Plant / Controller Optimization with applications to Integrated Surface Sizing and Feedback Controller Design for Autonomous Underwater Vehicles (AUVs)

摘要

This paper describes a solution to the following plant controlleroptimization (PCO) problem: given an autonomousunderwater vehicle (AUV) - with a fixed baselinebody configuration - that is required to operate overa finite number of representative trimming conditions inthe vertical plane, determine the optimal size of the bowand stern control surfaces so that a weighted average Jof the power required at trimming is minimized, subjectto the conditions that: i) a given set of open looprequirements are met, and ii) stabilizing feedback controllerscan be designed to meet desired time and frequencyclosed loop performance requirements about eachtrimming point. The solution proposed is rooted in thetheory of Linear Matrix Inequalities (LMIs) and leadsto efficient PCO algorithms that build on a recently releasedLMI Toolbox.