Combined Plant/Controller Optimization with Applications to Autonomous Underwater Vehicles (AUVS)

Abstract This paper describes a solution to the following plant-controller optimization (PCO) problem: given an autonomous underwater vehicle (AUV) - with a fixed baseline body configuration - that is required to operate over a finite number of representative trimming conditions in the vertical plane, determine the optimal size of the bow and stern control surfaces so that a weighted average J of the power required at trimming is minimized, subject to the conditions that: i) a given set of open loop requirements are met, and ii) stabilizing feedback controllers can be designed to meet desired time and frequency closed loop performance requirements about each trimming point. The solution proposed is rooted in the theory of Linear Matrix Inequalities (LMIs) and leads to efficient PCO algorithms that build on a recently released LMI Toolbox.