Multi-Objective Optimization Based Collision Avoidance Algorithm for an Intelligence Marine Navigation
Hamed Faroqi, A. Sadeghi Niaraki · Journal of Applied Sciences · 2015
Currently, the collision avoidance algorithm is considered as an essential part of the marine navigation systems.The collision avoidance algorithm determines and decides to change the speed and direction of a ship in the presence of obstacles.Avoidance of collision problem can be considered as a spatial multi objective problem due to it needs to handle with different objectives by searching a space of possible routes and find a set of optimal routes.Solving collision avoidance problem in ship's routing is a complicated problem because of different objectives, inconsistency between them and presence of dynamic and static obstacles in sea.In this study, the Multi-Objective Evolutionary Algorithm (MOEA) as an optimization technique has been used for solving the problem in the context of GIS.Two objectives, highest safety and lowest deviation from main route, have been considered for simulating and solving the problem.The proposed algorithm has been implemented for solving the problem in five simulated situations, including various modes of dynamic and static obstacles.The results have been analytically and numerically evaluated.Determination of the optimal routes set, suitable rapidity of the algorithm (about 40 sec), efficient convergence trend and repeatability of results (80%) are the positive and promising consequences of evaluating the algorithm's results.