Web-Based Residence Energy Consumption Monitoring and Seasonal Variation Analysis of Domestic Appliances

Abstract

The consumption of energy in residential areas has been unpredictable due to recent weather changes, changing consumption habits, and the increasing power of electrical appliances. The culminating reading from their electric meter and their invoices (for the previous month's activities) are currently the only resources available to users to monitor their total consumption of energy, and there is no available historical pattern data. The consumers may not fully understand how they consume energy or have sufficient knowledge to make informed decisions while trying to minimize their consumption due to lack of knowledge regarding these fluctuations. Considering seasonality in electrical appliance consumption, this study proposes the development of a web-based system for monitoring and forecasting electrical consumption in households through a set of structured processes that includes data collection, data preprocessing, forecasting, and visualization. For forecasting power consumption, the previous electric energy data will be cleaned and arranged in a time series fashion. The Seasonal Autoregressive Integrated Moving Average (SARIMA) model, which can describe both the trend and seasonal trends of home electric consumption data, will be used. The ability of the system to forecast home electric consumption correctly (with an average forecasting error of $8-10 \%)$ and to capture both the forecasted seasonality of demand and total use versus actual usage was tested using actual residential electrical use data. The output is presented in an easy-to-use web portal that allows users to understand how much energy has been used and receive suggestions to reduce unnecessary consumption, enabling better awareness and more effective energy management.