A Route Planning Method for Intelligent Ships Based on the AS-DDQN Algorithm

Abstract

In view of the problems of the traditional Whale Optimization Algorithm (WOA) when dealing with ship path planning tasks, such as easy to fall into local optimum and weak later search ability, an improved Whale Optimization Algorithm (IWOA) was proposed. Firstly, the algorithm introduces Tent chaotic map to initialize the population, enrich the diversity of the population, and enhance the global search ability. Secondly, the stochastic differentiation algorithm is introduced to update the individual position through the cross-mutation operation, retain the optimal solution, and improve the convergence speed and robustness. Then, the Levy flight strategy is introduced, which combines random step size and stagnation disturbance mechanism to balance long-distance and short-distance search and avoid local optimum. Finally, combined with the fused ship motion constraints (path smoothness, maximum steering angle, safety margin), etc., the ship path planning is carried out in the rasterized nautical chart. Simulation experiments show that the improved whale optimization algorithm improves 6% and 34% in terms of path length and running time compared with the traditional whale optimization algorithm, which can better meet the needs of ships in complex navigation environment.