Scholarlink

<p>To overcome the limitations of existing path planning methods in handling dynamic obstacles and sudden disturbances, this paper introduces a hierarchical model predictive control (MPC) framework for intelligent vehicle obstacle avoidance. The strategy adopts a dual-layer control structure: a global planner uses the Hybrid A* algorithm to generate a kinematically feasible reference trajectory with continuous curvature, while a local controller employs a particle filter (PF) to predict the future motions and uncertainties of dynamic obstacles in real-time. These predictions are formulated into spatiotemporal constraints with probabilistic safety bounds, which are embedded in an MPC-based rolling optimization process. This framework coordinates trajectory tracking with vehicle dynamics, maintaining safe distances from obstacles while improving tracking accuracy and efficiency. Co-simulation results in CarSim/Simulink validate that the proposed approach enhances obstacle avoidance performance, driving stability, and ride comfort in complex scenarios. Moreover, it outperforms conventional methods in terms of adaptability, smoothness, and safety.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Dynamic Obstacle Avoidance Strategy for Intelligent Vehicles with Hierarchical Control and Model Predictive Control

  • 期刊论文

  • 发表日期：

    2026年3月31日

查看全文
  • Gang Liu,
  • Wen Wen Shen,
  • Bo Ding

摘要

To overcome the limitations of existing path planning methods in handling dynamic obstacles and sudden disturbances, this paper introduces a hierarchical model predictive control (MPC) framework for intelligent vehicle obstacle avoidance. The strategy adopts a dual-layer control structure: a global planner uses the Hybrid A* algorithm to generate a kinematically feasible reference trajectory with continuous curvature, while a local controller employs a particle filter (PF) to predict the future motions and uncertainties of dynamic obstacles in real-time. These predictions are formulated into spatiotemporal constraints with probabilistic safety bounds, which are embedded in an MPC-based rolling optimization process. This framework coordinates trajectory tracking with vehicle dynamics, maintaining safe distances from obstacles while improving tracking accuracy and efficiency. Co-simulation results in CarSim/Simulink validate that the proposed approach enhances obstacle avoidance performance, driving stability, and ride comfort in complex scenarios. Moreover, it outperforms conventional methods in terms of adaptability, smoothness, and safety.