Integration of cooperative mobile robots and networked sensor fusion in Safety/Security-Critical tasks

美国几位著名的研究人员为美国国防部（DoD）的多学科大学研究计划（MURI）提出了“传感器网络结构工程的独立融合”研究。美国陆军研究办公室（ARO）最近资助了这个项目。美国的提议已经确定不列颠哥伦比亚省大学的Clarence de Silva博士为计划中研究的唯一加拿大合作者。成功的MURI申请为我们提供了一个独特的机会，在这个研究项目中进行合作，这是一个重大的共同利益。美国研究人员将主要关注非结构化城市环境中的动态传感器网络和信息融合，以实现安全/安保关键操作。考虑到这些动态网络信息采集和融合技术，拟议的SRO项目将集中研究开发多机器人合作技术，以在非结构化城市环境中执行对人类安全/保障至关重要的任务。具体来说，SRO项目将侧重于：基于行为的机器人自确定性合作学习;使用遗传算法（GA）优化机器人结构和动作;将GA与强化学习（RL）集成以生成最佳合作策略;使用人工免疫系统（AIS）实现系统鲁棒性和自组织;智能控制的软计算和基于知识的决策;并整合美国首创的动态传感器网络与融合研究。这些活动将涉及分析研究;计算机模拟;在多个移动的机器人的原型实验室系统中进行技术开发和实施;将物理系统与用于通信和信息融合的动态网络拓扑结构（由美国研究人员开发）集成在一起;实际安全/安全关键活动的实际演示;以及通过严格的物理测试进行系统评估。

Several prominent US researchers have proposed research on "Engineering of Sensor Network Structure for Dependable Fusion" for the Multidisciplinary University Research Initiative (MURI) of the US Department of Defense (DoD). The US Army Research Office (ARO) has recently funded this project. The US proposal has identified Dr. Clarence de Silva of the University of British Columbia as the only Canadian collaborator of the planned research. The successful MURI application provides a unique window of opportunity for us to collaborate in this research project which is of significant mutual interest. The US researchers will primarily focus on dynamic sensor networking and information fusion in an unstructured urban environment for safety/security critical operations. Given these technologies of dynamically networked information acquisition and fusion, the proposed SRO project will concentrate on research to develop technologies for multi-robot cooperation in carrying out tasks critical for human safety/security in an unstructured urban environment. Specifically the SRO project will focus on: Behaviour-based Learning for Self-deterministic Cooperation of Robots; Use of Genetic Algorithms (GA) for Optimization of the Robotic Structure and Actions; Integration of GA with Reinforcement Learning (RL) for Generating Optimal Cooperation Policies; Use of Artificial Immune Systems (AIS) for System Robustness and Self-organization; Soft Computing and Knowledge Based Decision Making for Intelligent Control; and Integration of US-originated Research on Dynamic Sensor Networking and Fusion. The activities will involve analytical research; computer simulation; technology development and implementation in a prototype laboratory system of multiple mobile robots; integration of the physical system together with a dynamic networking topology for communication and information fusion (developed by US researchers); practical demonstration of a realistic safety/security critical activity; and system evaluation through rigorous physical testing.