Event-triggered sliding mode control via digital redesign

Abstract

This paper presents the digital redesign of the event-triggered sliding mode control, which emulates the analog (control) signal for a sampled data system. The proposed controller consists of a sensor module and an actuator module, communicating with each other intermittently on the occurrence of events at the sensor. Both of these modules comprise a signal generator that mimics the plant model exactly in the absence of disturbance, thereby achieving analog performance. By using this proposed methodology, we achieve two major outcomes over the existing approaches. The first one is that the triggering mechanism achieves a larger inter-event time compared to that of zero-order hold-based design, leading to a sparser sampling sequence. As a result, the communication burden in the feedback loop is reduced significantly due to the reduction in the transmission cost. The second outcome is that the triggering sequence admits a uniform lower bound globally, i.e., Zeno behavior is avoided globally. This guarantees that the trajectories of the closed-loop system starting from anywhere can be bounded within any arbitrary bound by using the proposed event-triggered controller. Finally, we demonstrate our design methodology with the help of a numerical example.