Event-triggered Feedback in Control, Estimation, and Optimization

Networked control systems often send information across the communication network in a periodic manner. The selected period, however, must assure adequate system performance over a wide range of operating conditions and this conservative choice may result in significant over-provisioning of the communication network. This observation has motivated the use of sporadic transmission across the network’s feedback channels. Event-triggering represents one way of generating such sporadic transmissions. In event-triggered feedback, a sensor transmits when some internal measure of the novelty in the sensor information exceeds a specified threshold. In particular, this means that when the gap between the current and the more recently transmitted sensor measurements exceeds a state-dependent threshold, then the information is transmitted across the channel. The state-dependent thresholds are chosen in a way that preserves commonly used stability concepts such as input-to-state stability or L2 stability. This approach for threshold selection therefore provides a systematic way of triggering transmissions that provides some guarantees on overall control system performance. While early work in event-triggering focused on control applications, this technique can also be used in distributed estimation and distributed optimization. This chapter reviews recent progress in the use of state-dependent event-triggering in embedded control, networked control systems, distributed estimation, and distributed optimization.