Neuro-fuzzy looper control with T-operator and rule tuning for rolling mills: theory and comparative study

Looper control is widely used for rolling tension control, but conventional control performance usually degrades with the rolling parameter variations. Fuzzy control systems were proposed in Janabi-Sharifi and Fan (2000) and outperformed conventional control systems in terms of disturbance rejection and decreased steady state error. In this paper, hybrid neuro-fuzzy control is proposed for the T-operator and rule-tuning. Simulation results are presented to compare the performance of different operators and rule-tuning versus T-operator tuning. Also, the responses of the proposed systems and the above mentioned fuzzy control method are compared and practical conclusions are made