Application of Factorial Hidden Markov Model (FHMM) to Non-intrusive Load Monitoring



This project apply the factorial hidden markov model (FHMM) to intrusively disaggregate the loads. Each individual load is modeled by the hidden markov mode with Gaussian obervations, and the aggregated loads are characterized by a FHMM, then the FHMM can predict the states of different types of individual loads in an unknown household, such as fridges, lights, microwaves.