Load modeling from smart meter data using neural network methods

Electricity load modeling plays a critical role to conduct load forecasting or other applications such as non-intrusive load monitoring. For such a reason, this paper investigates a comparison study of two common artificial neural network methods (Multilayer perceptron (MLP) and radial basis function neural network (RBF-NN) for home load modeling application. The accuracy of load modeling using neural network methods highly depends on chosen variables as the input data set for the networks. For this purpose, data including weather, time, and consumer behavior are considered as the input dataset to train the networks. The results of this study show that the RBF-NN model has higher accuracy in training data. On the other side, the MLP model outperforms in test data. To sum up, the results prove that the load model obtained by MLP has a better performance in terms of mean square and root mean square error indices.