A hybrid enhanced A* and DWA approach for dynamic path planning of manufacturing logistics robots

Abstract

Smart manufacturing and logistics robots require integrated navigation capabilities. Optimized global routing, fast obstacle avoidance, and stable trajectory generation are all necessary. In this paper, we develop a hybrid path planner. An enhanced A* algorithm is combined with the Dynamic Window Approach (DWA). We intend to satisfy such practical demands. Four key aspects are addressed to optimize the conventional A* algorithm. Evaluation function optimization, neighborhood search optimization, Floyd based path smoothing, and segmented path processing. Path smoothing is also enhanced. These modifications cut computational costs. Invalid node expansion is reduced. Sudden steering changes are mitigated as well. Comparative simulations on two test maps were carried out. The superiority of our improved method was validated. The results show that path turns drop by 33.3% and 37.5%, respectively. Search time is reduced by 22.2% and 31.3%, respectively. Unlike pure static planning strategies, the hybrid framework delivers faster system response. Improved real-time obstacle avoidance capability is also achieved. Consequently, it fits well with practical industrial scenarios. Automated material handling and intelligent warehouse logistics in smart factories are covered.