Position Tracking System of Filling Machine Based on Compound Control Strategy

Abstract:

In view of the complexity and periodic motion of automatic filling machine, a novel compound control strategy based on single neuron PID model reference adaptive control and repetitive control is proposed. Diagonal recurrent neural network (DRNN) is used as on-line identifier of system for the single neuron PID controller to adjust its weights and PID parameters by self-learning and self-adapting. The dynamic state performance can be improved by adaptive PID controller based on DRNN on-line Identification and the steady state performance is improved by modified repetitive controller. Simulation results show that the control system has good ability of restraining disturbances and high position tracking precision and good robustness.

• Periodical: Applied Mechanics and Materials (Volumes 380-384)
• Pages: 321-324
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.380-384.321
• Online since: August 2013
• Authors: Zhi Qiang Wei, Dan Jin
• Keywords: DRNN, Filling Machine, Position Tracking, Repetitive Control, Single Neuron

References

[1] C.C. K and K.Y. Lee. Diagonal recurrent neural networks for dynamic systems control[J]. IEEE Trans On Neural Networks, January 1995, Vol. 6, pp.144-155.

DOI: 10.1109/72.363441

[2] Cuiyan Li, Dongchun Zhang, et al. Repetitive control-a survey[J]. Electric Machines and Control, January 2005, Vol. 9, No. 1, pp.37-44.

[3] Jinmei Li, Xingqiao Liu, et al. Application of an adaptive controller with a single Neuron in control of multi-motor synchronous system[C]. IEEE International Conference on Industrial Technology, ICIT 2008, April 2008, pp.1-6.

DOI: 10.1109/icit.2008.4608462

[4] Mingguang Zhang, Xinggui Wang, , et al. The self-tuning PID decoupling control based on the diagonal recurrent neural network[C]. Proceedings of the Fifth International Conference on Machine Learning and Cybernetics, August 2006, pp.3016-3020.

DOI: 10.1109/icmlc.2006.258357

[5] Shaosheng Fan, Xiao Hui. Diagonal recurrent neural network based predictive control for active power filter[J]. International Conferrence on Power System Technology, POWERCON 2004, Vol. 1, pp.759-762.

DOI: 10.1109/icpst.2004.1460093

[6] Hara S, Yamamoto Y, et al. Repetitive control system: a new type servo system for periodic exogenous signals[J]. IEEE Transactions On Automatic Control, 1988, Vol. 33, No. 7, pp.659-668.

DOI: 10.1109/9.1274

[7] Xi Xiao, Yongdong Li, et al. Performance control of PMSM drives using a self-tuning PID[C]. Electric Machines and Drives Conference, IEMDC 2005, May 2005, pp.1053-1057.

DOI: 10.1109/iemdc.2005.195852