CONSTRAINED OPTIMIZATION ALGORITHMS TOWARDS INTEGRATED DESIGN OF MECHATRONIC SYSTEMS

Abstract

Integrated design of mechatronic systems that concurrently considers both control and mechanical specifications is a promising trend to achieve high performance for next generation of the mechatronics. However, the optimal mechatronic design by taking all the subsystems as a whole leads to a composite gain matrix which is under certain structural constraints, and this problem cannot be solved by the standard optimal control methods. This thesis presents several constrained optimization algorithms for a class of integrated mechatronic design problems, including a constrained linear quadratic optimization algorithm, a constrained H2 optimization algorithm, and a cutting-plane-based optimization algorithm for the H2 guaranteed cost control problem. A flexure-linked dual-drive H-gantry is used as a case study to validate the theoretical results and the effectiveness of all the proposed algorithms.