Deep Structure of Gaussian Kernel Function Networks for Predicting Daily Peak Power Demands
This paper proposes a novel method of predicting daily peak power demands using the deep structure of Gaussian kernel function networks (GKFNs). For the prediction model, the whole time series is divided into multiple parts and each part is trained using a GKFN. Then, the trained GKFNs are combined using the deep structure of GKFNs to minimize the mean square errors (MSEs) of prediction model. As a consequence, the proposed deep structure of GKFNs provides an improved performance of prediction accuracy compared with canonical GKFNs. The simulation for predicting daily peak power demands in Korea reveals that the proposed prediction model has the merits in prediction performances compared with the GKFN model and also other prediction models such as the k-NN and SVR.
KeywordsDaily peak power demand Prediction model Gaussian kernel function network Deep structure
This work was partly supported by Institute for Information & communications Technology Promotion (IITP) grant funded by the Korea government (MSIT) (No. 2016-0-00133, Research on Edge computing via collective intelligence of hyperconnection IoT nodes).
- 1.Srinivasan, D., Lee, M.A.: Survey of hybrid fuzzy neural approaches to electric load forecasting. In: IEEE International Conference on Systems, Man and Cybernetics, pp. 4004–4008. IEEE Press, New York (1995)Google Scholar
- 2.Srivastava, A.K., Pandey, A.S., Singh, D.: Short-term load forecasting methods: a review. In: IEEE International Conference on Emerging Trends in Electrical Electronics & Sustainable Energy Systems, pp. 130–138. IEEE Press, New York (2016)Google Scholar
- 14.Kil, R., Park, S., Kim, S.: Time series analysis based on the smoothness measure of mapping in the phase space of attractors. In: International Joint Conference on Neural Networks, vol. 4, pp. 2584–2589. IEEE Press, New York (1999)Google Scholar