A multi-layer information model and adaptive algorithms for monitoring hybrid energy supply in IoT-based telecommunication networks
Abstract
Most deployed energy-monitoring systems only measure electrical parameters and transmit them to a remote server, whereas equipment condition assessment and automatic optimisation remain undeveloped. This paper proposes the architecture and the mathematical model of a hybrid monitoring platform uniting IoT sensing, edge computing, cloud analytics and digital-twin technology in one six-layer framework. A multi-stage edge-side conditioning scheme is developed, the digital-twin state residual is introduced as a diagnostic feature for early fault detection, and an integral Energy Efficiency Index with a carbon-footprint model is formulated so that technical and environmental criteria are evaluated in one loop. The testbed on a 0.4 kV three-phase feeder and the 180-day dataset of 17 280 records collected on it are described. The deep-learning forecasting layer built on this architecture is treated in a companion study.