Decentralized Event-Triggered Control Over Wireless Sensor Actuator Network

Design for Energy Conservation & Implementation

More Info
expand_more

Abstract

Over the last decade there has been a shift in focus from traditional sampled-data control to event-triggered control (ETC) as it promises to be better suited for systems with shared communication media, especially wireless systems. The mote lifetime in case of wireless sensor actuator networks (WSAN) is crucial and can be improved by utilizing efficient control algorithms. Unlike sampled-data control, event-triggered control (ETC) algorithms require updates from sensors only when it is essential to maintain the stability of the system and ensure satisfactory control performance. Hence, ETC approach to control over WSAN systems can provide significant energy and transmission savings. Moreover, wireless motes spend considerable energy just by idle listening as the communication radio is periodically switched on as per the currently used duty cycling technique. Hence, providing a passive wake-up radio extension to wireless motes can prevent this energy wastage by switching on the radio only when communication is required. This thesis presents event-triggered control as a viable option to implement control over wireless sensor actuator networks (WSAN). To achieve this, a Hardware-in-Loop system is developed, in which dynamics of a plant are simulated on a real-time hardware and controlled via a wireless sensor actuator network. The system also provides a LabVIEW user interface depicting relevant real-time information. A comparative experimental study is done between periodic sampling and event-triggered approach of controlling the system. It is observed that the event-triggered approach achieves similar control performance and provides sensor transmission savings of more than 40%. Further, to reduce the energy wasted by the sensors in idle listening mode passive wake-up radio extensions are added to the sensors. Energy savings of 38% are obtained by using ETC with passive wake-up radios.

Files

Msc-thesis-saurav.pdf
(pdf | 12.2 Mb)
Unknown license