Event-triggered real-time metering in smart grids

This paper introduces an event-triggered approach to realtime metering that significantly reduces the amount of reported measurements. The proposed method reports measurements only when the load profile changes due to a triggering event, and ensures that the maximum difference between reported and true measurement values is always bounded in magnitude. Moreover, it employs a change detection based filter to resolve the estimation-tracking conflict associated with non-stationary signals containing abrupt changes. Simulation results show that the proposed reporting method provides a convenient tradeoff between average reporting error as well as the reporting frequency.