An Adaptive-Neuro Fuzzy Inference System for Load Disaggregation in Residential Households

Abstract

Optimization of energy cost and consumption is a vital topic in today’s world. Normally, smart meters are used to record the total energy consumption at the customers’ end across the entire building and the users only receive aggregate electricity bills at the end of each month providing the information of their energy consumptions. Optimization of energy cost can be done using a feedback system, which is realizable through Non-Intrusive Load Monitoring (NILM). NILM is a process of identifying household appliances by disaggregating the mains power measurement into each appliance individually. Due to diversity of appliances, NILM is a complex classification problem. In this research, the NILM problem to identify the activation of household appliances has been solved using a hybrid technique termed as Adaptive-Neuro Fuzzy Inference System (ANFIS). This article aims to identify regular household appliances from the smart meter measurement. A publicly available UK Domestic Appliance-Level Electricity (UK-DALE) dataset has been employed to test and verify the effectiveness of the proposed method using different performance evaluation parameters. The results are compared with the existing literature to demonstrate the effectiveness of the proposed technique for the NILM problem.