LAUSR

This article proposes a new data-driven optimization of integrated control for flexible systems to achieve high-performance automatic control. The integrated control that is composed of feedback control, feedforward control, and disturbance observer is adopted in this article as the control framework that can effectively address the control problems of the flexible system. However, it is difficult to optimize all the parameters of the integrated control, because the number of the parameters to be optimized is larger than the conventional feedback control, which complicates the optimization procedure. In this article, the optimization procedure of the integrated control as well as the mathematical background of it is proposed. At first, the closed-loop characteristics of the integrated control are analyzed and its convexity with respect to control parameters is theoretically investigated. The proposed optimization method is designed taking into consideration the convexity of the control configuration to guarantee the global optimality of the obtained parameters. Moreover, the proposed method can simultaneously optimize all the parameters of the integrated controller based on the experimental data. The effectiveness of the proposed algorithm is experimentally confirmed using a flexible system under the following two conditions: first, change of initial parameters and second, change of plant conditions.