Research on Map Building and Path Planning Algorithms for Autonomous Navigation Robots in Tunnel Scenarios
Han Kou, Huagang Liang, Rui Kang, Huixia Zhao · 2023
With the continuous advancement of highway tunnel construction, safety issues and inspection requirements in tunnel maintenance are increasing. In response to the problems of low efficiency and difficulty faced by traditional manual inspections, as well as the dilemma of inflexible deployment and high maintenance costs for traditional guided rail robots inspection, an improved Cartographer algorithm and improved A* fused with DWA algorithm for autonomous navigation of tunnel inspection robots are proposed in this paper. Firstly, the Cartographer algorithm is improved by introducing the Lazy Decision strategy and frontend optimization methods to enhance mapping accuracy. Secondly, a novel evaluation function for the A* algorithm is proposed, which incorporates relative cost factors and angle factors to better assess path quality. Meanwhile, a curve smoothing algorithm is utilized to optimize the path and improve its smoothness. Furthermore, a hybrid path planning algorithm is formed by combining the DWA to achieve autonomous navigation capability. Experimental results demonstrate that the improved Cartographer algorithm achieves a 1.66% increase in mapping accuracy, an 85.3% reduction in search time for the improved A* algorithm, an 81.4% reduction in traversed node count, and a 58.3% decrease in turning point count. The proposed algorithm meets the requirements of navigation accuracy, response time, and stability in tunnel inspection applications.