Adaptive Motion Coordination and Control of Robotic and Vehicular Networks

由于单一机器人解决方案的不足，当今大多数工业自动化，室内（清洁，操作，厨房等）服务和监控的机器人应用涉及多个机器人相互作用。为了安全有效地运行这些应用程序，需要机器人之间通过通信和传感网络的分布式协调和控制进行可靠的合作。智能乘客与任务车辆的合作也是如此，以提高交通和任务性能。为了以比现有解决方案更好的性能更可靠地满足这一需求，本研究计划将研究新的运动协调和控制系统设计，包括(1)通过车载传感器网络的智能车辆和自动化/服务机器人的协作和自适应参数识别和控制系统；(2)分布式协作最优运动规划器；(3)分布式自适应运动协调和编队控制系统，用于(a)协同目标定位和跟踪，(b)协同地形测绘和导航，(c)优化机器人服务覆盖，(4)基于自适应和学习的运动规划，路径跟踪和机动控制系统，用于通过智能车辆网络进行协同驾驶。该设计将具有基于学习的动态协同运动规划、网络协调、在线参数和状态估计以及自适应低级运动/机动控制的交互层。项目成果将培育新的系统识别和自适应控制方法，作为现有系统的有利替代方案；它们将在机器人运动规划、车辆运动控制和多智能体系统协调之间形成桥梁，为复杂的运动协调和控制问题提供集成的高性能解决方案。此外，这些成果将开启合作目标寻找和监视的新时代。在该计划中开发的系统和算法将在制造、汽车、监控和应急响应等高要求领域，以及加拿大和全球利益的几个战略领域，包括协作（半）自动驾驶和驾驶员辅助系统、制造协调、自动化、自动驾驶和自动驾驶系统等领域，推动加拿大工业和政府技术组织在能力和应用方面的先进水平。服务机器人，生态系统干预和气候变化影响的网络检测，空气和水的废物和污染监测，网络监测。他们将推动这些行业和组织的发展，使下一代自动驾驶和用户/驾驶辅助产品能够由他们开发。通过该计划培训的HQP将成为这些部门工作的理想人选，带来他们通过该计划获得的技能、经验和技术背景。

Most of the robotic applications of today for industrial automation, indoor (cleaning, manipulation, kitchen, etc) services, and surveillance involve multiple robots interacting with each other, due to insufficiency of single robot solutions. For safe and efficient run of these applications, reliable cooperation among the robots is needed, via distributed coordination and control over a communication and sensing network. Similar situation holds for cooperation of intelligent passenger and mission vehicles for enhancing traffic and mission performance. To address this need with better performance more reliably than the existing solutions, this research program will investigate new motion coordination and control system designs, including (1) cooperative and adaptive parameter identification and control systems of intelligent vehicles and automation/service robots, via onboard sensor networks, (2) distributed cooperative optimal motion planners, (3) distributed adaptive motion coordination and formation control systems for (a) cooperative target localization and tracking, (b) cooperative terrain mapping and navigation, (c) optimal robotic service coverage, (4) adaptive and learning based motion planning, path tracking, and maneuver control systems for collaborative driving via intelligent vehicular networks. The designs will have interacting layers for learning based dynamic cooperative motion planning, network coordination, on-line parameter and state estimation, and adaptive low-level motion/maneuver control. The program outcomes will foster new system identification and adaptive control methods as advantageous alternatives to the existing ones; they will form bridges among robotic motion planning, vehicle motion control and multi-agent system coordination, leading to integrated high-performance solutions for complex motion coordination and control problems. Further, these outcomes will initiate new eras in cooperative target seeking and surveillance. The systems and algorithms developed in the program will advance the state-of-art in capabilities and applications of a wide range of Canadian industries and governmental technology organizations, within the highly demanding sectors of manufacturing, automotive, surveillance, and emergency response, along several strategic areas of Canadian and global interest, including collaborative (semi-)autonomous driving and driver assistance systems, coordination of manufacturing, automation, and service robots, networked detection of ecosystem interventions and climate change effects, waste and contamination monitoring of air and water, networked surveillance. They will  advance such industries and organizations, enabling the next generation of autonomous and user/driver assistance products to be developed by them. The HQP trained through the program will be ideal candidates to work in these sectors, bringing in the skill-sets, experience, technical background they gain through the program.