Non-intrusive identification of harmonic polluting loads in a smart residential system

Abstract

Smart meter technology has been developed rapidly in modern industrial environment in the context of smart grid connected residential load system. For the smart meter’s application, knowledge about instantaneous load pattern is crucial. Non-intrusive load monitoring (NILM) is a well-known method to assess the power consumption of individual load as well as its operating behavior.​ Since the modern household appliances can inject unwanted harmonics to the system, identification of harmonic polluting loads has also become an issue for such load monitoring schemes. In this article, an improved technique to identify the harmonic polluting loads has been presented using only input aggregated voltage–current data of a residential system. A search-based optimization, i.e., Artificial Bee Colony (ABC) algorithm is used in the proposed load monitoring technique. Unlike the existing methods, this technique does not require any heavy training mechanism for the system. The algorithm is verified using the PLAID datasets and results are compared to some state-of-the-art techniques. Suitable simulations and experimentally created dataset analysis have been carried out on a residential system to demonstrate the ability of the proposed methodology.