Spatial load forecasting based on unstructured information processing and multi — Attribute deep learning

A spatial load forecasting method based on unstructured information processing and multi — attribute depth learning is proposed. In order to solve the problem that the unstructured attributes have great influence on the load density but can't be directly put into calculation, the natural language processing (NLP) technique is used to structure those attributes. In view of the on characterization of the high dimension attributes by traditional method, Stacked Denoising Auto Encoder (SDAE) deep learning network is used to forecast the spatial load density. And the Rectified Linear Unit(ReLU) function is used as the excitation function of the network as well as the network structure gets improved to overcome the gradient disappearance and over-fitting. The results of case study show that the method of spatial load forecasting is effective and feasible.