Bionic Sensing with Artificial Lateral Line Systems for Fish-Like Underwater Robots

In this book, the authors first introduce two fish-like underwater robots, including a multiple fins-actuated robotic fish and a caudal fin-actuated robotic fish with a barycenter regulating mechanism. They study how a robotic fish uses its onboard pressure sensor arrays based-ALLS to estimate its trajectory in multiple locomotions, including rectilinear motion, turning motion, ascending motion, and spiral motion. In addition, they also explore the ALLS-based relative position and attitude perception between two robotic fish in a leader-follower formation. Four regression methods-multiple linear regression methods, support vector regressions, back propagation neural networks, and random forest methods-are used to evaluate the relative positions or attitudes using the ALLS data.

The research on ALLS-based local sensing between two adjacent fish robots extends current research from one individual underwater robot to two robots in formation, and will attract increasing attention