Safe Path Planning Method Based on Collision Prediction for Robotic Roadheader in Narrow Tunnels

Safe path planning is essential for the autonomous operation of robotic roadheader in narrow underground tunnels, where limited perception and the robot’s geometric constraints present significant challenges. Traditional path planning methods often fail to address these issues. This paper proposes a collision prediction-integrated path planning method tailored for robotic roadheader in confined environments. The method comprises two components: collision prediction and path planning. A collision prediction model based on artificial potential fields is developed, considering the non-convex shape of the roadheader and enhancing scalability. By utilizing tunnel design information, a composite potential field model is created for both obstacles and the roadheader, enabling real-time collision forecasting. The A* algorithm is modified to incorporate the robot’s motion constraints, using a segmented weighted heuristic function based on collision predictions. Path smoothness is achieved through Bézier curve smoothing. Experimental results in both obstacle-free and obstacle-laden scenarios show that the proposed method outperforms traditional approaches in terms of computational efficiency, path length, and smoothness, ensuring safe, efficient navigation in narrow tunnels.