NAVIGATION AND PATH-PLANNING OF MICRO-AUTONOMOUS UNDERWATER VEHICLES IN ENCLOSED AND CLUTTERED UNDERWATER ENVIRONMENTS

Abstract

Deploying a low-cost, micro autonomous underwater vehicle (µAUV) remains a challenge in modern underwater robotics. In the nuclear industry, there is a desire for µAUVs to replace people in the process of gathering information from cluttered layouts and take measurements of pH, radioactivity level, temperature and other relative parameters. The key challenges in successfully deploying µAUVs in spent fuel storage ponds are underwater navigation and collision-free path-planning. This thesis developed a solution for these challenges. In this thesis, a navigation approach, that combines storage pond map acquisition and path-planning, is proposed. The proposed approach integrates environment surveys, environment reconstruction, and path-planning as a complete task that involves learning and planning. An echo-sounding based aerial mapping is developed to gather the topological height map of the storage pond; an efficient map construction method is developed to process the obtained raw depth measurements data to establish a 3D grid map; the A* algorithm is evaluated to be used in the 3D grid map of the cluttered pond environment.

Date of Award	31 Dec 2019
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Peter Green (Supervisor) & Joaquin Carrasco Gomez (Supervisor)