Supporting Temporal Coordination in Wirelessly Networked Control Systems: Fundamental Theory, Algorithms, Architecture, and Experimentation

点击翻译按钮获取中文摘要

Intellectual Merit: When networked computers interact with the physical world, the problem of consistent time-stamping of sensor measurements, and temporal coordination of actuators, emerges as a core technical problem for high performance. In control systems, state-estimators need accurately time-stamped information, and distributed control actions must be coordinated temporally. It is also important for sensor networks, e.g., localization based on time of flight, and communication networks, where slotted synchronization and packet scheduling lead to greater throughputs. However, there are fundamental limitations to clock synchronization. This project investigates what is fundamentally feasible or infeasible in clock synchronization. It also investigates algorithms that scale well with system size.Broader Impact: Control systems are widespread in industries, including process control and manufacturing, and platforms such as cars and aircraft. We are on the cusp of a third generation of networked control systems that combines all the technological advances made in the forty years since the advent of digital control -- powerful computational hardware, complex software, and networked systems. It will lead to large scale control systems that are easily deployable and reconfigurable, and expected to be of great importance for future large scale projects such as smart transportation systems or energy grids, as well as for greater efficiencies in critical industries, such as the automobile industry, where cars already feature about 50 embedded computers. This project will train graduate students, host hundreds of students from local high schools at the annual Engineering Open House, and disseminate results in the open literature.