Ship Collision Avoidance Decision-Making Using an Improved Black Widow Optimization Algorithm

Abstract

In order to enhance intelligence and acquire the effective avoidance angles of the ship collision avoidance system, an improved decision-making scheme for ship collision avoidance is proposed by using an improved swarm optimization algorithm and random linear weighting method. Firstly, the key ship avoidance strategy is employed for analyzing the factors involved in ship collision risk. Then, taking the Convention on the International Regulations for Preventing Collisions at Sea (COLREGs), the risk of ship collision, and voyage loss into consideration, a multi-objective function optimization technique based on a random linear weighting method is proposed to overcome the influence of subjective weight assignment. Finally, an improved black widow optimization algorithm with novel position strategies is employed to determine the optimal course alteration for ship collision avoidance. The feasibility and effectiveness of the proposed ship collision avoidance scheme are demonstrated by the ship's intelligent collision avoidance platform. The proposed collision avoidance scheme may provide a reasonable avoidance angle as a reference for the officer on watch in the practice of navigation.