A New Short Term Load Forecasting Approach for Future and Smart Grids

In recent years, the key challenge for domestic distribution network is the highly uncertain load profile and increased emphasis to use renewable power. To ensure appropriate operation and planning of the power system, accurate load forecasting plays vital role. Moving forward towards smart grid system, short term load forecasting has become even more significant due to increasing share of renewable energy sources which is highly uncertain due to climatic variations. This paper proposes a new approach in which co-relations of historical demand with "real feel index" and some additional weather factors are used to forecast demand with enhanced level of accuracy. In addition, this paper emphasizes as to how the real feel index proves to be a better factor rather than simply using ambient dry bulb temperature and relative humidity. This paper provides an algorithm for short-term load forecasting of actual UDM(Unrestricted Demand) and describes it for demand response applications. The performance of the proposed model is validated on the Unrestricted Demand of the state of Uttar Pradesh, India. This model offers its compatibility to prevalent grid regulations of the Indian power market.