Robust gait control for steady swimming of a carangiform fish robot

Abstract

The sinusoidal gait, which is commonly seen in the steady swimming of fish, is applied and controlled on a carangiform fish robot in this paper. Because of the unmodeled hydrodynamics and the uncertainties among parameters of mathematical model of fish tail, it is critical for the motorized tail system to track the sinusoidal gait functions in water environment. The periodical oscillation of the caudal fin results in continuously varying drag forces from water, which brings strong disturbance to the control system. Two methods are used to solve this problem. Firstly, the disturbance is treated as a linear velocity damping with a constant damping coefficient. A PD controller is applied to regulate the tracking system. However, this method shows limited capabilities when the frequency and amplitude of gait functions change. The second method is to incorporate the mathematical model of the disturbance into the controller design. The robust tracking is then provided in an extended error space. Compared with the result of PD control, the system response with the robust design shows superior performance in tracking sine input and reducing disturbance caused by the interaction force with water. Derivation of the equations for the controller design is presented. Experiment setup and results are also described.