CAREER: Power Optimization in Autonomous Microsystems via Integrated Motion Control

本学院早期职业发展（Career）奖的研究目标是探索新的控制策略，以最大限度地减少自主微系统的功耗。微系统是亚毫米尺寸的工程设备；在这种规模下，许多在大型设备中被忽略的现象变得重要起来，包括电路中的许多能量损失。新的控制策略将通过协调运动控制与电力电子和传感器电路的主动调节来显着减少总能源使用。这些策略将基于切换控制理论，其中微系统的各个组成部分通过有限数量命令的适当定时进行调节。这些切换命令可以在小型化系统中特别有效地实现。开关控制理论将得到加强，以在潜在物理组件行为变化的情况下提供稳健的性能，并随着时间的推移适应不断变化的条件。交付成果将包括全面的控制器设计和优化程序、微尺度组件建模工具、开关控制理论的创新文档、微机器人试验台的实验验证、工程学生教育以及面向中学生的交互式在线学习工具。本研究的成功实现将增加对互联微尺度动态系统功耗与性能之间基本权衡的理解。主要的应用是控制类昆虫的陆地微型机器人，其小尺寸极大地限制了电力的可用性。最小功率控制和优化算法将是为应急、健康和维护工作者创造微型机器人工具的重要一步，并将广泛应用于其他微系统，如无人值守传感器节点和植入式医疗设备，其中资源或能源节约至关重要。在此过程中，微型机器人将作为开发互动式课堂课程和在线课程的主题，向中学生教授工作和能量等概念，并将交互式在线活动细化，以便在互联网上广泛传播。

The research objective of this Faculty Early Career Development (CAREER) award is to explore novel control strategies for minimizing power consumption of autonomous microsystems. Microsystems are engineered devices with sub-millimeter dimensions; at this scale, many phenomena become important that are ignored in larger devices, including many energy losses in circuitry. New control strategies will dramatically reduce total energy usage by coordinating motion control with active regulation of power electronics and sensor circuitry. These strategies will be based on switching control theory, where the various components of a microsystem are regulated through proper timing of a finite number of commands. These switching commands can be implemented especially efficiently in miniaturized systems. Switching control theory will be enhanced to provide robust performance in the presence of variation in the behavior of underlying physical components, and to adapt to changing conditions over time. Deliverables will include comprehensive controller design and optimization procedures, micro-scale component modeling tools, documentation of innovations in switching control theory, experimental verification on micro-robotic testbeds, engineering student education, and interactive on-line learning tools for secondary students. Successful realization of this research will increase understanding of fundamental trade-offs between power consumption and performance of interconnected micro-scale dynamic systems. The primary application is the control of insect-like terrestrial micro-robots, where small size dramatically limits power availability. Minimal-power control and optimization algorithms will be a major enabling step towards the creation of micro-robotic tools for emergency, health, and maintenance workers, and will have wider application to other microsystems such as unattended sensor nodes and implantable medical devices, where resource or energy conservation is critical. In the process, micro-robotics will be used as a theme for developing interactive in-class lessons and online programs teaching concepts such as work and energy to secondary students, with interactive online activities refined for broad dissemination over the Internet.