复杂动态场景下融合改进A*-DWA移动机器人路径规划

Path Planning of Mobile Robots via Combining Improved A*-DWA Algorithms in Complex Dynamic Scenarios

摘要:

植物工厂场景变化多、运行空间狭窄、障碍物不规则，该复杂动态环境给移动机器人路径规划带来较大挑战。为提高路径规划效率与安全性，本文提出一种融合改进A*和动态窗口法(Dynamic window approach，DWA)的路径规划方法。面向栅格地图提出一种包含邻域节点选择-评价函数优化-路径质量精化策略的改进A*算法，将剩余距离因子与安全距离约束融入路径规划代价函数，优化全局路径生成机制，提升路径安全性并减少拐点数量。针对移动机器人动态避障，改进DWA速度评价函数，新增全局导引指标项并设计权重自适应调整策略，实现了动态避障安全与全局规划效率协调。最后开展移动机器人仿真对比试验及实景验证试验，试验结果表明本文路径规划方法在有效性与安全性方面具备较大优势。

Abstract:

The dynamic and complex environment of plant factories, characterized by frequent scene changes, narrow operational spaces, and irregular obstacles, poses significant challenges to the path planning of mobile robots. To enhance both the efficiency and safety of navigation in such environments, a hybrid path planning method that integrated an improved A* algorithm with the dynamic window approach (DWA) was proposed. Firstly, an enhanced A* algorithm was developed for grid-based maps by incorporating a neighborhood node selection strategy, an optimized evaluation function, and a path quality refinement mechanism. In particular, a remaining-distance factor and a safety-distance constraint were embedded into the cost function to improve the global path generation, aiming to enhance safety and reduce the number of turning points. Secondly, to address the issue of dynamic obstacle avoidance, the conventional DWA was improved by redesigning the velocity evaluation function. A global guidance term was introduced, along with an adaptive weight adjustment strategy, to better balance local dynamic avoidance with global path optimality. Finally, a series of simulation-based comparative experiments and real-world validation trials were conducted. Experimental results demonstrated that the proposed method exhibited superior performance in terms of both effectiveness, path smoothness and safety, offering a promising and robust solution for mobile robot navigation in the constrained and dynamic environments typical of intelligent plant factories.