Comparison of Various Smoothing Parameter Techniques for Forecasting Power System States

Abstract

This paper presents a comparison of various smoothing parameter techniques for power system state forecasting. The time behavior of the power system states have been forecasted through several smoothing parameter techniques i.e. Brown’s one parameter method, Three parameter winter, Three parameter multiplicative method and Holt’s two-parameter method. The filtering problem has been resolved through a novel approach of Dynamic State estimation (DSE) relied on linear Kalman filter algorithm developed through optimally placed PMU measurements. Optimal location of PMUs has been obtained utilizing Integer linear programming (ILP) based approach. The proposed approach of DSE has been applied to IEEE 57, IEEE 118 and an Indian practical system i.e. 38 bus system of Damodar Valley Corporation (DVC). Among the various prediction techniques, Brown’s one parameter method provides better prediction compared to other forecasting methods.