Prediction of transmission line overloading using intelligent technique

With the worldwide deregulation of power system, fast line flows or real power ( MW) security assessment has become a challenging task for which fast and accurate prediction of line flows is essential. Since last few years, limit violation of voltage and line loading has been responsible for undesirable incidents of power system collapse leading to partial or even complete blackouts. Accurate prediction and alleviation of line overloads is the suitable corrective action to avoid network collapse. The control action strategies to limit the transmission line loading to the security limits are generation rescheduling/load shedding. In this paper, an intelligent technique based on cascade neural network (CNN) is presented for identification of the overloaded transmission lines in a power system and for prediction of overloading amount in the identified overloaded lines. The effectiveness of the proposed CNN based approach is demonstrated by identification and prediction of line overloading for different generation/loading conditions in IEEE 14-bus system. Once the cascade neural network is trained properly, it provides accurate and quick results for previously unseen loading scenarios during testing phase. (C) 2007 Elsevier B.V. All rights reserved.