基于改进模糊聚类并行蜻蜓算法的无人水面艇航迹规划  

Unmanned Surface Vessel Path Planning Based on Improved Fuzzy Clustering Parallel Dragonfly Algorithm

作　　者：门雅范 杨宗仁[1,2] 陈嫄玲 MEN Ya-fan;YANG Zong-ren;CHEN Yan-ling(Department of Industry and Information Technology of He'nan Province,He,nan Zhengzhou 450008,China;He'nan Information Engineering School,He'nan Zhengzhou 450008,China;Zhengzhou University of Light Industry,He*nan Zhengzhou 450008,China)

机构地区：[1]河南省工业和信息化厅,河南郑州450008 [2]河南信息工程学校,河南郑州450008 [3]郑州轻工业大学,河南郑州450008

出　　处：《机械设计与制造》2023年第1期101-106,共6页Machinery Design & Manufacture

基　　金：河南省教育科学规划基金(2019-JKJXA-142)。

摘　　要：对无人水面艇(Unmanned Surface Vessel,USV)航迹规划问题进行研究,提出了一种基于改进模糊聚类并行蜻蜓算法的无人水面艇航迹规划方案。首先,构建基于航距、能耗、航迹弯折度、航行安全度评价指标和网格坐标转换的USV航迹规划模型,在确保航行安全、降低问题求解维度的同时,最大限度地提高航迹平滑度和降低航行距离与能耗。其次,设计改进模糊聚类并行蜻蜓算法(Improved Fuzzy Clustering Parallel Dragonfly Algorithm,IFDDA),利用改进的并行模糊聚类算法,自适应确定蜻蜓静态邻域规模,以提高种群样本多样性;引入非线性进化因子,动态调整算法权重因子,以平衡算法局部搜索和全局搜索。最后,采用IFDDA算法对USV航迹规划模型进行求解,经典测试函数和具体航迹规划实例仿真结果表明,IFDDA具有良好的复杂高维函数优化性能,相比于其它航迹规划算法,基于IFDDA得到航程缩短了约(7.3~55.2)%、能耗降低了约(17.1~56.5)%。The path planning of unmanned surface vessel(USV)is studied,and an unmanned surface vessel path planning scheme based on improved fuzzy clustering parallel dragonfly algorithm(IFDDA)is proposed.First,a USV path planning model based on range,path curvature,energy consumption,path bending,navigation safety and grid coordinate transformation is constructed to ensure navigation safety while reducing problem solving dimension,minimizing navigation distance and navigation energy consumption.Secondly,an improved fuzzy clustering parallel dragonfly algorithm(IFDDA)is designed.The improved parallel fuzzy clustering algorithm is used to adaptively determine the size of the dragonfly static field to increase the diversity of population samples.The nonlinear evolution factor is introduced to dynamically adjust the algorithm weight factor,which helps to balance the local search and global search of the algorithm.Finally,the IFDDA algorithm is used to solve the USV path planning model.The simulation results of classic test functions and specific path planning examples show that IFDDA has good optimization performance of complex high-dimensional functions.Compared with other track planning algorithms,the range is shortened by(7.3~55.2)%and the energy consumption is reduced by(17.1~56.5)%.

关 键 词：无人水面艇 航迹规划 蜻蜓算法 模糊聚类 航程 能耗 

分 类 号：TH16[机械工程—机械制造及自动化] TP249[自动化与计算机技术—检测技术与自动化装置]