Elasticity Analysis Modeling by Least Squares Support Vector Machines Applied to System Load

Abstract:

Load elasticity analysis is effective for quantitative dynamic analysis on the load volatility in electric power planning. However traditional linear methods are always based on the definite mathematical or physical models of the observed problem, which is not suitable for the complicated real cases, especially electric load. So this paper describes a novel model for the elasticity analysis on the electric load. First to overcome the limitations of traditional regression methods, load simulation model are proposed by LS-SVM that is a strong nonlinear regression tool and outperforms standard support vector machine in the regression accuracy and modeling velocity. Based on that model, numerical simulation is performed to simulate the changes in load and its influence factors in consistency with the latent mapping relationship. Then electric load sensitivity models are deduced. Finally, the case studies are used to test the proposed model.

• Periodical: Advanced Materials Research (Volume 186)
• Pages: 165-169
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.186.165
• Online since: January 2011
• Authors: Hua Zheng, Li Xie, Jun Xiong
• Keywords: Electric Power Load, Least Squares Support Vector Machines (Ls-SVM), Sensitivity Analysis, Simulation

References

[1] LUO Zhi-qiang, ZHANG Yan, and ZHU Jie, in: Application of Rough Set Theory in Electricity Power Load Forecasting, submitted to Power System Technology(2004), Vol. 28, pp.29-32.

[2] R.W. Dahlgren, C. -C. Liu, and J. Lawarree, in: Volatility in The California Power Market: Source, Methodology And Recommendations, submitted to IEE Proc. -Gener. Transm. Distrib. (2001), Vol. 148, pp.189-193.

DOI: 10.1049/ip-gtd:20010339

[3] A.B. Smith, C.D. Jones, and E.F. Roberts, in: Voltage And Current Sensitivities Of Rradial Distribution Network _ A New Approach, submitted to IEEE Transactions on Circuits and System-II: Express Briefs(2004), Vol. 51, pp.421-425.

[4] Fangxing Li, Richard E. Brown, and Lavelle A. A. Freeman, in: A Linear Contribution Factor Model of Distribution Reliability Indices And Its Applications in Monte Carlo Simulation And Sensitivity Analysis, submitted to IEEE Transactions on Power System(2003).

DOI: 10.1109/tpwrs.2003.814906

[5] D.Z. Fang, S.Q. Yuan, T.S. Chung and K.P. Wong, in: Sensitivity-based Hybrid Approach for Assessment of Stability-constrained Generation Limit, submitted to IEE Proc. -Gener. Transm. Distrib. (2006), Vol. 153, pp.368-374.

DOI: 10.1049/ip-gtd:20045304

[6] V. Kumar, I. Gupta, H.O. Gupta and C.P. Agarwal, in: Voltage and Current Sensitivities of Radial Distribution Network: A New Approach, submitted to IEE Proc. -Gener. Transm. Distrib. (2005), Vol. 152, pp.813-818.

DOI: 10.1049/ip-gtd:20045068

[7] Marcelo Espinoza, Johan A. K Suykens, and Bart De Moor, in: Least Squares Support Vector Machines and Primal Space Estimation, 2003 IEEE Conf. on Decision and Control, pp.3451-3456.

DOI: 10.1109/cdc.2003.1271680

[8] L.J. Cao, Francis E.H. Tay, in: Support Vector Machine With Adaptive Parameters in Financial Time Series Forecasting, submitted to IEEE Trans. on Neural Networks(2003), Vol. 4, pp.1506-1518.

DOI: 10.1109/tnn.2003.820556