Real-time planning method of turning path for IGVs based on virtual obstacle

In order to improve the safety of intelligent guided vehicles (IGVs) operating on curves approaching railway crossings, this paper proposes a real-time planning method for turning paths based on cubic Bezier curves. Firstly, a virtual obstacle for path planning was preset ahead of a crossing, based on the global path and road conditions. Then, the information of the virtual obstacle is used to define the selection range of control points for Bezier curves, resulting in a cluster of turning paths that satisfied the curvature constraints. Finally, the minimum curvature of the path points was used as the criterion to identify the optimal turning path. In the simulations, the proposed method generated smooth turning paths, with the maximum curvature of the path points not exceeding 0.12 m^-1, which satisfies the turning radius requirements for IGV operation. Additionally, these results can effectively support the attitude adjustment of IGVs before entering railway crossings, thereby enhancing operational safety.