【作者】 吴海列；

【导师】 欧长劲；

【作者基本信息】 浙江工业大学 ， 机械制造及自动化， 2006， 硕士

【摘要】 气体辅助注塑成型是一种为了克服传统注塑成型的局限性而发展起来的利用高压气体在注塑件内部产生中空截面的成型技术。由于在成型工艺过程中提出了新的要求，其中又以气体注射压力对气辅成型产品的质量影响最大，这使得对气体压力的精确控制变得尤为重要。本文概述了气辅注射成型原理及装置，并重点介绍了气体压力控制系统。在其物理模型的基础上，建立了气辅注射气体压力控制系统的数学模型，对系统的数学模型进行了稳定性和可控性分析，并采用Ackerman状态反馈设计方法，对系统的稳态性能和动态性能进行改善。根据气辅成型压力控制的特点，进行不同气体压力控制方法的研究，提出了采用模糊控制、模糊神经网络控制对经过状态反馈后的气体注射压力系统进行智能控制的策略，并对其优缺点进行研究和分析。在Simulink环境下建立气体压力控制系统的模型。对PID控制、模糊控制、模糊神经网络控制三种控制器的性能做了仿真研究，并对气辅注射的三段气体压力控制进行仿真，结果表明，模糊神经网络具有更好的控制性能。文章最后提出了气体压力控制系统的实验软件设计方法，并对系统的模块生成以及运行情况进行分析。更多还原

【Abstract】 Gas-assisted injection molding (GAIM) technology, being an innovative injection molding technology, uses compressed gas to create hollowed voids in the plastic part. Compared with conventional injection molding, there are some new gas-related processing parameters, among which, the pressure of injected gas plays the most important role. To significantly improve the quality of product, it requires the accurate monitoring and control of gas injection pressure.The principle and the equipment of gas injection molding are outlined first, and then the gas pressure control system (GPCS) is elaborated. Thus a mathematical model of GPCS that can describe the gas pressure during the process is derived; furthermore, the stability and controllability based on this model are analyzed. To improve the dynamic performance and steady-state performance, a controller using Ackerman state feedback approach is designed. This paper applies some advanced control methods, such as fuzzy control and fuzzy neural network control to the control of gas injection pressure. The advantages, disadvantages and algorithms of the two control methods are discussed.Simulations are carried out by computer program written in Simulink, a toolbox of Matlab. The performances of the three control methods mentioned above on the control of gas pressure under three-segment reference signal are compared. The results demonstrate that fuzzy neural network control performs best in GAIM process.At last, a method of how to design the experimental software for GPCS is presented in this thesis. And a friendly man-machine interface designed by Visual C++ shows good operation of this system in project.更多还原