Identification of Hydrodynamic Parameters for a Remotely Operated Vehicle Using Projective Mapping Method

Abstract

This study proposed a sequential procedure for identifying all hydrodynamic parameters of a remotely operated vehicle (ROV). A simplified nonlinear ROV dynamic model with twelve unknown hydrodynamic parameters in six degrees of freedom was used for simulations. Uncoupled motions such as surge, heave, sway, yaw, and coupling surge and sway were obtained from ROV dynamic model to reduce number of parameters needed to be estimated simultaneously. A vision-based system was utilized to capture time-series images of ROV planar simple motions and an onboard pressure sensor was used to record ROV heave motion. Based on observations of captured time-series images, ROV planar movements were resolved by projective mapping. Then, hydrodynamic parameters were estimated with numerical optimizations by comparing the responses of the ROV and the mathematical model to the same inputs. Sequential identifications were performed, from the simple to the complex, in the order of first surge, next heave, yaw, then sway, and finally coupling surge and sway motions, and the results were satisfactory. As well, the results have been shown that the parameters identified from the coupling surge and sway motion have worse accuracy than those identified from single-degree-of-freedom motions.