Concurrent H2/H∞feedback control design with optimal sensor and actuator selection

摘要

This paper presents a systematic approach to con-currently design H₂/H_∞dynamic output feedback controllers and find the optimal number of sensors and actuators for continuous linear time-invariant models of complex multi-variate systems. The method presented here formulates this concurrent design and selection as a multi-objective optimization problem where ℓ₁norm regularization is performed on the controller input and output matrices. Two case studies illustrate the potential of the developed approach. Preliminary results, obtained from the two case studies, justify the interplay of obtaining a trade-off between the enhanced closed-loop performance and the implementation cost.