Electrical load forecasting based on self-adaptive chaotic neural network using Chebyshev map
- 150 Downloads
The importance of electrical load forecasting stems from energy planning and formulating strategies in power system. In this paper, a novel chaotic back-propagation (CBP) neural network algorithm based on the merit of Chebyshev map is proposed. To improve the accuracy of proposed algorithm, self-adaptive gradient correction method is used to eliminate the precocious phenomenon of network. An additional inertial term including chaotic sequence is increased in the process of optimizing the weight value and threshold value of network. The ergodicity of chaotic variables within the range of [−1, 1] can decrease the oscillation trend of network, accelerate the learning speed and overcome the fake saturation problem so as to greatly improve the forecasting ability of proposed algorithm. The simulation results of actual cases indicate that the proposed CBP neural network is advantageous in many respects in comparison with the previous methods studied.
KeywordsLoad forecasting Chebyshev map Chaotic neural network Self-adaptive gradient correction
This paper is funded by the National Natural Science Foundation (No.71401049), the Anhui Provincial Natural Science Foundation (No.1408085QG137, 1408085MG136), the CRSRI Open Research Program (Program SN CKWV2014213/KY), the Open Research Fund of State Key Laboratory of Simulation and Regulation of Water Cycle in River Basin(China Institute of Water Resources and Hydropower Research, Grant NO IWHR-SKL-201605), and the Specialized Research Fund for the Doctoral Program of Higher Education (No. 20130111120015).
- 20.EUNITE network (2001) http://neuron.tuke.sk/competition/index.php
- 21.Zhai YJ, Liu L, Wang P (2012) Mid-term load forecasting based on ADE-SVM and fuzzy theory. Power Syst Prot control 40(8):110–115Google Scholar
- 22.He YY, Xu QF, Yang SL, Yu BG (2013). A power load probability density forecasting method based on RBF neural network quantile regression. In Proceedings of the CSEE 33(1), 93–98Google Scholar
- 23.Liu R, Fang GF (2012) Short-term load forecasting with comprehensive weather factors based on improved Elman neural network. Power Syst Protect Control 40(22):113–117Google Scholar
- 27.He YY, Xu QF (2012) Short-term power load forecasting based on self-adapting PSO-BP neural network model. In: Proceedings—4th international conference on computational and information sciences, pp 1096–1099Google Scholar