基于模糊动态窗口的无人艇多动态障碍物路径规划  

Multi-dynamic obstacle path planning of unmanned boat based on fuzzy dynamic window

作　　者：林泽琼 刘海生 LIN Zeqiong;LIU Haisheng(Naval Architecture and Ocean Engineering College,Dalian Maritime University,Dalian 116026,China)

机构地区：[1]大连海事大学船舶与海洋工程学院,辽宁大连116026

出　　处：《舰船科学技术》2025年第6期62-68,共7页Ship Science and Technology

基　　金：高性能船舶技术教育部重点实验室开放基金课题资助项目(gxnc23052803)。

摘　　要：针对无人艇在多动态障碍物下避障、快速规划路径进行导航决策的要求,传统的动态窗口算法(DWA)存在运行效率低、容易陷入局部优化、复杂环境下搜索能力差等问题。在具有多个动态障碍物的复杂环境中,引入一种模糊逻辑算法对DWA算法进行改进,提高了路径评估函数的灵活性。模糊控制系统通过输入无人艇、终点和周围障碍物的信息,输出评价函数的相应系数权值。在无人艇搜索路径过程中,实时调整轨迹评价函数的权重,以达到优化路径的目的。仿真实验表明,模糊改进的动态窗口方法可以使规划的路径更平滑,获得更短的路径,同时有效地减少了无人艇的路径搜索时间。在具有多个动态障碍物的环境中,可以有效地实现动态避障。For the requirements of unmanned boats to avoid obstacles under multi-dynamic obstacles and quickly plan paths for navigational decision-making,traditional dynamic window algorithms suffer from the problems of low operating efficiency,easy fall into local optimization,and poor search ability in complex environments.In this paper,a fuzzy logic algorithm is introduced to improve the DWA algorithm in a complex environment with multiple dynamic obstacles,which improves the flexibility of the path evaluation function.The fuzzy control system outputs the corresponding coefficient weights of the evaluation function by inputting information about the unmanned boat,the endpoint,and surrounding obstacles.The weights of the trajectory evaluation function are adjusted in real time during the search path of the unmanned boat to achieve the purpose of optimizing the path.The simulation experiments show that the fuzzy improved dynamic window method can make the planned path smoother and obtain shorter paths while effectively reducing the time of searching paths for unmanned craft.Dynamic obstacle avoidance can be effectively realized in an environment with multiple dynamic obstacles while searching for the globally optimal path.

关 键 词：无人艇 动态窗口法 模糊算法 多动态障碍物 路径规划 

分 类 号：U666.11[交通运输工程—船舶及航道工程]