Model predictive control of precision stages with nonlinear friction

In the manufacturing industry especially in the semiconductor, the requirement for high-speed, fast-response and high-precision performances is critical. In this paper, the model predictive control is introduced to the positioning control of ultra-precision stage driven by a linear actuator. The linear actuator is the non-resonant ultra-sonic motor (SPIDER) which has the strong non-linearity due to the friction drive mechanism. Therefore, the control object is at first compensated by the friction control based on the bang-bang control. Next, the plant considering the friction is identified by the system identification strategy. Then, the model predictive control is applied to the identified plant. The effectiveness of the application of the model predictive control to the ultra-precision stage as well as the proposed control approach is verified through the simulation and experiments.