Time Series Pattern Learning and Forecasting for Long-Term Peak Electricity by Spectral Mixture Gaussian Kernel

Tuchsanai Ploysuwan; Prasit Teekaput; Pramukpong Atsawathawichok

doi:10.4028/www.scientific.net/AMM.781.245

Paper Titles

Laser Cleaning of Polypropylene Contamination on Magnetic Head Slider Surface
p.227

Determination of Transition Jitter Noise by Time Domain Analysis
p.231

The Synthesis of Nano-Porous Alumina by Anodization Process
p.235

Modelling and Simulation of the 150° Conduction Mode Quasi Space Vector PWM Inverter Fed Permanent Magnet Brushless DC Motor
p.241

Time Series Pattern Learning and Forecasting for Long-Term Peak Electricity by Spectral Mixture Gaussian Kernel
p.245

Grounding Design Improvement to Reduce Back Flashover Rate of 69 kV Subtransmission Line in Power’s Distribution System
p.250

Resonance Impact on Power Factor Correction System in Power System with Harmonic Distortion
p.254

Substation Expansion Considering Reliability of Various Bus Configurations and Load Growth
p.258

Model of Power Factor Charge for Photovoltaic Generation System Based on its Contribution to Power Systems
p.262

HomeApplied Mechanics and MaterialsApplied Mechanics and Materials Vol. 781Time Series Pattern Learning and Forecasting for...

Time Series Pattern Learning and Forecasting for Long-Term Peak Electricity by Spectral Mixture Gaussian Kernel

Abstract:

This paper presents the mathematical model for forecasting of future long-term peak electricity load from January 2014 to December 2024 with totally 132 months from the past knowledge data of training 156 months. The new kernel method is proposed by the combination ofsummed weight spectral mixture Gaussian in the frequency domain and squared exponential in the time domain, which are used as components in the answer of Gaussian Process (GP). Finally, the results show the prediction error mean absolute percentage error (MAPE) by 2.3283%.

You might also be interested in these eBooks

View Preview

Info:

Periodical:

Applied Mechanics and Materials (Volume 781)

Pages:

245-249

DOI:

https://doi.org/10.4028/www.scientific.net/AMM.781.245

Citation:

Cite this paper

Online since:

August 2015

Authors:

Tuchsanai Ploysuwan, Prasit Teekaput, Pramukpong Atsawathawichok*

Keywords:

Gaussian Process, Kernel Function, Long-Term Load Forecasting, Peak Load

Export:

RIS, BibTeX

Price:

Permissions CCC:

Request Permissions

Permissions PLS:

Request Permissions

Сopyright:

Citation:

* - Corresponding Author

References

[1] Annual Report , Electricity Generating Authority of Thailand (EGAT) 2000-(2013).

Google Scholar

[2] M. A. Mamun and K. Nagasaka, Artificial neural networks applied to long-term electricity demand forecasting, in Hybrid Intelligent Systems, 2004, pp.204-209.

DOI: 10.1109/ichis.2004.27

Google Scholar

[3] Z. Shi and W. Dingwei, Medium and Long-Term Load Forecasting Based on PCA and BP Neural Network Method, in Energy and Environment Technology, 2009, pp.389-391.

DOI: 10.1109/iceet.2009.559

Google Scholar

[4] L. Ghelardoni, A. Ghio, and D. Anguita, Energy Load Forecasting Using Empirical Mode Decomposition and Support Vector Regression, IEEE Trans on Smart Grid vol. 4, pp.549-556, (2013).

DOI: 10.1109/tsg.2012.2235089

Google Scholar

[5] P. Atsawathawichok, P. Teekaput, and T. Ploysuwan, Long term Peak Load Forecasting in Thailand using Multiple Kernel Gaussian Process, in IEEE ECTI-CON, 2014, pp.210-213.

DOI: 10.1109/ecticon.2014.6839869

Google Scholar

[6] C. E. Rasmussen and C. K. I. Williams, Gaussian processes for machine learning. Cambridge, Mass.: MIT Press, (2006).

Google Scholar

[7] S. Bochner, Lectures on Fourier integrals; with an author's supplement on monotonic functions, Stieltjes integrals, and harmonic analysis: Princeton University Press, (1959).

DOI: 10.1515/9781400881994-012

Google Scholar

[8] A. G. Wilson and R. P. Adams, Gaussian process covariance kernels for pattern discovery and extrapolation , in ICML Conference, (2013).

Google Scholar