Event-predictive Control for Energy Saving of Wireless Networked Control System

This paper discusses on a new concept of Event-predictive control, which is an extension of the Event-triggered control. The aim of the research is to apply such asynchronous sampling control strategy to a wireless networked control system to save battery energy consumption in the wireless nodes. The key idea is to maximize the control interval, under the condition of appropriate control performance, to save communication energy cost for battery life of wireless network nodes incorporated in the control system. A heuristic new concept of "predictive event" is introduced which is pre-determined as a cross-point of future predictive response of the controlled system and the stable region trajectory named "admissible set trajectory" to ensure recoverability to stable region. Then the sleep and wake mode of the wireless node is controlled according to the predictive events. Some considerations on stability condition are discussed, which is extendable to a class of nonlinear plants under some assumptions. A numerical example is illustrated to show the effectiveness of the proposed method.