Online Optimization-Based Time-Optimal Adaptive Robust Control of Linear Motors With Input and State Constraints

Linear motors are widely used in the manufacturing industry, in which time-optimal desired motion and high-accuracy motion tracking are both needed for higher productivity and better quality of product produced. However, the unavoidable uncertainties and input/state hard constraints in actual operations make the traditional separate treatment of desired motion planning and high-accuracy motion tracking control either too conservative or causing closed-loop instability. To address these issues, a two-layer control structure is proposed in this article. In the upper layer, a time-optimal control problem taking into account the system model and hard constraints is solved online instead of offline to make full use of the accurate parameter estimations and real-time knowledge of the system initial state of lower layer adaptive robust control (ARC) controller. By doing so, not only the conservativeness of traditional offline motion planning is overcome but also the closed-loop stability of overall system can be guaranteed through seamless integration of the lower layer ARC controller design and upper layer constrained optimization online planning. Comparative experiments conducted on a linear motor confirm the effectiveness of the proposed method.