Research On Temperature Distribution Measurement And Control In Extrusion Process

Polymer extrusion molding is an important process in the polymer manufacturing industry. The advancement of modern science and technology and market requirement impel extrusion technology towards parallel development of three directions in high speed, high efficiency and precise extrusion. Dynamic extrusion process was a complex, nonlinear and strong coupling multivariable process. Temperature of molten polymers is one important process parameters in extrusion process. Melt temperature also is not only affected stabilization of extrusion and rheological property but also directly affected the quality of extrusion product.So far, there is not effective technology of measurement in molten temperature distribution. In this paper, a self developed sensor based thermocouple nets and the temperature acquiring system were used research on temperature distribution measurement in dynamic extrusion process. Based on a large number of experimental, we researched influence of die segment temperature, screw speeds and electric magnetic vibration parameters on LDPE molten temperature distribution and the effect of polymer melt export pressures were analyzed. In this paper, we put forward four modeling methods based on BP neural networks, RBF neural networks, samples matrix nonlinear regression analysis and adaptive network-based fuzzy inference system. The molten temperature distribution prediction models were established in dynamic extrusion process by using the four methods combined with numerical analysis method. Prediction and simulation on melt temperature distribution results showed that, the four modeling methods were feasible and effective. The merits and demerits of each model are also analyzed.Extruder temperature control system has multivariable, large time delay, nonlinear, strong coupling characteristics. It is hard to establish its accurate model by using traditional methods. In this paper, extruder temperature control system was used by fuzzy control method. Fuzzy controller rules and control algorithm were established. The Fuzzy control method was simulated by MATLAB, simulation results that fuzzy control has a good dynamic response, stability characteristics, strong anti-interference ability and good robustness.The study aimed to further research on dynamic rheological theory in dynamic extrusion process. It applied a new method and theoretical foundation for inline measurement and control in temperature distribution of extrusion. Furthermore, it is extended significance to implement precision extrusion technology and precision injection molding technology.