Evaluation of Short-Term Load Forecasting Methods Using Dynamic Neural Networks

This paper presents forecasting of short-term electricity load using dynamic neural networks, DNN, and an assessment of the neural network stability to ascertain continued reliability. This includes an assessment and comparative study of three different neural networks, feedforward, Elman and the radial basis neural network. The performance and stability of each DNN is evaluated by means of an extensive simulation study using actual hourly load data. The neural networks weights are dynamically adapted. Stability for each of the three different networks is determined through Eigen values analysis. Evaluation of the neural network methods is done in terms of estimation performance, stability and the difficulty in training a particular network. The results show that the radial basis method performs better than the rest Eigen value analysis also shows that it is more reliable as it remains stable as the input varies.