An Improved SA-Based BP-ANN Technique for Annual Runoff Forecasting Under Uncertain Environment

In this paper, the author presents an integrated approach combining the simulated annealing method and the feed forward neural network to forecast the annual runoff in power system under uncertain environment. The type of neural network used in this method is a multi-layer pre-trained by the SA. Finally, we use the SA-based ANN to see if we actually could reduce the error of annual runoff forecasting. The proposed Simulated Algorithm-based Error Back Propagation Artificial Neural Net (SA-based BP-ANN) annual forecasting scheme was tested using data obtained from a case study including 24 h time periods. The result demonstrated the accuracy of the proposed annual runoff forecasting.