Study on collision avoidance methods used for unmanned surface vehicle

• Main Author: Adithya, Malyavantam Ramesh
• Other Authors: Cheah Chien Chern
• Format: Theses and Dissertations
• Language: English
• Published: 2019
• Subjects: Engineering::Electrical and electronic engineering::Control and instrumentation::Robotics
• Online Access: http://hdl.handle.net/10356/78905
• Institution: Nanyang Technological University

Unmanned surface vehicles (USVs) are intelligent Marine vehicles which can navigate autonomously on the water with minimal human interaction such that it can complete its assigned tasks by itself. For effective navigation of the USV, it requires modules such as obstacle detection, obstacle avoidance, path navigation and other mission-related equipment on board. Trajectory planning in navigation can be classified as global path planning and local path planning approaches. It is also very important that the USV manoeuvre in such a way that it follows navigation rules on sea hence The Coastguard Regulations on Prevention of Collision at Sea (COLREGs) guidelines must be implemented within the approaches. An attempt to explain recent developments in available collision control approaches are discussed in this thesis report. You would be able to see methods involving mathematical approaches, optimization problem approach and intelligent approaches. Some of the Different methods explained here are artificial potential field method which constructs an artificial potential between the USV and its environment such that it is attracted to the object and repelled by the obstacles. Metaheuristic methods which are a type of optimization approach that can solve the problem quickly by searching over large solutions. Lastly, a method based on deep neural network is discussed in object avoidance of the USV, this method is still in the development stage but worth to know because of the potential of its applications. Simulation is performed on simple cases with the use of potential field method and optimization method and the results are compared. The inference point towards the advantages of combining two methods. The negatives of the optimization method were overcome by the artificial potential field method when the two were combined. Some of the Inference made on the methods available are, the USV is operated in a dynamic environment with unpredictable encounter situations, the weather conditions play a strong role in the completion of USV mission, but these are not considered in the discussed methods.