Helicopter multichannel vibration active control method based on secondary channel damping compensation

Abstract

A helicopter multi-channel vibration active control method based on secondary channel damping compensation comprises the following steps: acquiring acceleration response error signals of each control point of a helicopter body, inputting the acceleration response error signals into a band-pass filter corresponding to harmonic waves to obtain sub-error signals of each order of harmonic waves, and taking the minimization of global vibration energy as a target to obtain a sub-error signal of each order of harmonic waves; updating the FxLMS controller weight corresponding to each order of harmonic wave by adopting a gradient descent algorithm, and inputting the reference signal into the FxLMS controller so as to generate a feedforward control signal; the response of each secondary channel under the master control signal is predicted through a secondary channel identification model, a predicted value is generated to serve as the input of an acceleration negative feedback controller, the output of the acceleration negative feedback controller is a feedback control signal, and therefore damping compensation of the secondary channels in the FxLMS controller is achieved in a decoupling mode; the feedforward control signal and the feedback control signal are superposed to form a total control signal which serves as an input signal of an actuator in the next cycle, the input signal drives the actuator to generate main power, the main power responds at a fuselage control point through a secondary channel to counteract the response generated by rotor excitation, and the main control signal is used as an output signal of the actuator in the next cycle. The counteracted residual vibration is the collected acceleration response error signal, and continuously circulating the control period until the acceleration response root mean square of all the control points of the fuselage reaches the minimum value under the control of a plurality of actuators. According to the method, the convergence speed and the stability can be improved under the condition that the structure of the feedforward FxLMS controller is not changed.

Description

Technical Field

[0001] The present invention relates to a technology in the field of vibration control, specifically a multi-channel vibration active control method for helicopters based on secondary channel damping compensation. Background Art

[0002] The main rotor of a helicopter generates vibrations with a frequency of kNΩ (k = 1, 2, 3...; N is the number of rotor blades, and Ω is the rotor speed) under the action of alternating aerodynamic forces and inertial forces. Changes in flight attitude and fluctuations in instantaneous power demand result in a time-varying characteristic of the rotor vibration load, and the body vibration response includes a steady-state response and a transient response. Although existing vibration active control methods can effectively suppress the steady-state harmonic response, there are still deficiencies in controlling the transient response. Existing optimized active noise reduction technologies based on FxLMS only consider the control problems of single-input single-output systems, cannot solve the control problems of more complex multi-input multi-output systems, and cannot optimize the performance of the secondary channel. At the same time, the real-time computational load of the hardware system is relatively high. Summary of the Invention

[0003] Aiming at the problem that the existing feedforward filtered least mean square (FxLMS) controller cannot effectively control the transient response and has a slow convergence speed and poor stability when facing a secondary channel with small damping, the present invention proposes a multi-channel vibration active control method for helicopters based on secondary channel damping compensation, which can improve the convergence speed and stability without changing the structure of the feedforward FxLMS controller.

[0004] The present invention is realized through the following technical solutions:

[0005] The present invention relates to a multi-channel vibration active control method for helicopters based on secondary channel damping compensation. In one control cycle, the acceleration response error signals of each control point on the fuselage are collected and input into the band-pass filters corresponding to the respective harmonics to obtain the sub-error signals of each harmonic order. With the goal of minimizing the global vibration energy, the gradient descent algorithm is used to update the weights of the FxLMS controllers corresponding to each harmonic order. The reference signal is input into the FxLMS controller to generate a feedforward control signal. Then, the response of each secondary channel under the total control signal is predicted through the secondary channel identification model, and the predicted value is used as the input of the acceleration negative feedback controller. The output of the acceleration negative feedback controller is the feedback control signal, thereby realizing the damping compensation of the secondary channel in the FxLMS controller in a decoupled form. The feedforward control signal and the feedback control signal are superimposed to form the total control signal, which is used as the input signal of the actuator in the next cycle. The input signal drives the actuator to generate an active force, and the active force generates a response at the control point on the fuselage through the secondary channel, canceling the response generated by the rotor excitation. The remaining vibration after cancellation is the acceleration response error signal collected. The control cycle is continuously repeated until the root mean square of the acceleration responses of all control points on the fuselage reaches the minimum value under the control of multiple actuators.

[0006] The secondary channel mentioned above refers to the vibration transmission channel from the actuator to the acceleration sensor at the control point.

[0007] The reference signal mentioned above refers to the harmonic signal generated by the harmonic signal generator corresponding to the rotor passing frequency and its multiples.

[0008] The acceleration negative feedback controller is: where: g is the controller gain, ζ n is the damping ratio, ω n is the natural frequency of the acceleration negative feedback controller.

[0009] The damping compensation for the secondary channel mentioned above means that the acceleration negative feedback controller takes the acceleration response of the secondary channel as the controller input, and the controller output is in the same phase as the velocity response of the secondary channel, so the equivalent damping of the secondary channel can be increased.

[0010] The decoupled form mentioned above means that the input signal of the acceleration negative feedback controller is the predicted value of the response of the secondary channel by the secondary channel identification model, rather than the signal obtained by superimposing the rotor excitation response and the actuator active force response measured by the acceleration sensor at the control point during the control process.

[0011] In the total control signal, the feedforward control signal suppresses the harmonic response at the control point, and the feedback signal suppresses the transient response generated by the secondary channel at the control point. Therefore, the present invention achieves rapid attenuation of the transient response caused by the change in the rotor vibration load by performing damping compensation on the secondary channel, thereby effectively improving the suppression speed of the harmonic response generated by the helicopter rotor and the stability of the system.

[0012] The present invention relates to a system for realizing the above method, comprising: a bandpass filter unit, a secondary channel identification model unit, a damping-compensated secondary channel identification model unit, an acceleration negative feedback controller unit and a multi-harmonic feedforward controller unit, wherein: the bandpass filter unit processes the collected acceleration response error signal to obtain a sub-error signal corresponding to each order harmonic; the secondary channel identification model unit predicts the response of the secondary channel under the action of the total control signal to obtain the input of the acceleration negative feedback controller; the damping-compensated secondary channel identification model unit filters the reference signal to offset the influence of the secondary channel on the gradient estimation value; the acceleration negative feedback controller unit generates a feedback control signal according to the predicted value of the secondary channel response, and suppresses the transient response by increasing the equivalent damping of the secondary channel; the multi-harmonic feedforward controller unit updates the FxLMS controller weight of the corresponding harmonic according to the reference signal and the sub-error signal by using the gradient descent method to generate a feedforward control signal; and the feedforward control signal and the feedback control signal are superimposed to form a total control signal. Technical Effects

[0013] The present invention combines an acceleration negative feedback controller to achieve damping compensation of the secondary channel in a deconstructed form without changing the structure of the multi-channel FxLMS algorithm. Compared with the prior art, the present invention can quickly attenuate the transient response caused by the change of the rotor load, reduce the overshoot, and significantly improve the convergence speed of the multi-harmonic response, so that the system can quickly reach a steady state under complex flight conditions, enhance the system stability, and avoid the problems of slow transient response attenuation, difficulty in balancing convergence speed and system stability, and fluctuations in the convergence process existing in the existing FxLMS controller. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is the principle diagram of the present invention;

[0015] Figure 2 It is the schematic diagram of damping compensation of secondary channel;

[0016] Figure 3 It is a schematic diagram of a two-channel embodiment;

[0017] FIG4( a) is a diagram showing the secondary channel identification process from the first actuator to the first sensor;

[0018] Figure 4(b) is the identification process diagram of the secondary channel from the first actuator to the first sensor after damping compensation;

[0019] Figure 5 It is the tap comparison diagram obtained by identifying the secondary channels from the first actuator to the first sensor before and after damping compensation respectively;

[0020] Figure 6 It is the time-domain diagram before and after control of helicopter airframe control points 1 to 4 under the condition of changing rotor load in the experiment. Specific implementation manner

[0021] As Figures 1 - 3 shown, this embodiment relates to a multi-channel vibration active control method for helicopters based on secondary channel damping compensation, including:

[0022] Step 1, due to the large number of secondary channels in the helicopter vibration active control system, considering the controller's feasibility, determine some secondary channels by comparing the decay rates of the unit impulse responses of all secondary channels: use the unit impulse signal to excite the secondary channels, analyze the response data measured by the acceleration sensor, calculate the peak value of the residual response after the unit impulse excitation. For the secondary channels whose peak value of the residual response is greater than 40% of the initial response after 1 s of unit impulse excitation, they will be selected for damping compensation.

[0023] The secondary channel mentioned above is H ji (z), representing the secondary channel from actuator j to the airframe control point i.

[0024] Step 2, secondary channel identification: sequentially use white noise to drive each actuator to excite the airframe, collect the input signals of each actuator and the acceleration response signals of each sensor, and based on the collected signals, use the normalized least mean square algorithm to identify each secondary channel to minimize the squared error between the input signal and the output signal until the optimal secondary channel identification model is obtained.

[0025] The secondary channel identification model mentioned above is representing the secondary channel identification model from actuator j to the airframe control point i. For example, the secondary channel identification model from the first actuator to the airframe control point 1 is Specifically, it is where: are the model coefficients, and M is the model order.

[0026] Step 3, construct an acceleration negative feedback controller: construct an acceleration negative feedback controller for the main contribution modes of the airframe response under the first-order through-frequency harmonic excitation of the helicopter rotor. For a certain order of the mode to be processed, set the natural frequency ω of the acceleration negative feedback controller nTune to the natural frequency of the mode to be processed, and further construct the controller gain g and damping ratio ζ n , and it is also necessary to construct a first-order all-pass filter H ap (z) to cancel the phase lag caused by the non-collocated configuration of the actuator and the sensor. The first-order all-pass filter is connected in series with the acceleration negative feedback controller and in parallel with the acceleration negative feedback controller for damping compensation of each order of the main contribution modes.

[0027] The main contribution modes of the helicopter rotor body response under the first-order through-frequency harmonic excitation can be determined based on the identification model of the secondary channel. Specifically, in the implementation, among the frequency responses of the identification model of the secondary channel, the third-order overall body mode closest to the first-order harmonic excitation frequency is used as the main contribution mode.

[0028] The phase lag is the phase gain at the natural frequency of the mode to be processed in the secondary channel, and can be obtained from the frequency response of the identification model of the secondary channel.

[0029] Step 4, identify the closed-loop system formed by the secondary channel, the first-order all-pass filter, and the acceleration negative feedback controller: sequentially use white noise to drive each actuator to excite the closed-loop system, collect the input signals of each actuator and the acceleration response signals of each sensor, and based on the collected signals, use the normalized least mean square algorithm to identify all closed-loop systems to obtain the identification models of all closed-loop systems.

[0030] The closed-loop system mentioned above refers to the secondary channel after damping compensation using the acceleration negative feedback controller represents the secondary channel after damping compensation from actuator j to the fuselage control point i.

[0031] The identification models of all the closed-loop systems mentioned above are represents the identification model of the secondary channel after damping compensation from actuator j to the fuselage control point i. For example, if it has been selected to perform damping compensation on the secondary channel from the first actuator to the fuselage control point 1, then the identification model of this secondary channel after damping compensation is Specifically, it is where: are the model coefficients, and M is the model order.

[0032] Step 5, error signal measurement: When performing active vibration control on the helicopter, collect the acceleration response error signal e i (n) at each control point of the fuselage and input it into the band-pass filter B k (z) constructed for each order of harmonics to obtain the sub-error signals corresponding to each order of harmonics where: k = 1, 2, 3 are the harmonic orders.

[0033] The acceleration response error signal e at each control point of the fuselage i (n), measured by the acceleration sensor, specifically: the response d i (n) of the rotor excitation at control point i and the response of all actuator forces at control point i superimposed, where: p is the number of actuators.

[0034] According to the frequency characteristics of the rotor excitation, when choosing to control the first three harmonics, the band-pass filter is specifically B1(z) with a frequency band range of 15 - 25 Hz, B1(z) with a frequency band range of 35 - 45 Hz, and B1(z) with a frequency band range of 50 - 70 Hz.

[0035] Step 6, generate the reference signal: Obtain the rotor speed signal Ω, and use the harmonic signal generator to generate harmonic signals x k (n) with frequencies of kNΩ as the reference signals for the feedforward FxLMS controllers for controlling each harmonic, where: k = 1, 2, 3 are the harmonic orders.

[0036] Step 7, update the weights of the feedforward FxLMS controllers corresponding to each harmonic: According to the sub-error signals corresponding to each harmonic obtained in Step 5 and the harmonic signals generated in Step 6, use the gradient descent algorithm to update the weights of the feedforward FxLMS controllers corresponding to each harmonic with the goal of minimizing the global vibration energy to obtain the actuator control signals for controlling each harmonic

[0037] The minimum global vibration energy mentioned above means: in a multi-input multi-output system, under the joint control of multiple actuators, the root mean square value of the acceleration sensor responses at multiple control points of the helicopter fuselage is minimized, where: m is the number of control points.

[0038] The update of the weights of the feedforward Fx-LMS controllers corresponding to each harmonic means: where: is the iteration step size corresponding to the j-th actuator in the k-th harmonic feedforward Fx-LMS controller, is the control factor corresponding to the j-th actuator in the k-th harmonic feedforward FxLMS controller, is the filtered -x signal obtained by filtering the reference signals of each harmonic generated in Step 7 through the secondary channel identification model with damping compensation from actuator j to the fuselage control point i in Step 4, and the signal length is M, specifically:

[0039] The iteration step size satisfies where: λ is the eigenvalue of the autocorrelation matrix of the filtered - x signal. In implementation, a smaller iteration step is first selected, and then it is gradually increased according to the convergence effect and stability of the system.

[0040] Step 8, secondary path response prediction: When performing active vibration control on the helicopter, the acceleration sensor response at control point i is the superposition of the response d i (n) of the rotor excitation at the control point and the corresponding secondary path response . The response y ji (n) of each secondary path is predicted through the secondary path identification model obtained in Step 2, and the predicted value is

[0041] Step 9, the total control signal is formed by superimposing the control signal of the feed - forward Fx - LMS controller and the total control signal of the acceleration negative feedback controller. Specifically: where: the total control signal of the feed - forward Fx - LMS controller where: is the actuator control signal for controlling each harmonic, is the controller weight obtained in Step 7 for controlling each harmonic, x k (n) is the reference signal of the feed - forward Fx - LMS controller for controlling each harmonic obtained in Step 6, l is the sequence number of the main contribution mode described in Step 3, is the output of the acceleration negative feedback controller corresponding to each main contribution mode in the secondary path from actuator j to sensor i, is the acceleration negative feedback controller constructed for each main contribution mode in the secondary path from actuator j to sensor i in Step 3, H ap (z) is the first - order all - pass filter constructed for the secondary path from actuator j to sensor i in Step 3, is the predicted value of the response of the secondary path from actuator j to sensor i in Step 8. Specifically:

[0042] Step 10, the total control signal for driving the actuator obtained in Step 9 is used as the input signal of the actuator for the next cycle of iteration. The input signal drives the actuator to generate an active force, and the active force generates a response signal through the secondary path at the control point on the helicopter fuselage to be controlled, so as to cancel the response of the rotor excitation at the control point. The remaining vibration after cancellation is the error signal e i(n), continuously execute steps 5 to 9 in a loop until the root mean square of the acceleration responses of all control points on the fuselage reaches the minimum value under the joint control of multiple actuators.

[0043] According to practical experience, in the helicopter vibration active control test bench, the controller step sizes corresponding to the first three passing frequencies of the rotor in this method are set to 0.0003, 0.0004, and 0.00065 respectively. The gain of the acceleration negative feedback controller constructed for each main contribution mode of each secondary channel is set to 0.3, and the damping ratio is constructed to be 0.45. By the above method, when the rotor excitation changes suddenly, the transient vibration of the fuselage at the control points can decay rapidly, and the convergence speed of the harmonic response is also faster. In order to ensure the stability of the control system, the maximum step sizes corresponding to the first three passing frequencies in the traditional feedforward FxLMS controller can only be set to 0.00012, 0.0003, and 0.0055.

[0044] As shown in Figure 4(a), a 2048-order finite impulse response model is used to identify the secondary channel from the first actuator to control point 1, and the identification range is set to 15 - 100 Hz. As shown in Figure 4(b), a 2048-order finite impulse response model is used to identify the damped compensated secondary channel from the first actuator to control point 1, and the identification range is set to 15 - 100 Hz. Comparing the identification results in Figure 4(a) and Figure 4(b), it can be seen that when using the same order identification model, identifying the damped compensated secondary channel has a smaller steady-state error, indicating higher identification accuracy.

[0045] As Figure 5 shown, the identification results of the secondary channel from the first actuator to control point 1 before and after damping compensation are compared. The unit impulse response of the secondary channel before damping compensation decays slowly, and it is difficult for the 2048-order finite impulse response model to fully characterize the dynamic characteristics of the secondary channel. While after damping compensation, the unit impulse response of the secondary channel decays rapidly, and at this time the 2048-order finite impulse response model can better characterize the dynamic characteristics, that is, the identification accuracy is higher.

[0046] As Figure 6As shown, it is the time-domain diagram before and after the control of helicopter airframe control points 1 to 4 under the condition of changing rotor load in the experiment. The change of rotor load is divided into 7 stages, and each stage is 30 s. The first stage is from 0 to 30 s, and the excitation condition is the vertical first three harmonic loads; the second stage is from 30 to 60 s, and the amplitude of the excitation force increases by 100%; the third stage is from 60 to 90 s, and the phase of the excitation force changes by 180°; the fourth stage is from 90 to 120 s, and the frequency of the excitation force increases by 2.5%; the fifth stage is from 120 to 150 s, and the phase of the excitation force changes by 180°; the sixth stage is from 150 to 180 s, and the amplitude of the excitation force decreases by 100%; the seventh stage is from 180 to 210 s, and the control is turned off. Compared with the traditional feedforward FxLMS algorithm, this method can set a larger step size while ensuring the stability of the control system. Therefore, it can converge faster after the sudden change of the excitation force, and there is no obvious fluctuation phenomenon during the convergence process.

[0047] The present invention realizes the damping compensation for the secondary channel by combining an acceleration negative feedback controller, effectively solving the problem of insufficient vibration suppression ability of the existing feedforward FxLMS controller when dealing with the change of rotor load during the flight of the helicopter. Through the optimization of the damping characteristics of the secondary channel, the present invention significantly improves the transient response suppression ability of the control system, can quickly attenuate the transient response caused by the sudden change of the excitation force, and reduce the overshoot; at the same time, it improves the convergence speed of the feedforward FxLMS controller for suppressing the multi-harmonic response, enabling the system to quickly reach a steady state under complex flight conditions; in addition, it enhances the system stability by broadening the step size selection range, avoiding the problem that the existing feedforward FxLMS controller is prone to divergence under a large step size. Compared with the problems existing in the application of the existing feedforward FxLMS technology, such as the slow attenuation of the transient response, the difficulty in balancing the convergence speed and the system stability, and the fluctuation during the convergence process, this method shows advantages when the rotor excitation changes. This method can not only quickly attenuate the transient response caused by the time-varying excitation, effectively reduce the overshoot of the system, but also suppress the harmonic response with a faster convergence speed, while avoiding the fluctuation phenomenon during the convergence process. It can be verified from the experimental results of the active vibration control based on time-varying excitation that this method can configure a larger convergence step size for the feedforward FxLMS controller used to control the harmonic responses of each order, thereby broadening the step size selection range and enhancing the stability of the system.

[0048] The above specific implementation can be locally adjusted in different ways by those skilled in the art without departing from the principles and purposes of the present invention. The protection scope of the present invention is subject to the claims and is not limited by the above specific implementation. All implementation solutions within its scope are subject to the present invention.

Claims

1. A helicopter multi-channel vibration active control method based on secondary channel damping compensation, characterized in that: In one control cycle, the acceleration response error signal of each control point of the fuselage is collected and input into the bandpass filter of the corresponding harmonic to obtain the sub-error signal of each order harmonic. With the goal of minimizing the global vibration energy, the gradient descent algorithm is used to update the weight of the FxLMS controller corresponding to each order harmonic, and the reference signal is input into the FxLMS controller to generate a feedforward control signal; then the response of each secondary channel under the total control signal is predicted by the secondary channel identification model, and the predicted value is generated as the input of the acceleration negative feedback controller. The output of the acceleration negative feedback controller is the feedback control signal, thereby realizing the damping compensation of the secondary channel in the FxLMS controller in a decoupled form; the feedforward control signal and the feedback control signal are superimposed to form a total control signal, which is used as the input signal of the actuator in the next cycle. The input signal drives the actuator to generate the main force, and the main force generates a response at the fuselage control point through the secondary channel to offset the response generated by the rotor excitation. The residual vibration after offset is the collected acceleration response error signal, and the control cycle is continuously repeated until the root mean square of the acceleration response of all control points of the fuselage under the control of multiple actuators reaches the minimum value; The secondary channel refers to: a vibration transmission channel from the actuator to the acceleration sensor at the control point; The reference signal refers to: a harmonic signal corresponding to the rotor passing frequency and its multiples generated by a harmonic signal generator; The damping compensation for the secondary channel means that the acceleration negative feedback controller uses the acceleration response of the secondary channel as the controller input, and the controller output is in phase with the speed response of the secondary channel, thereby increasing the equivalent damping of the secondary channel.

2. The helicopter multi-channel vibration active control method based on secondary channel damping compensation according to claim 1 is characterized in that: The acceleration negative feedback controller is: Where: g is the controller gain, ζ n is the damping ratio, ω n is the natural frequency of the acceleration negative feedback controller.

3. The helicopter multi-channel vibration active control method based on secondary channel damping compensation according to claim 1 is characterized in that: The decoupling form means that the input signal of the acceleration negative feedback controller is the predicted value of the secondary channel response by the secondary channel identification model, rather than the signal obtained by superimposing the rotor excitation response and the actuator force response measured by the acceleration sensor at the control point during the control process.

4. The helicopter multi-channel vibration active control method based on secondary channel damping compensation according to claim 1 is characterized in that: In the total control signal, the feedforward control signal suppresses the harmonic response at the control point, and the feedback signal suppresses the transient response generated by the secondary channel at the control point. By performing damping compensation on the secondary channel, the transient response caused by the change in the rotor vibration load is quickly attenuated, thereby effectively improving the suppression speed of the harmonic response generated by the helicopter rotor and the stability of the system.

5. The helicopter multi-channel vibration active control method based on secondary channel damping compensation according to claim 1 is characterized in that: The damping compensation is determined by comparing the attenuation speed of the unit pulse response of all secondary channels. Specifically, the secondary channels are excited by a unit pulse signal, and the residual response peak after the unit pulse excitation is calculated by analyzing the response data measured by the acceleration sensor. The secondary channels whose residual response peak after the unit pulse excitation is greater than 40% of the initial response after 1 second of the unit pulse excitation will be selected for damping compensation.

6. The helicopter multi-channel vibration active control method based on secondary channel damping compensation according to claim 1 is characterized in that: The secondary channel identification refers to: using white noise to drive each actuator to stimulate the body in turn, collecting the input signal of each actuator and the acceleration response signal of each sensor, and based on the collected signals, using the normalized least mean square algorithm to identify each secondary channel to minimize the square error between the input signal and the output signal until the optimal secondary channel identification model is obtained; The secondary channel identification model is The secondary channel identification model from actuator j to fuselage control point i includes: the secondary channel identification model from the first actuator to fuselage control point 1 is Specifically in: is the model coefficient, and M is the model order.

7. The helicopter multi-channel vibration active control method based on secondary channel damping compensation according to claim 1 is characterized in that: The continuous cycle control cycle refers to: Step a, error signal measurement: When actively controlling the vibration of the helicopter, the acceleration response error signal e at each control point of the fuselage is collected. i (n) and input the bandpass filter B constructed for each order of harmonics k (z), and obtain the sub-error signals corresponding to each order of harmonics Where: k = 1, 2, 3 are harmonic orders; Step b, generate reference signal: obtain rotor speed signal Ω, use harmonic signal generator to generate harmonic signal x with frequency kNΩ k (n) as the reference signal of the feedforward FxLMS controller for controlling each order of harmonics, where: k = 1, 2, 3 are the harmonic orders; Step c, updating the weight of the feedforward FxLMS controller corresponding to each order harmonic: based on the sub-error signal corresponding to each order harmonic obtained in step a and the harmonic signal of each order generated in step b, the weight of the feedforward FxLMS controller corresponding to each order harmonic is updated by using the gradient descent algorithm with the global vibration energy minimization as the control target, and the actuator control signal for controlling each order harmonic is obtained. Step d, secondary channel response prediction: When active vibration control is performed on a helicopter, the acceleration sensor response at control point i is the response d of the rotor excitation at the control point i (n) and the corresponding secondary channel response The superposition of the secondary channel identification model to each secondary channel response y ji (n) is predicted, and the predicted value is Step e, the signal is controlled by the feedforward Fx-LMS controller And the total control signal of the acceleration negative feedback controller The total control signal is superimposed, which is: Where: Total control signal of feedforward Fx-LMS controller Total control signal of acceleration negative feedback controller in: is the actuator control signal used to control each order harmonic, is the controller weight for controlling each order harmonic obtained in step c, x k (n) is the reference signal of the feedforward Fx-LMS controller used to control each order harmonic obtained in step b, l is the serial number of the main contributing mode, is the output of the acceleration negative feedback controller corresponding to each main contribution mode in the secondary channel from actuator j to sensor i, is the acceleration negative feedback controller constructed for each main contributing mode in the secondary channel from actuator j to sensor i, H ap (z) is a first-order all-pass filter constructed for the secondary path from actuator j to sensor i, is the predicted value of the secondary channel response from actuator j to sensor i in step d, specifically: Step f, using the total control signal of the driving actuator obtained in step e as the input signal of the actuator in the next loop iteration, the input signal drives the actuator to generate the main force, and the main force generates a response signal at the control point to be controlled of the helicopter fuselage through the secondary channel To offset the response of the rotor excitation at the control point, the residual vibration after offset is the error signal e measured in step a i (n), continuously executing step a to step e in a loop until the root mean square of the acceleration response of all control points of the fuselage reaches a minimum value under the joint control of the multiple actuators.

8. The helicopter multi-channel vibration active control method based on secondary channel damping compensation according to claim 7 is characterized in that: The acceleration response error signal e at each control point of the fuselage i (n), measured by the acceleration sensor, specifically: the response d of the rotor excitation at the control point i i (n) and the response of all actuator forces at control point i The superposition is formed, where: p is the number of actuators; The global vibration energy minimum means that in a multi-input multi-output system, under the joint control of multiple actuators, the root mean square value of the acceleration sensor response at multiple control points of the helicopter fuselage is Minimum, where: m is the number of control points; The weights of the feedforward Fx-LMS controller corresponding to each order harmonic are updated as follows: in: is the iteration step size corresponding to the jth actuator in the kth order harmonic feedforward Fx-LMS controller, is the control factor corresponding to the jth actuator in the kth order harmonic feedforward FxLMS controller, The reference signals of each order generated in step b are filtered by the secondary channel identification model after damping compensation through the actuator j to the fuselage control point i. The signal length is M, specifically: in: is the secondary channel identification model after damping compensation from actuator j to fuselage control point i, which is specifically in: is the model coefficient, M is the model order; The iteration step size satisfies Where: λ is the eigenvalue of the autocorrelation matrix of the filtered-x signal.

9. A helicopter multi-channel active vibration control system implementing the method described in any one of claims 1 to 8, characterized in that: include: A bandpass filter unit, a secondary channel identification model unit, a secondary channel identification model unit after damping compensation, an acceleration negative feedback controller unit and a multi-harmonic feedforward controller unit, wherein: the bandpass filter unit processes the collected acceleration response error signal to obtain a sub-error signal corresponding to each order of harmonics; the secondary channel identification model unit predicts the response of the secondary channel under the action of the total control signal to obtain the input of the acceleration negative feedback controller; the secondary channel identification model unit after damping compensation filters the reference signal to offset the influence of the secondary channel on the gradient estimation value; the acceleration negative feedback controller unit generates a feedback control signal according to the predicted value of the secondary channel response, and suppresses the transient response by increasing the equivalent damping of the secondary channel; the multi-harmonic feedforward controller unit updates the FxLMS controller weight of the corresponding harmonic according to the reference signal and the sub-error signal by the gradient descent method to generate a feedforward control signal; the feedforward control signal and the feedback control signal are superimposed to form a total control signal.

Images