Forecasting Residential Electricity Load Demand using Machine Learning

Since the emergence of different forms of sophisticated home appliances and smart home devices globally, the demand for residential energy is rapidly growing. This growing demand has led to create energy sustainability issues, which have been identified as one of the major concerns in the recent times as more consumers need more energy. Therefore, the forecasting of demand for electricity plays a crucial role in maintaining an equilibrium between the consumers demand and energy generated by the energy producing companies. In recent years, numerous machine learning algorithms have been employed to forecast the consumers electricity load demand. This study has been carried out as a comparative analysis of four different machine learning algorithms, Autoregressive Integrated Moving Average with Explanatory Variable (ARIMAX), Seasonal Autoregressive Integrated Moving Average with Explanatory Variable (SARIMAX), PROPHET and Long-Short Term Memory (LSTM) for a group of residential customer present at London. It is observed that the efficiency of the LSTM model is considerably greater than all the other implemented models. Keywords - Residential Energy, Forecasting, Electricity, Energy Demand, ARIMAX, SARIMAX, PROPHET, LST