| As computer-controlled systems become more and more ubiquitous in today's world, the physical challenges that must be overcome to ensure their reliable and efficient operation become extremely important. In general, controllers are designed assuming perfect information is available at all times and actuators can be updated continuously. When resources such as actuator power or energy consumption are not factors, this is not a problem because the controller can take samples and update the control signals fast enough so that the system behaves as close to ideally as possible. But now as we move steadfast into the age where we want everything to be smaller and more portable, physical constraints such as battery life become major limiting factors to what we can achieve. Furthermore, when considering wireless networks of cyber-physical systems, the coupled physical constraints become an even larger challenge, making it unrealistic to continue blindly using controllers that assume ideal scenarios.;This naturally gives rise to the study of robustness of controllers. In other words, given a system and an ideal controller, how fast must the controller sample the state and update the actuator signals such that the system behaves in the intended way? Rather than answering the above questions directly, we are interested in finding control strategies that account for these uncertainties at the design stage. While it is certainly easier to design controllers for systems that have perfect information at all times, it is just not practical in the real world.;In this dissertation we explore various existing and new methods of endowing cyber-physical systems with sufficient autonomy to allow them to determine when and what kind of information is needed to perform a given task with a desired quality of service. The core of this dissertation is rooted at ideas developed from event- and self-triggered control strategies which we apply to a variety of different goals. We also provide a novel framework for a new method we call the team-triggered control strategy. This strategy combines the strengths of event- and self-triggered control and shows a lot of promise in efficiently controlling networks of cyber-physical systems. |
