关键词: 无人运输车, 路径规划, A*算法改进, 算法融合

Abstract: Aiming at the path planning requirements of unmanned transport vehicles on the decks of large ships in complex static and dynamic environments, A collaborative path planning method for unmanned transport based on the improved A* algorithm is proposed. Optimize the traditional A* algorithm at the global planning level by simplifying the search direction and reducing the number of node expansions in narrow areas. A dynamic weighted evaluation function is designed based on the density of the ship deck to achieve adaptive planning of the ship deck area under different working conditions. Propose a hierarchical safety expansion strategy, and set the expansion radius of obstacles in different areas differently according to the danger level. By integrating the long-span linear optimization method with the third-order Bezier curve smoothing, path simplification is achieved through cross-node obstacle detection. At the local planning level, the evaluation function of the dynamic window algorithm is improved, the regional density penalty term is introduced, and finally a global-local collaborative method is formed. The simulation results and real vehicle tests in typical scenarios show that the unmanned transport vehicle has the optimal movement path on the deck and has achieved obstacle avoidance.

Key words: Unmanned transport vehicle, Path planning, Algorithm improvement of A* ;Algorithm fusion