基于协同多目标算法的多机器人路径规划

多机器人路径规划是群体机器人协同工作的前提,其特点是在防碰撞与避障的前提下追求多方面资源的最小消耗.针对这一特点,提出协同非支配排序遗传算法,解决具有多个优化目标的多机器人路径规划问题;运用改进的多目标优化算法,克服多目标优化取权值的不足,同时考虑机器人能源与时间两大资源,以多机器人的路径总长度、总平滑度、总耗时为规划目标.同时引入合作型协同算法框架,将难以求解的多变量问题分组求解.每个机器人的路径视为子种群,子种群通过带精英策略的非支配排序遗传算法,进化并筛选出子种群的部分进入协同进化,每次迭代更新外部的精英解集,最终生成一组非支配路径解.仿真结果表明,在栅格地图环境下,本文算法可有效实现多移动机器人的多优化目标路径规划.

Multi-robot Path Planning Based on Cooperative Multi-objective Algorithm

Multi-robot path planning is the premise of the cooperative work of group robots,and it is characterized by the pursuit of multi-resource minimum consumption under the premise of collision prevention and obstacle avoidance.Aiming at this characteristic,a collaborative non-dominant sequencing genetic algorithm is proposed to solve the multi-robot path-planning problem with multiple optimization objectives.An improved multi-objective optimization algorithm is applied to overcome the deficiency of multi-objective optimization in weight selection,and the total path length,smoothness,and time of the multi-robot are considered.At the same time,a collaborative algorithm framework is introduced to solve the problem of multivariate groups;here,each robot path is regarded as a sub-population.Parts of the sub-populations are selected by the nondominated sorting genetic algorithm(NSGA-Ⅱ)with elitist strategy,and they evolve into cooperative coevolution.Each iteration update outside the elite solution set eventually forms a set of control path.Simulation results show that the proposed algorithm can effectively realize multi-objective path planning for multi-mobile robots in a known raster map environment.