Author(s)

Slawomir Jaszczak & Andrzej Piegat

Abstract

The paper presents results of investigations on application of artificial intelli- gence in course control of the underwater vehicle (UV) KRAB II, being prop- erty of the Maritime Technology Faculty of the TU Szczecin. Classical design methods of control systems don't come up to expectations in case of such plants as an UV, which has a complicated, nonlinear, strongly varying mathematical motion model. This complication motivates scientists to apply the neuro-fuzzy technics giving the possibility of the learning from in- put/output data of a real system typical for neural networks and transparency of fuzzy logic. The paper presents the construction and the training method of the neuro- fuzzy PID controller. The controller has a rule base with 27 inference rules and 36 degrees of freedom. Such a comp

Keywords