Dynamic Trajectory Tracking by Synergistic Dual-Stage Actuation and Control

• Published in: IEEE/ASME transactions on mechatronics Vol. 22; no. 6; pp. 2600 - 2610
• Main Authors: Chen, Cheng-Wei, Chang, Yen-Chi, Tsao, Tsu-Chin
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.12.2017; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Actuation; Actuators; Control theory; Dynamic characteristics; Dynamical systems; Engines; Feedforward control; Feedforward neural networks; Linear control; Linear synchronous motors; Linear systems; machine tool control; Machining; Master-slave; Mathematical models; mechatronics; Motion control; Nonlinear dynamics; Optimization; Parameter robustness; piezoelectric devices; Piezoelectricity; Predictive control; Real time; Repetitive control; Robust control; Robustness; Tracking control; Uncertainty
• DOI: 10.1109/TMECH.2017.2749510
• ISSN: 1083-4435

This paper addresses precision tracking control of dual-stage actuators for machining of noncircular engine piston profiles. To achieve desired tracking performance, the capacity limits, dynamic characteristics, external disturbances, and model uncertainties of the dual-stage actuators pose significant challenges in control design and real-time implementation. The master-slave linear control configuration previously applied to similar applications is inadequate when nonlinear dynamics are triggered by constraint violations. Optimal predictive control, which minimizes the tracking error subject to physical constraints, is desirable, but the substantial computation forbids real-time implementation at the sampling rate required by the application. A hybrid approach is proposed where an offline constrained optimal feedforward tracking utilizing the model-predictive control framework (FFMPC) is added to a two-parameter robust repetitive control (TPRRC) to achieve high-precision tracking performance, which would otherwise be unattainable solely by each individual control action. Multirate digital control implementation on a dual-stage actuator system, which is comprised of a long-stroke low-bandwidth linear motor and a short-stroke high-bandwidth piezoelectric actuator for tracking noncircular engine piston profiles is presented. The experimental results demonstrate the effectiveness of the approach, where the tracking error by TPRRC is reduced by 84.8 % when FFMPC is added.