Rule-Based Load Forecasting: Malaysian Electricity Company’s Journey

RuleBaseD loaD FoRecasTiNG: malaysiaN elecTRiciTy comPaNy’s JouRNey By Zuhaimy ismail and Fuad Jamaluddin Establishes rules for load forecasting by developing relationships between input and output … the morning peak demand occurs between 11:00 AM and 12:00 noon … rulebased forecasts at times give forecasts better than those statistically developed. S S hort-term load forecasting (STLF) — which is anywhere from one hour to one week — plays an important role in the operation of a power system. The Malaysian Electricity Company (MEC)’s core business is the generation, transmission, and distribution of electricity; as such, it is a key supplier of electricity in Malaysia. It uses various statistical models including Box Jenkins and exponential smoothing to forecast the load demand. In recent years, we successfully experimented with the rule- based forecasting approach to forecast load demand, which is the topic of this article. Forecasting is one of the important functions at MEC, and the short-term load forecasting is one of the problems. To solve this problem, we tested various time series and regression models. In recent years, we experimented with the rule-based ...