Collective robotics

越来越多的人正在寻求多机器人系统（MRS）来解决一些偏远（月球或火星）和危险环境（水下或灾区）的挑战，这些挑战不能由单机器人系统充分解决。由于并行执行的效率，冗余的鲁棒性，以及空间分布和异构性的多功能性，与单机器人相比，MRS具有巨大的优势。本提案所描述的研究将著重于开发多机器人系统（MRS）的分散控制算法，通过检查两个重要任务：集体决策和集体构建。由于MRS的集中控制算法在数学上是难以解决的，目前的大多数研究都是针对分散的解决方案。幸运的是，大自然为我们提供了存在性证明和特定的算法，可以根据各种感兴趣的问题构建去中心化的MRS算法。本文概述了两种MRS算法的发展步骤：集体决策算法（CDMA）和信息素辅助集体构建算法（PACC），前者的灵感来自蚂蚁的巢迁移算法，后者的灵感来自群居昆虫筑巢的两种基本机制：耻感和信息素。研究活动将涉及一些研究生。

Increasingly, people are looking to multi-robot systems (MRS) for solutions to some of the challenges in remote (Moon or Mars) and dangerous environments (underwater or disaster zones) that can not be adequately addressed by single-robot systems. A MRS enjoys tremendous advantages over its single-robot counterpart due to its efficiency through parallel execution, its robustness through redundancy, and its versatility through spatial distribution and heterogeneity. The research described in this proposal will focus on developing decentralized control algorithms for a multi-robot system (MRS) through the examination of two important tasks: collective decision making and collective construction. Because centralized control algorithms for a MRS are known to be mathematically intractable, most of the current research is directed toward decentralized solutions. Fortunately, nature has provided us with both existence proof and specific algorithms to follow for various problems of interest on which to build decentralized MRS algorithms.  This proposal outlines the steps toward the development of two MRS algorithms: Collective Decision Making Algorithm (CDMA) - motivated by the nest relocation algorithm of an ant species - and Pheromone Assisted Collective Construction (PACC) inspired by the two basic mechanisms used by social insects in building their nests: stigmergy and pheromone.  The research activities will involve a number of graduate students.