Point-to-point iterative learning control with optimal tracking time allocation: a coordinate descent approach

Iterative learning control (ILC) is a high performance control technique for systems operating in a repetitive manner. A novel design methodology is developed in this paper to incorporate optimal tracking time allocation within the point-to-point ILC framework for discrete time systems. This leads to significant performance improvements compared to fixed time points (e.g. energy reduction). An optimization problem is formulated based on the point-to-point tracking requirement and the via-point temporal constraints. A two stage design framework is proposed to solve this problem, yielding an algorithm based on norm optimal ILC and the coordinate descent method, which automatically minimizes control effort while maintaining high performance tracking. The proposed algorithm is implemented on a gantry robot experimental test platform, with results verifying its practical effectiveness in the presence of model uncertainty