运动规划
模型预测控制
路径(计算)
启发式
计算机科学
控制器(灌溉)
理论(学习稳定性)
运动学
跟踪(教育)
控制理论(社会学)
控制工程
算法
任意角度路径规划
车辆动力学
跟踪系统
控制(管理)
工程类
质量(理念)
弹道
实时计算
控制系统
智能控制
系统动力学
作者
Fei Lai,Chaoqun Huang,Junhao Sun
标识
DOI:10.1177/09544070251389255
摘要
This study addresses the challenges of insufficient real-time path planning and limited tracking accuracy for intelligent vehicles in complex dynamic environments by proposing a novel path planning and tracking control framework. The framework integrates an improved A* algorithm for path planning and a model predictive control (MPC) algorithm for tracking control. First, to overcome the limitations of the traditional Hybrid A* algorithm, we introduce the Direction-Hybrid A* algorithm, which incorporates directionality, optimizes the cost function, and refines the heuristic map function. These enhancements significantly improve the efficiency and quality of path planning. Second, for path tracking control, we design a controller based on the MPC algorithm. We develop both a kinematic coupled model predictive controller and a linear dynamics model predictive controller, each tailored to different driving conditions: the former for low-speed scenarios and the latter for medium- and high-speed scenarios. Finally, simulation and real-vehicle experiments confirm that the proposed algorithms achieve shorter planning times and more efficient path planning in low-speed scenarios. They also demonstrate improved stability in path tracking and reduced lateral deviation in medium- and high-speed scenarios. Overall, the improved algorithms significantly enhance the path planning and tracking control performance of intelligent vehicles.
科研通智能强力驱动
Strongly Powered by AbleSci AI