Leader-following formation control of multiple vision-based autonomous robotic fish

Abstract

A framework for the deployment of multiple autonomous robotic fish to achieve leader-following formations with Bezier trajectory is presented. Each follower robot estimates the position and orientation angle of its leader with a fast color-tracking vision system, and establishes a Bezier trajectory between its current position and the position of its leader robot. Considering the nonholonomic properties of the robotic fish, the optimization of Bezier curve's curvature to choose appropriate scale factor is conducted by combination of penalty function and one-dimensional search methods to perform a smooth and stable trajectory. The optimal trajectories are accurate enough to estimate the angular velocity of the follower robot, while a fuzzy controller is used to adjust its linear velocity. By the introduction of virtual leaders, formations of different shapes can be generated. Experimental results show the effectiveness of our approach.