An Integrated Path Planning Framework for Multi-Obstacle Avoidance of the Multi-Axle Autonomous Vehicle With Enhanced Safety and Stability

The autonomous path planning for obstacle avoidance has garnered great interest of numerous researchers recently. In this field, the multi-obstacle issue is a challenging but critical topic that needs to be adequately investigated, especially for the autonomous heavy vehicle. An integrated path planning framework comprising a key reference target risk evaluator (KRE), a B-spline path optimizer (BPO) and a multidimensional constraint set (MCS) is proposed for a multi-axle distributed autonomous vehicle. In the KRE, the key reference target is recognized for risk classification, and the space to collision is constructed for lane changing feasibility judgement. Meanwhile, the probability-based vehicle ellipse is designed for influence range reflection. In the BPO, the path generation-optimization integration algorithm is designed based on B-spline parameterized method, by which the optimal path with smoothness and response is determined continuously with real-time planning objectives. To improve the comprehensive performance of the optimal path, an MCS considering the collision field, the planning corridor and ego vehicle dynamic is formulated, which guarantees the anti-collision ability and the spatial realizability, and enhances the vehicle dynamic properties involving anti-sideslip, rollover prevention, yaw stability and tire wear alleviation. Finally, the simulation and HIL test platform are established and the validation is conducted in different cases, which proves the effectiveness and real-time capability of the proposed framework.