基于双层路径优化算法的多机器人最优编队方法  

Multi-robot optimal formation method based on two-layer path optimization algorithm

作　　者：周佳加[1] 张强[1] 王宏健[1] 张洪泉[1] 王莹莹 ZHOU Jiajia;ZHANG Qiang;WANG Hongjian;ZHANG Hongquan;WANG Yingying(College of Automation,Harbin Engineering University,Harbin 150001,China)

机构地区：[1]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001

出　　处：《传感器与微系统》2021年第9期65-67,70,共4页Transducer and Microsystem Technologies

基　　金：国家自然科学基金资助项目(51609048,51909044,51809060);黑龙江省自然科学基金资助项目(QC2017051,E2017014);黑龙江省科技补贴项目(GX17A014);传感技术联合国家重点实验室开放项目(SKT1905)。

摘　　要：为解决能耗和时间约束下的多机器人编队集结点优化问题,提出一种将粒子群优化(PSO)与连续Hopfield神经网络(CHNN)结合的双层路径优化算法。针对时间约束,设计了一种PSO算法求解最优编队集合点;针对总能耗约束,利用Dubins路径将路径规划转换成指派问题,并基于CHNN搜索最优分配路径。三个同构欠驱动机器人的数值仿真结果表明:所提算法提高了搜索能力,且能够实现最优编队集结。To solve the problem of multi-robot formation assembly point optimization under energy consumption and time constraints,a two-layer path optimization algorithm is proposed,which combines particle swarm optimization(PSO)and continuous Hopfield neural network(CHNN).For the time constraint,a PSO algorithm is designed to solve the optimal formation set point.For the total energy consumption constraint,the Dubins path is used to transform the path planning into an assignment problem,and the optimal distribution path is searched based on CHNN.The numerical simulation results of three isomorphic underactuated robots show that the algorithm improves the search ability and can achieve the optimal formation aggregation.

关 键 词：欠驱动机器人 编队 粒子群优化算法 连续Hopfield神经网络(CHNN) 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]