Energy Disaggregation Using Multi-Objective Genetic Algorithm Designed Neural Networks

Energy-saving schemes are nowadays a major worldwide concern. As the building sector is energy consumer, and hence greenhouse gas emitter, research in home management systems (HEMS) has increased substantially during last years. One of primary purposes HEMS monitoring electric consumption disaggregating this across different appliances. Non-intrusive load (NILM) enables disaggregation without having to resort profusion specific meters associated with each device. This paper proposes low-complexity low-cost NILM framework based on radial basis function neural networks designed by multi-objective genetic algorithm (MOGA), design data selected an approximate convex hull algorithm. Results proposed residential house demonstrate models’ ability disaggregate devices excellent performance, which was consistently better than using other machine learning algorithms, obtaining F1 values between 68% 100% estimation accuracy ranging from 75% 99%. The approach enabled us identify operation appliances accounting for 66% total recognize that 60% could be schedulable, allowing additional flexibility operation. Despite reducing sampling one second minute, allow employment low complexity models enable its real-time implementation hardware, technique presented usage devices.