Identification of Hydrodynamic Coefficients of a Robotic Fish Using Improved Extended Kalman Filter

Abstract

Bio-inspired robotic fish have received an increasingly rapid development in recent years due to their advanced performances, such as high energy efficiency and high maneuverability. An accurate dynamic model is essential to the design and control of such robots. Hydrodynamic co-efficients play an important role in modeling the robot, which are usually obtained from theoretical calculation or water tunnel experiments. This paper proposes a novel method for hydrodynamic coefficients identification using an improved Kalman filter with angular velocity and distributed pressure measurements, which are typically available from the robot’s on-board sensors. Simulation based on a Joukowski airfoil shaped robotic fish demonstrates the proposed method.