Short-term load forecasting model based on deep neural network

• Published in: 2017 2nd International Conference on Power and Renewable Energy (ICPRE) pp. 589 - 591
• Main Authors: Hui, Xue, Qun, Wang, Yao, Li, Yingbin, Zhang, Lei, Shi, Zhisheng, Zhang
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.09.2017
• Subjects: Biological neural networks; deep neural network; electric power system; Forecasting; genetic algorithm; Load forecasting; Load modeling; Predictive models; short-term load forecasting
• DOI: 10.1109/ICPRE.2017.8390603

The short-term load forecasting model based on deep neural network is constructed in this paper. The forecasting model contains multiple hidden layers, which can extract the deep characteristics of the data. This allows the deep neural network to contain more information, extends the modeling ability of neural networks, and improves the predictive ability of the model. Genetic algorithm is used to optimize the weights and the thresholds of deep neural network. Using actual load data of electric power system for simulation test, compared with the conventional BP neural network, the test results show that the short-term load forecasting model based on deep neural network has good prediction effect.