Optimization-Based Trajectory Planning for Autonomous Driving at Mining Area with Irregular Boundary

Autonomous driving technology has been widely used in mining areas, and how to get a safe, smooth and easy-to-track path has gradually become the focus of research. Currently, most planners are oriented to structured roads and perform poorly in unstructured road scenarios. Due to the irregular and variable boundary geometry of the mining scene, the commonly used solution is to decouple the trajectory planning problem into path planning and the speed planning problem. In other words, the planning module first performs path planning and then finds the corresponding speed profile as the final result, which creates a path tracking task for the control module. This decoupling method may lose the optimality of the trajectory due to the problem of the matching degree between the path solution and the speed solution. In this paper, we present a trajectory planning method based on the idea of optimal control problem (OCP). Firstly, the hybrid A* method is used to obtain rough paths and convert them into rough trajectories based on the assumption of constant speed in the mining area. Then, an optimal control problem (OCP) is established by taking the state quantity and control quantity of the vehicle as the optimization objective, and it is discretized into a nonlinear problem (NLP). In order to solve the problem efficiently, the vehicle kinematic constraints are considered as the penalty term in the objective function. The non-convex collision avoidance constraints are transformed into convex terms by constructing a corridor. At the same time, a soft collision avoidance constraint is set as a penalty term to avoid the loss of path smoothness due to too conservative hard constraints. Finally, the proposed method is validated in two typical mining scenarios: loading and road driving areas. The results show that the proposed method enables finding a trajectory within the safety margin, and considerably improves the curvature and its rate.