Non-Intrusive Appliance Load Monitoring using Genetic Algorithms

Smart Meters provide detailed energy consumption data and rich contextual information which can be utilized to assist energy providers and consumers in understanding and managing energy use. Here, we present a novel approach using genetic algorithms to infer appliance level data from aggregate load curves without a-priori information. We introduce a theoretical framework to encode load data in a chromosomal representation, to reconstruct individual appliance loads and propose several fitness functions for the evaluation. Our results, using artificial and real world data, confirm the practical relevance and feasibility of our approach.