Improved RRT Algorithm with Adaptive Step-Size and Hybrid Sampling for Ship Path Planning

To achieve more efficient and reliable ship path planning, this paper presents an enhanced version of the RRT* algorithm, addressing the issues of blind search, insufficient path smoothness, and redundant nodes found in traditional algorithms. First, the entropy weight method is employed to assign weights to objectives such as path smoothness, number of nodes, path angle, and path length, optimizing the evaluation function. Then, the algorithm introduces an adaptive step-size strategy, dynamically adjusting the step size based on the environment, and combines uniform and Gaussian distribution sampling methods to enhance search efficiency and path quality. Finally, the path is smoothed using bidirectional search pruning and Hermite interpolation. Experimental simulations reveal that the improved algorithm surpasses conventional approaches in path length, smoothness, and safety, significantly enhancing the efficiency and effectiveness of ship path planning.