基于改进捆绑拍卖多机器人任务分配研究  

Task allocation of searching in multi-robot system using improved bundled auction

作　　者：姜来浩 戴学丰[1] 蔡标[1] 陈泽涛[1] 

机构地区：[1]齐齐哈尔大学计算机控制与工程学院,黑龙江齐齐哈尔161006

出　　处：《齐齐哈尔大学学报（自然科学版）》2014年第6期5-9,共5页Journal of Qiqihar University(Natural Science Edition)

基　　金：黑龙江省自然科学基金(F201331)

摘　　要：为了解决多机器人在搜索过程中多任务分配和多机器人利用率问题,提出了一种带有即时拍卖的K-means聚类捆绑式拍卖算法。首先通过K-means聚类算法解决多机器人系统中的多任务捆绑问题,再运用捆绑式拍卖机制把聚类分配给相应的机器人。考虑各聚类内的任务完成情况和各机器人利用情况,相应机器人判断是否产生即时拍卖。仿真证明,该方法在多机器人搜索过程中节省资源且各机器人利用均衡。In order to solve the problem of multi-robot in the search process of multi task allocation and multi-robot utilization, this paper presents a K-means clustering with instant auction algorithm based bundle auction.Firstly by using K-means clustering algorithm to solve the problem of multi tasks in multi robot binding ,then using bundled auction mechanism allocates the clusters to the corresponding robots. According to the completion of the tasks in each cluster ,the corresponding robot judge whether to make instant auction. Simulation results proves that this method of the multi-robot in the search process save resource and the robots use equilibrium.

关 键 词：搜索 任务分配 K-MEANS聚类 捆绑式拍卖 闲置消耗 即时拍卖 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]