TCRRT: A Novel Path Planning Approach of Field Robots for Optimizing the Traversal Cost in Diversified Terrains

Abstract

The identification of feasible regions is crucial for navigation and path planning of robots. When robots travel across different terrains, their speeds and energy consumption vary dramatically. Considering the different energy costs in various terrains, this study proposes a novel path planning strategy for robots that work in diversified complex terrains. First, the energy cost is evaluated under different terrains such as concrete, wood, and grass. Then, a terrain cost map is designed to record the difference in energy costs, in which the terrain is recognized and segmented by a light network Psp-MobileNet. According to the energy cost of terrains, a path planning approach named Terrain Cost Rapidly-exploring Random Trees (TCRRT) is proposed to optimize the traversal cost. In the above approach, the cost function could balance the energy cost and distance in various terrains and generate a dynamically feasible and low-cost path. Finally, the effectiveness of the TCRRT algorithm is verified by simulation and real-world navigation experiments on diversified terrains. © 2025 Institute of Electrical Engineers of Japan and Wiley Periodicals LLC.