Multivariable intelligent control strategies for an autonomous underwater vehicle

Abstract

This paper discusses the application of a novel multivariable control technique to the problem of autonomous underwater vehicle (AUV) autopilot design. Based on an adaptive network structure a multivariable Sugeno style fuzzy inference system is tuned to produce autopilots for simultaneous control of multiple degrees of freedom for an AUV. Simulation results illustrate the effectiveness of this new multivariable approach for course-changing roll-regulating, and y-positional-changing yaw-regulating control when compared with a traditional multi-input single-output control approach whereby control of each degree of freedom is considered separately and hence no provision is made for the inherent cross-coupling between AUV channel motions. © 1999 Taylor & Francis Group, LLC.