Tension Control Based On Soft Computing

Tension control is a very important part of process control in printing and packing service. The control of the stability tension is one of the crucial techniques ensuring the product quality. The tension control system is a non-linear system with the characteristics of multi-input, coupling and time variable, a satisfactory control effect can hardly be attained from tension control system designed with conventional control theory. While the neuro-fuzzy control system has the characteristics of adapting to any indefinite system, the theory of the neuro-fuzzy is applied on the tension control system. It is applied on the jointing-printing tension control system.As only few researches on the jointing-printing tension control system of the Laminating Machine are published, this article analyses the composition and structure, working principle and dynamic model of the jointing-printing tension control system, and have established the tension control system's models. This paper puts forward the tension control method based on soft computing theory; the structure of fuzzy neural network is built; and it tunes membership functions, rules and operators to improve control effect. The simulation of three tuning systems is carried out, and the simulation results are compared among the tuning system, non-tuning system and PID system. There analyses those advantages and disadvantages.It has been testified by the simulation analysis and the field produce, the results prove that the tension control system based on neuro-fuzzy control theory effectively overcome the defaults of the conventional control method and increase the response velocity and the control stability of the control precision of the whole system, the run is reliable and the deviation of tension is small, thus reach the required property parameter of the system.