Motion Planning of Automatic Underwater Vehicle Based on 3D Dipolar Vector Field

Abstract

This paper investigates a 3-dimensional (3D) motion planning problem of an automatic underwater vehicle (AUV) in ocean environment with multiple stationary/moving spherical obstacles. To solve the 3D motion planning problem, a family of novel 3D analytic vector fields are employed, the representation of which depends on the selection of the integral curve according to the parameter λ. More specifically, when the value of λ is known, the vector field has a unique singularity of the dipole type and its integral curves are able to attract an AUV to a predefined target point. Furthermore, the vector field integral curve with continuous singular points enables the AUV to repel spherical obstacles. By blending the above two kinds of vector fields, a mixed vector field is generated in 3D space such that an AUV can simultaneously reach the target point and avoid spherical obstacles. The efficacy of the proposed 3D motion planning approach is demonstrated via simulation results.