Introduction of Availability Based Tariff (ABT), signifies the importance of frequency prediction by bringing in the concept of frequency sensitive unscheduled interchange (UI) charge of energy drawn in deviation from the pre-committed daily schedule. Accurate predicted frequency facilitates the system operators in the decision process of precise generation scheduling (GS). Traditional approaches of frequency prediction are not producing satisfactory results. In this paper we considered the dependency of frequency on various parameters that affect the frequency regime in power system. An Artificial Neural Network (ANN) based model (Back propagation network) has been used in this paper to solve this problem. The data obtained from North Regional Load Dispatch Center (NRLDC) for the period from January 2005 to December 2007 has been used for training, validating and testing the ANN model. The performance of proposed model has been analyzed using the error indices; Absolute Percentage Error (APE) and Mean Absolute Percentage Error (MAPE). Simulation results show the superiority of the proposed ANN model to solve the frequency prediction problem over the traditional techniques, in terms of reduced MAPE.

No abstract