The intrinsic uncertainties associated with demand
forecasting become more acute when it is required to provide an invaluable
dimension to the decision-making process in a period characterized by fast and dynamic changes. This dynamic nature of the demand for
electricity call for more research in areas such as the estimation methodology of peak demand, factors affecting load demand, the system used, the organizational structure involved in forecasting and the types of forecasting methods. One of the main factors that influence electric demand is types of day such as weekday, weekend and holiday. The behavior patterns of the end-user are based on a few actual observations, which are divided into daily and annual weekly load profile. The annual daily and weekly load profile is used to explain the general lifestyle throughout the year. The daily load profiles for a typical week and during festive season, is used to explain behavioral pattern during a typical week and an hourly load profile is used to explain the lifestyle on a normal working day. This paper presents the influence of types of day in forecasting electricity load demand in Malaysia. We use separate forecasting model to forecast day and named the weekday and weekend model. In this study we include the analysis of load demand using
exponential smoothing model of electricity load forecasting. We use several exponential techniques to forecast short-term electricity demand and compare performance among them. The data used was the 1826 daily maximum demand time series from 1 January 2001 to 31 December 2005. A case study was conducted at Tenaga Nasional Berhad (TNB), a public listed electric company who builds, operates and maintains electricity transmission and distribution network in Peninsular Malaysia. TNB has the largest generation capacity of over 10450 MW that accounts for about 60% of the total power installed (17300MW). The other 40% is provided by the Independent Power Producers (IPPs). The result shows that Log seasonal Exponential Smoothing is a reasonable good model for forecasting daily load demand.