A path planning method for collision avoidance of ships in complex sea areas

Abstract

At present, applying the existing path planning methods of collision avoidance of ships to complex sea areas is difficult. This paper presents a path-planning method that effectively considers collision avoidance in complex sea areas. First, the K-means method is used to divide the collision avoidance of ships into different stages with a reasonable risk discrimination index. Second, the track deduction is conducted in accordance with the Nomoto model to predict the track of the ships after a certain time. The predicted tracks with collision risk and violation of the collision avoidance rules in the collision avoidance stage are eliminated, and the track with the lowest risk is selected as the optimal predicted track in the other predicted tracks; Finally, the selection function is established in the normal navigation stage to select the optimal predicted track, and all the optimal predicted tracks are composed to generate the path. Experimental results show the following. 1) Compared with the current advanced improved artificial potential field method, the proposed method can effectively improve the efficiency of planning in a simple environment with a spot obstruction area. 2) The proposed method can consider the course change of incoming ships and plan the collision avoidance path in line with the International Regulations for Preventing Collisions at Sea (COLREGS) in remarkably complex sea areas with various navigation obstruction areas.