Title :
Electric Load Prediction Using a Bilinear Recurrent Neural Network
Author :
Kim, Jae-Young ; Park, Dong-Chul ; Woo, Dong-Min
Author_Institution :
Dept. of Electron. Eng., Myong Ji Univ., Yongin, South Korea
Abstract :
A prediction scheme of electric load using a Bilinear Recurrent Neural Network (BRNN) is proposed in this paper. Since the BRNN is based on the bilinear polynomial, BRNN has been successfully used in modeling highly nonlinear systems with time-series characteristics. Dynamic BRNN further improves the convergence of BRNN and the Dynamic BRNN can be a natural candidate in predicting electric load. The performance of the proposed BRNN-based electric load prediction scheme is evaluated and compared with the conventional MultiLayer Perceptron-type Neural Network (MLPNN)-based predictor in this paper. Experiments are conducted on a load data set from the North-American Electric Utility (NAEU). The results show that the Dynamic BRNN-based predictor outperforms the MLPNN-based predictor in terms of the Mean Absolute Percentage Error (MAPE).
Keywords :
load forecasting; neural nets; power system analysis computing; BRNN; MLPNN; North-American Electric Utility; bilinear polynomial; bilinear recurrent neural network; electric load prediction; mean absolute percentage error; multilayer perceptron-type neural network; Artificial neural networks; Autoregressive processes; Convergence; Economic forecasting; Load forecasting; Multi-layer neural network; Neural networks; Nonlinear systems; Predictive models; Recurrent neural networks; electricity; forecasting; neural network; recurrent;
Conference_Titel :
Computer Modelling and Simulation (UKSim), 2010 12th International Conference on
Conference_Location :
Cambridge
Print_ISBN :
978-1-4244-6614-6
DOI :
10.1109/UKSIM.2010.81