ,Right now = The control signal at time k. As the current input trigger moment The target control signal is sent to the actuator, wherein the control signal at the current moment is... = , To control incremental terms, for The control signal at the last trigger moment; Otherwise, the target control signal from the previous input trigger time will be used. As the current input trigger moment The target control signal is sent to the actuator, that is... = .

7. The data-driven, event-triggered sliding mode control method for wind turbines according to claim 6, characterized in that, The control increment item The calculation methods include: Estimated by pseudopartial derivative Calculate the control increment term ,as follows: ， in, Indicates the input trigger time The dynamic threshold parameter, Indicates the input trigger time aerodynamic torque error. express Quantized error value; For error-driven control terms, , It is a moment One step to control the increment, , It is the triggering time. The specified performance function value at the next moment. It is the linearized length constant of the controlled system. Represents the weighting constant; For sliding mode switching control item, , It is a step size vector. It is the triggering time. Estimates of pseudo-partial derivatives; As compensation item, , It is a positive definite matrix. Indicates the time-varying damping ratio; The pseudo-partial derivative estimate update rate : ； in, As a weighting factor, The step size is a constant. This represents the state change at time k. The quantization function represents the change in state, and the change in state... Quantification processing is performed; For historical control increment vectors, .

8. A data-driven, event-triggered sliding mode control system for wind turbines, characterized in that, The data-driven, event-triggered wind turbine sliding mode control system includes the following modules: The data acquisition and error calculation module is configured to: acquire the actual aerodynamic torque of the wind turbine at each moment in real time, calculate the expected aerodynamic torque, and obtain the aerodynamic torque error based on the actual aerodynamic torque and the expected aerodynamic torque; The quantization processing and error constraint module is configured to: construct a quantizer function based on quantization, quantize the aerodynamic torque error through the quantizer function, and obtain the quantized aerodynamic torque error; A specified performance function is established, and based on the specified performance function, a smoothing function is introduced to converge the quantized aerodynamic torque error to a preset range, thereby obtaining the equivalent converted aerodynamic torque error. The sliding surface construction and controller generation module is configured to: construct a sliding surface function based on the equivalent transformed aerodynamic torque error, construct a torque controller for the wind turbine, input the quantized aerodynamic torque error at the current moment into the torque controller, and output the control signal at the current moment; The event trigger judgment and signal update module is configured to: design an event trigger mechanism, optimize the target control signal through the event trigger mechanism; determine whether the control signal at the current moment satisfies the event trigger mechanism; if it does, set the current moment as the current input trigger moment, and send the control signal at the current moment as the target control signal at the current input trigger moment to the actuator; Otherwise, the target control signal at the previous input trigger moment will be sent to the actuator as the target control signal at the current input trigger moment; The method for setting the event triggering mechanism includes: definition To meet the input triggering time of the event triggering mechanism, Indicate the number of successful triggers of the event triggering mechanism, and design the input trigger time that satisfies the event triggering mechanism. as follows: ； in, This is for aerodynamic torque error. To preset the error threshold, Indicates the event triggering error. , Indicates the input trigger time aerodynamic torque error. express Quantized error value; Indicates time aerodynamic torque error. express Quantized error value; Indicates the dynamic threshold parameter. , Represents the step size constant. Indicates spurious bias quantity The first element, , yes The estimated value; Represents the weighting constant; , This represents the expected increment at time k+1. It is the linearized length constant of the controlled system. , This represents the estimated value of the noise or interference term. Indicates the historical control increment; The efficiency optimization module is configured to determine whether the preset running time or error converges to a threshold, and maximize power generation efficiency through iterative optimization.

9. An electronic device, characterized in that, The electronic device includes a processor, a memory, and a bus system. The processor and the memory are connected through the bus system. The memory is used to store instructions, and the processor is used to execute the instructions stored in the memory to implement the data-driven event-triggered wind turbine sliding mode control method according to any one of claims 1 to 7.

Images