Search

Abstract The vibration isolation system is crucial to high-precision space systems. This paper studies dynamics and control of a six-axis vibration isolator via a flexible Stewart platform. The parasitic stiffness induced by flexible joints is considered in dynamics modeling and compensated in control for the first time. The explicit dynamic equations are established based on the pseudo-rigid-body model and the principle of virtual power. After validation by ADAMS/Flex and the finite element method, the dynamic equations are used for designing a decoupled controller. The control force is synthesized based on the leg's force and position feedback. The impact of bending and torsional stiffness of flexible joints is compensated and the MIMO system is consequently decoupled in modal space, making parameter design and performance implementation more convenient and effective. The identical sub-controller for every SISO system makes the six-DOF isolator achieve the capability of vibration isolation simultaneously for the six modes. With a proportional plus integral compensator as the sub-controller, the vibration isolation bandwidth and active damping can be regulated separately. [ABSTRACT FROM AUTHOR]

Parallel piezostages have been prevalently applied in the micro-/nano-positioning field, but, unluckily, suffer from the undesired effects, like creep, hysteresis, vibration, cross coupling, and input saturation. To eliminate these nonlinearities, a new discrete-time sliding mode predictive control (DSMPC) with decoupling and damping properties (DSMPC-DDP) is proposed for a parallel six-axis piezostage devoted to cell micromanipulation. Firstly, a state-space-based decoupler is applied to make the MIMO system decoupled into six SISO subsystems, and then a damping compensator via the notch filter principle is selected to suppress the high-frequency vibration dynamics of six subsystems. After that, a feedback DSMPC law is designed for each decoupled and damped subsystem. Unlike the traditional linear model predictive control (MPC), the feedback DSMPC is developed based on a nonlinear discrete-time terminal sliding mode prediction model, leading to stronger robustness and faster convergence speed. Through the utilization of the receding horizon optimization, the sliding mode states are optimized to explicitly deal with the input constraints. The stability of the closed-loop subsystems is analyzed based on the terminal constraint theory. Finally, the effectiveness and superiority of the proposed controller is validated by a series of experimental tests. Results show that the proposed DSMPC-DDP scheme can achieve higher tracking accuracy and control bandwidth than the existing proportional–integral (PI) and MPC methods in the context of abovementioned nonlinearities. • A new DSMPC-DDP is proposed for a parallel six-axis piezostage. • Attenuating the creep, hysteresis, cross coupling, vibration, and input saturation simultaneously. • Providing rigorous stability analysis for the closed-loop control system. • Comprehensive comparative experiments are conducted for validating the superiority. [ABSTRACT FROM AUTHOR]