On a robust PID controller parameters optimization applied to a mass-spring system

摘要

H?? synthesis is often used to realize robust and high performance controllers. This synthesis allows to have full order H?? controller. The ability to formulate it as a convex optimization problem makes it relatively easy to solve using semi-definite programming techniques. The H?? controller order is often much higher than that of the plant model. However, implementing such controllers for high order systems is a rather complex task. Hence, the reduced order H?? controller synthesis is becoming of great interest for practical applications. Contrary to the full-order H?? controller synthesis case, the H?? optimization problem of static output feedback controllers cannot be reparameterized as a convex optimization problem. The cone complementarity linearization algorithm is used to overcome the nonconvexity problem due to the constraint on the controller order. The proposed algorithm is applied to the tuning of SISO (Single-Input Single-Output) PID (Proportional plus Integral plus Derivative) controller parameters for the control of a mass-spring system.