Viterbi algorithm with sparse transitions (VAST) for nonintrusive load monitoring

Implementation of smart grid provides an opportunity for concurrent implementation of nonintrusive appliance load monitoring (NIALM), which disaggregates the total household electricity data into data on individual appliances. This paper introduces a new disaggregation algorithm for NIALM based on a modified Viterbi algorithm. This modification takes advantage of the sparsity of transitions between appliances' states to decompose the main algorithm, thus making the algorithm complexity linearly proportional to the number of appliances. By consideration of a series of data and integrating a priori information, such as the frequency of use and time on/time off statistics, the algorithm dramatically improves NIALM accuracy as compared to the accuracy of established NIALM algorithms.