The Application of GA-RBF Neural Network Generalized Predictive Control Strategy in Circulating Fluidized Bed Unit

Abstract:

Circulating fluidized bed boiler, with heat storage capacity, nonlinear, large time delay, multi-variable and strong coupling characteristics, is a complex quantity control system. It is difficult to establish accurate model when the parameters of control object is uncertainty because of all disturbances. This article uses one way to identify model based on genetic algorithm and RBF network combined with generalized predictive control strategy, through online rolling optimization and feedback correction, to achieve predictive control. The results of the simulation show that it has strong robustness when the load condition changes, the parameter changes or lag affects.

• Periodical: Applied Mechanics and Materials (Volumes 496-500)
• Pages: 1582-1586
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.496-500.1582
• Online since: January 2014
• Authors: Yin Bo Liu*, Hu Jun Ling, Jing Hua Li, Chun Yuan Su, Zhong Yang Yuan
• Keywords: Circulating Fluid Bed Unit, Generalized Predictive Control, Genetic Algorithm (GA), RBF Neural Network

References

[1] Yugeng Xi, Predictive Control [M], National Defence Industry Press, 6. (1993).

[2] Li Ning, Ling Hujun. Study of an algorithm of GA-RBF neural network generalized predictive control for Generating Unit [C]. Piscataway, NJ: IEEE Computer Society, 2011, 1723-1726.

DOI: 10.1109/iceice.2011.5777819

[3] Kamalabady, A. Sabet, Salahshoor, L. New SISO and MIMO adaptive Generalized Predictive controllers based on self-organizing RBF neural networks[C]. NJ: Inst. of Elec. and Elec. Eng. Computer Society, 2008, 374-379.

DOI: 10.1109/icnsc.2008.4525243

[4] Xinwei Geng, Yufei Zhang. Intelligent PID Design for Coordinated Control System of 300 MW CFB Power Unit [J]. Power Equipment, 2012, 02: 105-107.