Stall flutter closed-loop control method based on skin optimal phase vibration

Abstract

The present invention discloses a stall flutter closed-loop control method based on skin optimal phase vibration, and belongs to the technical field of aeroelastic control. The present invention calculates the instantaneous amplitude, frequency and phase according to the pitch angle time domain response of the stall flutter system, and then obtains the optimal phase difference between the skin active vibration and the airfoil pitch angle according to the instantaneous amplitude and frequency, superimposes the optimal phase difference with the instantaneous phase of the airfoil pitch angle as the phase of the skin active vibration, calculates the skin active vibration displacement, drives the skin deformation driver to generate corresponding displacement, and exerts control on the stall flutter system. The present invention can significantly improve the stall flutter mitigation control effect; it solves the defect that the existing method of obtaining the instantaneous amplitude, frequency and phase of the signal cannot be applied to the stall flutter closed-loop control based on the skin optimal phase vibration.

Description

Technical Field

[0001] The invention belongs to the technical field of aeroelasticity control, and relates to an aeroelasticity closed-loop control method, and in particular to a stall flutter closed-loop control method based on skin optimal phase vibration. Background Art

[0002] Stall flutter is an aeroelastic instability phenomenon closely related to flow separation, which manifests as a large pitch single-degree-of-freedom limit cycle oscillation of the airfoil. It is widely found in large flexible wings of high-altitude and long-endurance aircraft, helicopter rotors, and wind turbine blades. Stall flutter is highly destructive and seriously affects the performance and structural safety of related equipment. Therefore, it is crucial to use appropriate control methods to mitigate stall flutter response.

[0003] Skin active vibration control uses the active vibration of the local skin to disturb the boundary layer, change the dynamic characteristics of vortex shedding in the stall flutter flow field, and thus achieve the purpose of mitigating stall flutter. The phase difference between the skin active vibration and the stall flutter pitch angle is a key parameter that affects the mitigation control effect. After the control is applied, the stall flutter pitch angle oscillation amplitude and frequency will change, and the corresponding optimal phase difference will also change. How to establish the corresponding relationship between the stall flutter response characteristics (pitch angle oscillation amplitude and frequency) and the optimal phase difference is a key issue to improve the skin active vibration stall flutter mitigation control effect.

[0004] After obtaining the corresponding relationship between the stall flutter pitch angle oscillation amplitude, frequency and the optimal phase difference, when performing stall flutter control, it is first necessary to obtain the instantaneous oscillation amplitude and frequency of the pitch angle in real time, and obtain the optimal phase difference at the current moment according to the determined relationship between the stall flutter pitch angle oscillation amplitude, frequency and the optimal phase difference, and then obtain the instantaneous phase at the current moment, superimpose the optimal phase difference, and use the sum of the two as the phase of the skin active vibration drive signal at the current moment, and drive the skin vibration to slow down the stall flutter. Therefore, how to obtain the instantaneous oscillation amplitude, frequency and phase of the pitch angle at the current moment in real time is another key issue in the stall flutter closed-loop control based on the skin optimal phase vibration. There are two main methods for obtaining instantaneous amplitude, frequency and phase: one is short-time Fourier transform. Since the stall flutter response frequency is low and equivalent to the frequency resolution, it is impossible to accurately distinguish the change of instantaneous frequency; the other is the decomposition method based on variational mode decomposition. However, this method requires that the time derivative of the amplitude is much smaller than the frequency. Since the amplitude will be significantly attenuated after the control is applied, and the response frequency of the stall flutter is small, this condition cannot be met. The existing methods for obtaining the instantaneous amplitude, frequency and phase of the signal are not applicable to the stall flutter closed-loop control based on the optimal phase vibration of the skin. Summary of the invention

[0005] Aiming at the existing problems of stall flutter closed-loop control based on optimal phase vibration of skin, including how to establish the corresponding relationship between stall flutter response characteristics (pitch angle oscillation amplitude and frequency) and optimal phase difference, and how to obtain the instantaneous amplitude, frequency and phase of stall flutter pitch angle response at the current moment in real time. The present invention proposes a stall flutter closed-loop control method based on optimal phase vibration of skin, by developing an instantaneous amplitude, frequency and phase calculation method for stall flutter response, establishing the corresponding relationship between stall flutter response characteristics (instantaneous oscillation amplitude and frequency) and the optimal phase difference of skin active vibration, giving the optimal phase of skin active vibration according to the instantaneous amplitude and frequency at the current moment, driving the skin active vibration deformation, and realizing stall flutter closed-loop mitigation control.

[0006] The stall flutter closed-loop control method of the present invention is based on the optimal phase vibration of the skin, and the control process is as follows:

[0007] Step 1: Given the control start time t c , the amount of data used for fitting N m , filter parameter μ, filter sampling interval N F , calculate the total time t To .

[0008] Step 2: Perform fast Fourier transform on the stall flutter pitch angle response without skin active vibration control to obtain the peak frequency f peak , the time domain response function form of the stall flutter pitch angle to be identified is expressed as:

[0009]

[0010] Step 3: If the current time t is greater than the total calculation time t To , then the calculation ends; otherwise, calculate the stall flutter pitch oscillation response at the current moment and collect the pitch angle signal α at the current moment n Put it into the pitch angle time series α.

[0011] Step 4: If the number of data in the pitch angle time series α is N α Less than the amount of data used for fitting N m , then directly enter the loop of the next time step, otherwise proceed to step 5.

[0012] Step 5: Fit the pitch oscillation amplitude A at the current moment αm , Phase Average value α 0m , put into the corresponding time series A αm , α 0m middle.

[0013] Step 6: If the current time t is less than the control start time tc , then directly enter the loop of the next time step and return to step 3; otherwise, proceed to step 7.

[0014] Step 7: From the current moment, from time series A αm , α 0m Interval N F Sample forward until the index is less than 1 to get the sampling sequence

[0015] Step 8: Get the sampling sequence The data length N s , construct the filter function Φ.

[0016] Step 9: Use the filter function Φ to filter the sampling sequence Perform filtering.

[0017] Step 10: Using the pre-built proxy model, combined with the filtered amplitude, frequency and phase at the current moment obtained in step 9, the optimal phase difference corresponding to the instantaneous amplitude and frequency of the pitch oscillation at the current moment is obtained. Then the phase of active vibration of the skin is The active vibration displacement of the skin applied at the current moment is Then enter the next time step loop; in the above formula, is the phase at the current moment after filtering.

[0018] The advantages of the present invention are:

[0019] 1. The stall flutter closed-loop control method based on the optimal phase vibration of the skin of the present invention establishes a method for determining the corresponding relationship between the stall flutter response characteristics (pitch angle oscillation amplitude and frequency) and the optimal phase difference, which can significantly improve the stall flutter mitigation control effect;

[0020] 2. The stall flutter closed-loop control method based on the optimal phase vibration of the skin of the present invention develops a method for calculating the instantaneous amplitude, frequency and phase of the stall flutter response, which solves the defect that the existing method for obtaining the instantaneous amplitude, frequency and phase of the signal cannot be applied to the stall flutter closed-loop control based on the optimal phase vibration of the skin. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of stall flutter mitigation control based on active skin vibration of the present invention;

[0022] Figure 2 It is a framework diagram of the stall flutter closed-loop control method based on skin optimal phase vibration of the present invention;

[0023] Figure 3The flowchart of the method for calculating the optimal phase difference of skin active vibration based on the transient response characteristics of stall flutter is as follows;

[0024] Figure 4 It is a flow chart of the stall flutter closed-loop control method based on skin optimal phase vibration of the present invention;

[0025] Figure 5a The instantaneous amplitude is calculated based on the pitch angle time domain response of the stall flutter system;

[0026] Figure 5b The instantaneous phase is calculated based on the pitch angle time domain response of the stall flutter system;

[0027] Figure 5c To compare the reconstructed signal based on the transient response characteristics with the real signal;

[0028] Figure 6 This is the stall flutter closed-loop control effect based on the optimal phase vibration of the skin. DETAILED DESCRIPTION

[0029] like Figure 1 As shown in the figure, for a two-dimensional airfoil model with the pitching freedom released, the structural damping coefficient of the airfoil pitching oscillation is D θ , the stiffness coefficient is K θ , under the action of the incoming airflow, when the critical stall flutter speed is exceeded, the airfoil will undergo a large pitch single-degree-of-freedom limit cycle oscillation, namely stall flutter.

[0030] The stall flutter closed-loop control method of the present invention is based on the optimal phase vibration of the skin. Excitation is applied to the local areas of the skin on the upper and lower surfaces of the airfoil so that it can actively vibrate and deform according to a given signal, thereby disturbing the flow field boundary layer and achieving stall flutter mitigation control.

[0031] The functional form of skin active vibration is:

[0032]

[0033] in, f α Take the stall flutter pitch angle peak frequency f without skin active vibration control peak , then the key to stall flutter closed-loop control is to determine the phase of the skin active vibration In the present invention, Figure 2 As shown in the figure, the instantaneous amplitude, frequency and phase are calculated based on the pitch angle time domain response of the stall flutter system. Then, the optimal phase difference between the active vibration of the skin and the pitch angle of the airfoil is obtained based on the instantaneous amplitude and frequency, and the optimal phase difference is superimposed with the instantaneous phase of the pitch angle of the airfoil as the phase of the active vibration of the skin. The active vibration displacement of the skin is calculated, and the skin deformation actuator is driven to produce the corresponding displacement to control the stall flutter system.

[0034] The specific process of the stall flutter closed-loop control method based on the optimal phase vibration of the skin of the present invention is described below in conjunction with the accompanying drawings:

[0035] First, a proxy model between the aerodynamic torque periodic work and three given motion parameters needs to be constructed, which is used to be called during the execution of the method of the present invention to calculate the optimal phase difference of the skin active vibration based on the stall flutter transient response characteristics, such as Figure 3 As shown, the specific construction method of the model is as follows:

[0036] 1) Given the amplitude of pitch oscillation A α , frequency f α Range, skin active vibration phase difference Range (0-2π).

[0037] 2) Perform Latin hypercube sampling in the parameter space determined by the three parameter ranges given in step 1) to obtain a parameter sampling set.

[0038] 3) For each parameter combination (A α ,f α , ) performs CFD calculations for a given motion and obtains the aerodynamic torque time domain data M(t).

[0039] The given motion includes:

[0040] The given signal of the airfoil pitch oscillation motion is:

[0041] α=A α sin(2πf α t)

[0042] Where, t is time;

[0043] The upper surface skin actively vibrates, and its given signal is:

[0044]

[0045] The active vibration displacement of the lower surface skin is the same as that of the upper surface, but with opposite signs. The given signal is:

[0046] In the above formula, is the chord-wise coordinate x, x1 is the chord-wise coordinate of the left endpoint of the skin active vibration area, and x2 is the chord-wise coordinate of the right endpoint of the skin active vibration area.

[0047] 4) According to the obtained aerodynamic torque time domain data M(t), the aerodynamic torque period work is calculated as:

[0048]

[0049] Where, T is the period of pitch oscillation; is the pitch angular velocity.

[0050] 5) Based on the Kriging model, the proxy model between the aerodynamic torque periodic work and the three given motion parameters is constructed as follows:

[0051]

[0052] 6) Perform encrypted sampling on the parameter space to obtain a new parameter sampling set (A α,new ,f α,new , ), the new parameter sampling set is subjected to the same CFD calculation of the given motion by the above method, and the aerodynamic torque cycle work corresponding to each parameter combination is calculated, and the new parameter sampling set and the data of all previous parameter sampling sets are further used to construct a new proxy model:

[0053]

[0054] 7) Set the new parameter sampling set (A α,new ,f α,new , ) are respectively brought into the proxy model and Get vector and Calculate the relative error between these two vectors ||y i+1 -y i || / ||y i+1 || (|| || is the norm of the vector), if the relative error is less than 5%, the amplitude of the pitch oscillation is given by A α , frequency f α Solve the equation:

[0055]

[0056] The optimal phase difference corresponding to the pitch oscillation amplitude and frequency can be obtained

[0057] After constructing the above proxy model, the stall flutter closed-loop control can be performed, such as Figure 4 As shown, the specific method is as follows:

[0058] Step 1: Given the control start time t c , the amount of data used for fitting N m, filter parameter μ, filter sampling interval N F , calculate the total time t To .

[0059] Step 2: Perform fast Fourier transform on the stall flutter pitch angle response without skin active vibration control to obtain the peak frequency f peak , then the time domain response function of the stall flutter pitch angle to be identified can be expressed as

[0060]

[0061] In the formula, A αm (t) represents the pitch oscillation amplitude; represents the pitch oscillation phase; α 0m (t) represents the average value of the pitch angle;

[0062] Step 3: Determine whether the current time t is greater than the total calculation time t To , if it is greater than, the calculation ends; otherwise, the stall flutter pitch oscillation response at the current moment is calculated, and the pitch angle signal α at the current moment is collected n Put it into the pitch angle time series α;

[0063] Step 4: If the number of data in the pitch angle time series α is N α Less than the amount of data used for fitting N m , then directly enter the loop of the next time step, at this time t = t + dt, return to step 3, otherwise go to step 5.

[0064] Step 5: Solve

[0065]

[0066] Then follow the formula

[0067]

[0068] Calculate the pitch oscillation amplitude A at the current moment αm , Phase Average value α 0m At the same time, the current A αm , α 0m Put it into the corresponding time series A αm , α 0m middle.

[0069] Step 6: Determine whether the current time t is greater than the control start time t c , if it is not greater than, directly enter the loop of the next time step t=t+dt and return to step 3; otherwise, proceed to step 7.

[0070] Step 7: From the current moment, from time series A αm , α 0m Interval N F Sample forward until the index is less than 1 to get the sampling sequence

[0071] Step 8: Get the sampling sequence The data length N s , construct the filter function Φ according to the following formula:

[0072]

[0073] Where I is N s ×N s The unit matrix of ; μ is the filtering parameter.

[0074]

[0075] Step 9: Since the time series A obtained by direct fitting in step 5 αm , α 0m There is a certain amount of noise, so the sampling sequence is filtered using the filter function Φ After filtering, the sampling sequence after filtering is:

[0076]

[0077] The amplitude, frequency and phase at the current moment after filtering are:

[0078]

[0079] like Figure 5a , Figure 5b , Figure 5c , which is the calculation effect of the stall flutter transient response characteristics using the method of the present invention under a certain working condition. It can be seen that the method of the present invention can accurately extract the instantaneous amplitude, frequency and phase information of the stall flutter response.

[0080] Step 10: Using the proxy model built in step 1 Solution Get the optimal phase difference corresponding to the instantaneous amplitude and frequency of the pitch oscillation at the current moment Then the phase of the active vibration of the skin The active vibration displacement of the skin applied at the current moment is Then enter the loop for the next time step.

[0081] like Figure 6As shown, the stall flutter closed-loop control effect based on the optimal phase vibration of the skin of the present invention is adopted in a certain working condition. After the closed-loop control is applied at 150s, the stall flutter amplitude is reduced from 24.4 degrees to 8 degrees, and the amplitude is reduced by 67.2%.

Claims

1. A stall flutter closed-loop control method based on skin optimal phase vibration, characterized in that: The control flow is: Step 1: Given the control start time t c , the amount of data used for fitting N m , filter parameter μ, filter sampling interval N F The total calculation time t To ; Step 2: Perform fast Fourier transform on the stall flutter pitch angle response without skin active vibration control to obtain the peak frequency f peak , the time domain response function form of the stall flutter pitch angle to be identified is expressed as: In the formula, A αm (t) represents the pitch oscillation amplitude; represents the pitch oscillation phase; α 0m (t) represents the average value of the pitch angle; x is the chord-wise coordinate, x1 is the chord-wise coordinate of the left end point of the skin active vibration area, and x2 is the chord-wise coordinate of the right end point of the skin active vibration area; α(t) is the given signal of the airfoil pitch oscillation motion; Step 3: If the current time t is greater than the total calculation time t To , then the calculation ends; otherwise, calculate the stall flutter pitch oscillation response at the current moment and collect the pitch angle signal α at the current moment n Put it into the pitch angle time series α; Step 4: If the number of data in the pitch angle time series α is N α Less than the amount of data used for fitting N m , then directly enter the loop of the next time step, otherwise proceed to step 5; Step 5: Fit the pitch oscillation amplitude A at the current moment αm , Phase Average value α 0m , put into the corresponding time series A αm , α 0m middle; Step 6: If the current time t is less than the control start time t c , then directly enter the loop of the next time step and return to step 3; otherwise, proceed to step 7; Step 7: From the current moment, from time series A αm , α 0m Interval N F Sample forward until the index is less than 1 to get the sampling sequence Step 8: Get the sampling sequence The data length N s , construct the filter function Φ; Step 9: Use the filter function Φ to filter the sampling sequence Perform filtering; Step 10: Using the pre-built proxy model, combined with the filtered amplitude, frequency and phase at the current moment obtained in step 9, the optimal phase difference corresponding to the instantaneous amplitude and frequency of the pitch oscillation at the current moment is obtained. Then the phase of active vibration of the skin is The active vibration displacement of the skin applied at the current moment is Then enter the next time step loop; in the above formula, is the phase at the current moment after filtering.

2. The stall flutter closed-loop control method based on skin optimal phase vibration according to claim 1, characterized in that: The optimal phase difference calculation method corresponding to the instantaneous amplitude and frequency of the pitch oscillation of the agent model pre-built in step 10 is: 1) Given the amplitude of pitch oscillation A α , frequency f α Range and skin active vibration phase difference scope; 2) performing Latin hypercube sampling in the parameter space determined by the three parameter ranges given in 1) to obtain a parameter sampling set; 3) For each parameter combination in the parameter sampling set obtained in 2) Perform CFD calculations for a given motion to obtain aerodynamic moment time domain data M(t); 4) According to the obtained aerodynamic torque time domain data M(t), the aerodynamic torque period work is calculated as: Where, T is the period of pitch oscillation; is the pitch angular velocity; 5) Based on the Kriging model, the proxy model between the aerodynamic torque periodic work and the three given motion parameters is constructed as follows: 6) Encrypt the parameter space and obtain a new parameter sampling set The new parameter sampling set is subjected to the same CFD calculation of the given motion by the above method, and the aerodynamic torque cycle work corresponding to each parameter combination is calculated. The new parameter sampling set and the data of all previous parameter sampling sets are further used to construct a new proxy model: 7) Set the new parameter sampling set The parameter combinations in are respectively brought into the proxy model and Get vector and Calculate the relative error between two vectors ||y i+1 -y i || / ||y i+1 ||, if the relative error is less than 5%, the amplitude of the pitch oscillation is given by A α , frequency f α Solve the equation: Get the optimal phase difference corresponding to the pitch oscillation amplitude and frequency 3. The stall flutter closed-loop control method based on skin optimal phase vibration as claimed in claim 2, characterized in that: The given motion in step 3) includes: The given signal of the airfoil pitch oscillation motion is: α=A α sin(2πf α t) Where, t is time; The upper surface skin actively vibrates, and its given signal is: The active vibration displacement of the lower surface skin is the same as that of the upper surface, but with opposite signs. The given signal is: In the above formula, x is the chord-wise coordinate, x1 is the chord-wise coordinate of the left end point of the skin active vibration area, and x2 is the chord-wise coordinate of the right end point of the skin active vibration area.

4. The stall flutter closed-loop control method based on skin optimal phase vibration as claimed in claim 1, characterized in that: The pitch oscillation amplitude A at the current moment in step 5 αm , Phase Average value α 0m The method to obtain is: Solution Then according to the formula Calculated.

5. The stall flutter closed-loop control method based on skin optimal phase vibration as claimed in claim 1, characterized in that: In step 8, obtain the sampling sequence The data length N s , according to the following formula, the filter function is: Where I is N s ×N s The unit matrix of ; μ is the filtering parameter; Using the filter function Φ to filter the sampling sequence After filtering, the sampling sequence after filtering is: The amplitude, frequency and phase at the current moment after filtering are:

Images