Motion Planning and Control of Artificial Muscle Actuated Underwater Vehicle Using Nonlinear Neural Oscillator

Abstract

In this paper, we introduce the motion planning and control strategy for the underwater vehicle actuated by a soft artificial muscle actuator. The artificial muscle used for this underwater application is an Ionic Polymer Metal Composite (IPMC) which can generate bending motion in aquatic environments. In this research, the double ring structured nonlinear neural oscillator is proposed for the undulatory motion in the actuator. The overall dynamic model of the flexible IPMC actuator including its fluid interaction terms is used for the motion planning and open-loop controller design. The IPMC used in this study is a patterned or segmented type where the electrode surface of the actuator is encoded such that each segment can be controlled independently for effectively generating an undulatory motion in the water. Computer simulations show that the proposed neural oscillator based controller can be effectively used for the underwater locomotion applications, and can be extended to the closed-loop controller where the precise maneuver is needed in the unstructured aquatic environments.