ON COMPLEXITY ISSUES IN MULTIOBJECTIVE CONTROLLER DESIGN USING CONVEX OPTIMIZATION

In multiobjective control of linear systems, critical issues are the size of optimization constraints, number of variables, and computation time. This paper investigates in a quantitative way, how some prominent methods from the literature behave with respect to these critical factors for different problem sizes. Moreover, an alternative formulation is proposed and compared to the existing approaches, showing its favorable properties for a selection of typical applications.