Hydrodynamic Parameter Estimation for an Anguilliform-inspired Robot

This paper presents an optimization-based approach to estimate the hydrodynamic parameters namely drag and added mass coefficients from free-running experiments conducted on an in-house developed Anguilliform-inspired robot. The objective of the optimization problem is to estimate the hydrodynamic parameters that minimize the differences between the trajectories obtained from the simulations and the physical experiments when operated for identical gait parameters and controller gains for both the straight and the turning motions. The hydrodynamic parameters obtained from the developed approach leads to a maximum root-mean-square (RMS) position error of 0.183 BL and a maximum RMS velocity error of 0.03 BL/s between the trajectories obtained from simulations and experiments. Experimental results suggest that the parameters estimated using the developed approach can be useful in predicting the robot’s motion accurately. Accurate robot motion prediction is the fundamental requirement for localization, collision prediction, and motion planning algorithms which in turn are required for automated inspection, maintenance, and repair of sub-sea structures using Anguilliform-inspired robots.