Research On Vibration Control Of Piezoelectric Flexible Plate Based On Non-contact Measurement

Modern spacecraft often employs a variety of flexible structures such as solar array and antenna.The structures exhibit the characteristics of large dimension,lightweight,which leads to its low-frequency and wide-amplitude vibration.Under excitation,the vibration will last for a long period of time,which may accelerate the fatigue damage process of mechanisms,even affect the performance of systems.Therefore,low frequency mode vibration detection and suppression for the flexible plate structure is necessary.Binocular stereo vision and laser displacement sensors were combined to measure the vibration signals of flexible plate,a detection mark setting method of projecting laser dot lattice for photogrammetry was proposed.The calculation of three-dimensional coordinates was carried out based on convergent model instead of standard model,on one hand,there is not installation restriction of paralleling optical axes in vision system,on the other hand,the accuracy of measurement can be improved because the epipolar rectification procedure is leaved out.On the surface of flexible hinged plate,several circular cards were pasted as marks,therefore,the acquired information is more accurate compared with sticking on a whole printing paper,for the card joint achieves the closer attachment.Laser dot array was projected to form detection targets in flexible cantilevered plate system,which simplified the installation process,moreover,the styles of dot pattern can be easily adjusted,the efficiency is increased.In order to guarantee the reliability and stability,an algorithm of image processing was proposed to obtain the pixel coordinates of detection marks,and the measurement error brought with dot projection was analysed.To suppress the vibration effectively and quickly,designed active vibration controllers for the flexible cantilevered plate and hinged plate according to their structural characteristics,including nonlinear controller,sliding mode predictive controller and fuzzy neural network.Considering the model of cantileverd plate is more accurate,a kind of sliding mode predictive control strategy was investigated,the numerical simulation was conduct based on the built finite element model,for the verification of control effect,in addition,nonlinear control law was utilized to compensate the nonlinear characteristics of the real system,the parameters of nonlinear function was designed to make the control output decrease appropriately when the vibration is in large amplitude,increase when in residual vibration phase.For the nonlinear characteristics that can not be modeled in hinged plate system,a more robust controller based on Mamdani fuzzy neural network algorithm was designed for active suppressing.The parameters of the controller are updated in each sampling interval,which compensates the uncertainties of system.The experimental setups of the piezoelectric flexible hinged plate with binocular stereo vision and the piezoelectric flexible cantilevered plate based on laser projection vision,laser sensors are constructed,respectively.The vibration suppression experiments using the adopted controller are conducted.The experiments of the vibration control on the cantilevered plate using adaptive nonlinear,fuzzy tunning nonlinear and sliding mode predictive controllers were conducted.The experiments of the vibration control on the hinged plate using fuzzy neural network controller were conducted.The experimental results demonstrated that both for the low frequency mode vibrations of two flexible structure,the non-contact detecting method is effective,the designed controllers have greater attenuate performance,in comparison with proportional and derivative controller,particularly for the small-level residual vibration.