Abstract: This study aims to develop an enhanced Dijkstra algorithm to optimize navigation path planningfor unmanned shovel loaders in underground mining environments. In the traditional Dijkstra algorithmthe shortest path tree is generated by continuously selecting the nearest unvisited node from the startingpoint. However, in underground mining environments, additional factors need to be considered, such asterrain, mineral distribution, and safety factors. By deriving a new kinematic model and incorporating itinto the algorithm , the improved algorithm can provide safer and more eficient path planning. Throughpractical case studies, this paper demonstrates the efectiveness and potential applications of the enhancedDijkstra algorithm in underground mining navigation.

Key words: unmanned driving, path planning, kinematic model