Adaptive Weighting Decoupling Design for Constrained Generalized Predictive Control Using Improved Genetic Algorithm

Abstract:

The paper proposes a new method for decoupling multivariable system based on generalized predictive control (GPC) with constrains. It is the main idea of proposed control method that the error weight can change with output deviation caused by reference changes in order to reduce interactions in the system and improve dynamic performance of coupling loops. With improved genetic algorithm to optimize the performance index of GPC, the algorithm is applied to auto shape control and auto gauge control (ASC-AGC). The simulation results demonstrate the efficiency and correctness of approach proposed.

• Periodical: Applied Mechanics and Materials (Volume 441)
• Pages: 829-832
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.441.829
• Online since: December 2013
• Authors: Yong Bin Dai
• Keywords: Adaptive Error Weighting, Constraint, Decoupling, Generalized Predictive Control, Genetic Algorithm (GA)

References

[1] SU Baili, CHEN Zengqiang, YUAN Zhuzhi: Chinese J. Chem. Eng., Vol. 14, No. 2(2006), p.216. In Chinese.

[2] Wang dongfeng: stems Engineering and Electronics, Vol. 123 , No. 12(2001), p.65. In Chinese.

[3] Jin Qibing, Cheng Yunwa: Computer Measurement & Control, Vol. 13, No. 6( 2005), p.565. In Chinese.

[4] Martinez M: Engineering Applications of Artificial Intelligence, Vol. 11, No. 3(1998), p.355.