Abstract

Abstract:With theincreases in competitive business environment, forecasting of demand isdirectly linked up with the portfolio of the generating companies andconsidered as an important parameter of power system regulation. To takedecisions like bidding of the power blocks in a pool market, trade decisions,generation schedules, dispatch and unit commitment operations require preciseknowledge of the system demand. It is also essential to consider an accuracy ofload forecasting to maintain balance in power system regulation. In this paperwe propose sensitivity analysis with an application of adaptive supervisedlearning technique “Generalized Regression Neural Network” (GRNN) to know theeffect of the change in parameters in forecasting engine performance. Differentinputs are changed uniformly and fair comparison is carried out on the basis ofstandard error index Mean Absolute Percentage Error (MAPE).

Keywords– Load Forecasting, Power system regulation, SensitivityAnalysis, GRNN, MAPE