Preliminary research on predictive sparse reporting in low-voltage network PLC-based metering

Advanced Metering Infrastructures (AMI) play a significant role in remote energy meter reading. A typical Power Line Communication (PLC) metering scheme can suffer from data transfer bottlenecks. On the one hand, the Meter Data Management system (MDM) requires costly contracts with wireless providers to acquire data from Data Concentrator Units (DCU), on the other hand PLC can become unreliable as a means of communicating energy measurements between Smart Meters (SM), and DCUs. In this article, the authors propose and compare two possible approaches to developing a sparse reporting technique which could enable electrical energy distribution companies to sacrifice some measurement accuracy in order to significantly reduce data volumes that need to be transferred regularly over expensive or precarious networks. A number of forecasting models are described in detail and evaluated in simulation using nine months worth of energy data collected from a sample of five hundred smart meters.