Cooperative and Adaptive Mechatronic Systems: Identification, Control, and Optimization

该研究计划主要研究动态机电系统的分布式参数辨识、控制和优化，使用新的协作和自适应方法来处理一类任务，包括(I)通过车载传感器网络在线辨识机电系统的参数，例如自动汽车或移动机器人平台；(Ii)机电多智能体(车辆/机器人/工作单元)系统(MASS)的协调和控制；(Iii)极值搜索。对于这样的任务，该计划将开发一套广泛适用的(A)有效的分布式高级算法，(B)传感器和执行器网络体系结构，以及(C)用于在各种机电一体化质量上运行算法的低级实用控制器。通过在汽车、多车辆、可重构机器人和传感器网络测试平台上进行建设性的数学分析和实时测试和验证，所有结果都将在保证稳定性、性能和健壮性的情况下产生。 在传感器/估计器网络上使用开发的在线协作识别算法，与使用单个传感器/估计器单元相比，在不确定的机电系统的高级控制中，例如具有变化的道路和车辆条件的汽车系统，以及具有未知信号传播系数的环境和生物医学目标跟踪系统，将提高估计的可靠性和准确性。 所开发的协作和自适应MAS控制方案可用于(自动)汽车排的协作驾驶和紧急机动；使用无人驾驶飞行器(UAV)、无人地面飞行器(UGV)、自主水下航行器(AUV)、航天器或机器人团队的协作监视；移动无线传感器网络的协调；分离空间探索单位(如分立望远镜、天线和卫星)的编队捕获和调节。 极值搜索研究将集中于提高性能和可靠性；基于(电磁、声学、辐射、热)信号强度的目标跟踪，用于环境和生物医学监测；以及通过对汽车、机器人和工业自动化应用的测量数据进行成本最小化搜索来优化估计器/控制器参数。 该计划的成果将有益于加拿大和全球的几个战略应用领域，包括高性能和可靠的信息/通信技术(ICT)传感器系统设计，机电网络(特别是汽车、工业自动化和航空航天)网络的最佳协调，基于位置的联网信息系统，智能联网监测，人类干预和气候变化对生态系统的影响的联网检测，以及空气和水的废物和污染监测。

The proposed research program focuses on distributed parameter identification, control and optimization of dynamic mechatronic systems, using novel cooperative and adaptive approaches for a class of tasks, including (i) on-line parameter identification of mechatronic systems, e.g. autonomous cars or mobile robotic platforms, via onboard sensor networks; (ii) coordination and control of mechatronic multi-agent (vehicle/robot/work unit) systems (MASs); (iii) extremum seeking. For such tasks, the program will develop widely applicable sets of (a) effective distributed high level algorithms, (b) sensor and actuator network architectures, and (c) low level practical controllers for running the algorithms on various mechatronic MASs. All outcomes will be produced with guaranteed stability, performance, and robustness, via constructive mathematical analysis and real-time testing and verification on automotive, multi-vehicle, reconfigurable robot, and sensor network test platforms. Use of the developed on-line cooperative identification algorithms over a sensor/estimator network will enhance estimation reliability and accuracy compared to the use of a single sensor/estimator unit, in advanced control of uncertain mechatronic systems, such as automotive systems with varying road and vehicle conditions, and environmental and biomedical target tracking systems with unknown signal propagation coefficients. The developed cooperative and adaptive MAS control schemes have use in cooperative driving and emergency maneuvering of (autonomous) car platoons; cooperative surveillance using teams of unmanned aerial vehicles (UAVs), unmanned ground vehicles (UGVs), autonomous underwater vehicles (AUVs), spacecraft or robots; coordination of mobile wireless sensor networks; formation acquisition and regulation of fractionated space exploration units, e.g. fractionated telescopes, antennas, and satellites. The extremum seeking studies will focus on enhancing performance and reliability; (electromagnetic, acoustic, radiation, heat) signal intensity based target tracking by sensory MASs for environmental and biomedical monitoring; and estimator/controller parameter optimization via cost minimization search over measured data for automotive, robotic, and industrial automation applications. The program outcomes will be beneficial for several strategic Canadian and global application fields, including high-performance and reliable information/communication technology (ICT) sensor system design, optimal coordination of mechatronic (particularly automotive, industrial automation, and aerospace) networks, location-based networked information systems, intelligent networked surveillance, networked detection of human intervention and climate change effects to the ecosystem, and waste and contamination monitoring of air and water.