【摘要】 本文提出一种融合全局规划、局部避障与轨迹跟踪的车辆自主避障方法，以应对静态与动态障碍物共存的复杂行驶环境。首先，采用改进的跳点搜索算法（Jump Point Search, JPS）生成初始路径，并结合优化的贝塞尔曲线平滑处理，获得高连续性的参考轨迹；其次，设计一种改进的人工势场法（Artificial Potential Field, APF），使其能够根据障碍物状态动态调整路径，并引入量子退火算法高效规避传统人工势场法易陷入局部最优的问题；进一步地，将该势场信息嵌入模型预测控制（Model Predictive Control, MPC）的约束条件与代价函数中，实现轨迹规划与跟踪的一体化协同优化。最后，在MATLAB/Simulink与CarSim联合仿真平台上开展多场景测试。结果表明，所提方法有效克服了单一算法在鲁棒性、实时性或安全性方面的局限，显著提升了车辆在复杂环境中的运行安全性与轨迹稳定性。更多还原

【Abstract】 This paper proposes an integrated autonomous vehicle obstacle avoidance method that combines global path planning, local obstacle avoidance, and trajectory tracking to address complex driving scenarios involving both static and dynamic obstacles.First,an improved Jump Point Search(JPS) algorithm is employed to generate an initial path, which is subsequently smoothed using an optimized Bézier curve to produce a highly continuous reference trajectory. Second, an enhanced Artificial Potential Field(APF) method is developed,capable of dynamically adjusting the path based on obstacle states; moreover,quantum annealing is introduced to efficiently mitigate the risk of local minima inherent in conventional APF approaches. Furthermore, the potential field information is embedded into both the constraints and cost function of Model Predictive Control(MPC), enabling unified and cooperative optimization of trajectory planning and tracking. Finally,multi-scenario simulations are conducted on a MATLAB/Simulink—CarSim co-simulation platform. Results demonstrate that the proposed method effectively overcomes the limitations of single-algorithm approaches in terms of robustness,realtime performance, or safety, significantly enhancing vehicle operational safety and trajectory stability in complex environments.更多还原