The University of British Columbia

This work considers control of sampled distributed parameter systems. These large-scale systems have a specific structure which allows for simplification of robust controller design compared to the general MIMO system case. This paper introduces and analyzes a controller structure which has been developed expressly for these systems. The controller design and analysis is performed in terms of the dynamic frequencies and also the spatial Fourier components of the process response. The proposed controller tuning strategy uses frequency loop shaping ideas which are applied in the spatial variable domain. The designed controller is proven to be robustly stable for a wide range of model uncertainty structures commonly occurring in practical systems.