NLARX Model Identification for Grid Tied Inverter Using Genetic Algorithm

Abstract:

This research studies the parameter identification for establish the model of inverter control system which has grid-connected in distributed power systems. Identification is a method of measuring the mathematical description of a system by processing the observed inputs and outputs of the system. The model developed is based on nonlinear autoregressive with exogenous input (NLARX) for simulate the inverter control system which has grid-connected in distributed power systems. To analyze the model, we use one, three and five steps ahead analysis technique which is the tool in mathematical model analysis for model categories of NLARX. In this research, we present comparative analysis of the model using NLARX in 3 different categories that is wavelet network, tree partition network and sigmoid network. The simulation results show that the optimal category of NLARX is sigmoid network which has one, three and five steps ahead value is equal to 97.01, 91.50 and 88.90 percent respectively from output model and the measured output from the validation data set from grid tied inverter mathematical modelling using genetic algorithm. This optimal model can be used to simulate for analyze trends of the output data which can represent for the real system.

• Periodical: Applied Mechanics and Materials (Volume 804)
• Pages: 325-328
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.804.325
• Online since: October 2015
• Authors: Viritpol Vacharapanich*, Naris Pratinthong, Dhirayut Chenvidhya
• Keywords: Genetic Algorithm, Grid Tied Inverter, NLARX Model, Step Ahead

References

[1] W. Zheng, Dynamic converter modeling using black-box identification method for grid-connected photovoltaic systems, Journal of computational information systems. 9 (2013) 9083-9093.

[2] F.D. Kanellos, Wind parks equivalent ARX models of power systems with large wind power using identification theory, Electric power systems research. 81 (2011) 707-715.

DOI: 10.1016/j.epsr.2010.10.035

[3] X. Xiong, System identification for NARX model of photovoltaic grid-connected inverter, Power system technology. 37 (2013) 2440-2445.

[4] M. Karthik, Dynamic neural network based parametric modeling of PEM fuel cell system for electric vehicle applications, IEEE. 14394641 (2014) 1-5.

DOI: 10.1109/icaee.2014.6838559

[5] J.N.M. Francoise, DC to DC converter with neural network control for on-board electrical energy management, IEEE. 8280625 (2004) 521-525.

[6] O. Nelles, Nonlinear system identification: from classical approaches to neural networks and fuzzy models, Springer, Berlin, (2001).

[7] Information on http: /www. mathworks. com.

[8] X.S. Yang, Nature-inspired metaheuristic algorithms, Luniver Press, London, (2008).

[9] S. Koziel, X.S. Yang, Computational optimization and applications in engineering and industry, Springer, Berlin, (2011).

[10] X.S. Yang, Nature-inspired optimization algorithms, Elsevier, London, (2014).