Medium-term Load Forecasting Method with Improved Deep Belief Network for Renewable Energy
DOI:
https://doi.org/10.13052/dgaej2156-3306.3735Keywords:
IDBN-NN, renewable energy, power system, load forecasting, restricted Boltzmann machine, deep belief network.Abstract
With the continuous transition of the traditional power system to the new
power system, the composition of the power generation side in the power
system has gradually begun to be dominated by renewable energy (at least
more than 50%). Among the renewable energy sources, wind power is the
most susceptible to weather and environmental influences. These factors
increase the complexity of the power generation mode, and put forward
higher requirements for the accuracy and stability of load forecasting. This
paper proposes a medium-term renewable energy load forecasting method
based on an improved deep belief network (IDBN-NN). The method includes
the construction of a deep belief network, the layer-by-layer pre-training
and fine-tuning of model parameters, and the application of the model.
In the process of model parameter pre-training, Gauss-Bernoulli Restricted
Boltzmann Machine (GB-RBM) is used as the first part of the stacked deep
belief network, so that it can process multiple types of real-valued input data more effectively. In addition, IDBN-NN uses a combination of unsupervised
training and supervised training for pre-training. Finally, the actual load data
is used to analyze the calculation example. When the number of RBM layers
is 3, the number of fully connected layers is 1, and Dropout is equal to
0.2, the MSE and loss values are optimal, which are 0.0037 and 0.0104,
respectively. The experimental results show that the proposed method has
higher prediction accuracy when the training sample is large and the load
influencing factors are complex.
Downloads
References
C. Q. Kang, Q. Xia, M. Liu. Power system load forecasting. Automation
of Electric Power Systems, 2007, 6(16):457–467.
T. Hong, P. Wang, Willis H. L. A naive multiple linear regression
benchmark for short term load forecasting [C]// IEEE Power and Energy
Society General Meeting, July 24–29, 2011, Detroit, USA: 1–6.
Q. H. Wu, J. Jun, G. S. Hou, B. Han, K. Y. Wang. Online Recognition of
Human Actions Based on Temporal Deep Belief Neural Network. Power
system automation, 2016, 40(15):67–72.
L. Hernandez, C. Baladron, J. M. Aguiar, et al. Artificial neural net-
work for short-term load forecasting in distribution systems[J]. Energies,
, 7(3):1576–1598.
J. Liu, H. Gao, M. A. Zhao. Review and prospect of active distribu-
tion system planning[J]. Journal of Modern Power Systems and Clean
Energy, 2015, 3(4):457–467.
T. Y. Zhou, J. Q. Yi, Y. Yang, H. T. Zhang, X. F. Yuan. Online Recogni-
tion of Human Actions Based on Temporal Deep Belief Neural Network.
Acta Automatica Sinica, 2016, 15(7):1030–1039.
Z. Q. Geng, Y. K. Zhang. An Improved Deep Belief Network Inspired
by Glia Chains. Acta Automatica Sinica, 2016, 4(6):943–952.
G. E. Hinton, S. Osindero, Y. W. Teh. A fast learning algorithm for deep
belief nets[J]. Neural Computation, 2006, 18(7):1527–1554.
W. Liu, Z. Wang, X. Liu. A survey of deep neural network architectures
and their applications[J]. Neurocomputing, 2016, 234:11–26.
Z. H. Rong, B. Qi, C. R. Li, S. J. Zhu, Y. F. Chen. Combined DBN
Diagnosis Method for Dissolved Gas Analysis of Power Transformer
Oil. Power System Technology, 2019, 23(10):3800–3808.
J. Q. Shi, T. Tan, J. Guo, Y. Liu, J. H. Zhang. Multi-Task Learning
Based on Deep Architecture for Various Types of Load Forecasting in
Regional Energy System Integration. Power System Technology, 2018,
(3):698–707.
X. B. Zhang, J. Tang, C. Pan, X. X. Zhang, M. Jin. Research of Par-
tial Discharge Recognition Based on Deep Belief Nets. Power System
Technology, 2016, 40(10):3283–3289.
T. Kuremoto, S. Kimura, K. Kobayashi, et al. Time series forecasting
using a deep belief network with restricted Boltzmann machines[J].
Neurocomputing, 2014, 137(15):47–56.
X. Qiu, L. Zhang, Y. Ren, et al. Ensemble deeplearning for regression
and time series forecasting[C]// IEEE Symposium on Computational
Intelligence in Ensemble Learning, December 9–12, 2014, Orlando,
USA: 1–6.
Dedineca, S. Filiposka, A. Dedinec, et al. Deep belief network based
electricity load forecasting: an analysis of Macedonian case[J]. Energy,
, 115:1688–1700.
G. E. Hinton. A practical guide to training restricted Boltzmann
machines[J]. Momentum, 2012, 9(1):599–619.
X. Y. Kong, F. Zheng, Z. J. E, J. Cao, X. Wang. Short-term Load
Forecasting Based on Deep Belief Network. Power System Technology,
, 8(5):133–139.
J. X. Zhao, X. Zhang, F. Q. Di, S. S. Guo, X. Y. Li. Exploring the
Optimum Proactive Defense Strategy for the Power Systems from an
Attack Perspective. Security and Communication Networks, 2021(1):
–14.
G. E. Hinton, R. R. Salakhutdinov. Reducing the dimensionality of data
with neural networks[J]. Science, 2006, 313(5786):504–507.
Jinxiong Zhao, Sensen Guo, Dejun Mu. DouBiGRU-A: Software Defect
Detection Algorithm Based on Attention Mechanism and Double
BiGRU, Computers & Security, 2021, 13(5786):504–517.
You Weijing, Liu Limin, Ma Yue, et al. An Intel SGX-based Proof of
Encryption in Clouds. Netinfo Security, 2020, 20(12):1–8.
Xu Guotian, Shen Yaotong. A Malware Detection Method Based on
XGBoost and LightGBM Two-layer Model[J]. Netinfo Security, 2020,
(12):54–63.
Li Hongjiao, Chen Hongyan. Research on Mobile Malicious Adver-
sarial Sample Generation Based on WGAN[J]. Netinfo Security, 2020,
(11):51–58.
Guo Qiquan, Zhang Haixia. Technology System for Security Protec-
tion of Critical Information Infrastructures[J]. Netinfo Security, 2020,
(11):1–9.
