An improvement on multivariable PID controller design via iterative LMI approach. (Q1428150)

This note improves on a recent result on multivariable proportional integral derivative (PID) control design via convex linear matrix inequality (LMI) optimization. Building on the original idea, which is to transform the multivariable PID design problem into a static output feedback design problem, the note removes a technical invertibility constraint and simplifies the design procedure. This can also be extended to uncertain systems and robust PID control. PID design, like static output feedback design, boils down to nonconvex nonlinear matrix inequality optimization, so that a heuristic based on solving iteratively a sequence of convex LMI problems must be used, without guarantee of convergence.