Bio-inspired flow sensing and control: Autonomous rheotaxis using distributed pressure measurements

Abstract

This paper presents the design and use of an artificial lateral-line system for a bio-inspired robotic fish capable of autonomous flow-speed estimation and rheotaxis (the natural tendency of fish to orient upstream), using only flow-sensing information. We first present a feedback controller based on the difference between pressure measurements collected on opposite sides of the fish robot. We then describe a dynamic rheotaxis controller based on a potential-flow model and a Bayesian observer that uses two or more pressure sensors in an arbitrary arrangement. Pressure sensor placements are selected based on nonlinear observability analysis. Experimental results demonstrate the advantages of the proposed scheme, which include robustness to model error and sensor noise. The primary contribution of this paper is a framework for rheotaxis and flow-speed estimation based on pressure-difference information that does not require fitting model parameters to flow field conditions.