【作者】 王硕；

【导师】 陈子辰； 应济；

【作者基本信息】 浙江大学 ， 机械制造及其自动化， 2012， 博士

【摘要】 注射成型是塑料制品的主要加工方法,在工业中已经得到了广泛的应用。决定制品质量的关键因素在于对注射过程中一些关键变量进行有效的检测与控制,以及在此基础上对工艺参数的优化。充模过程中模腔内熔体的前端流动速度与保压过程中模腔压力是与制品质量相关的两项重要过程变量,然而目前在实际生产中始终缺乏对上述变量进行检测的成熟有效方法。本论文基于国家科技支撑计划项目“精密塑料注射成型装备电伺服液压传动及控制技术的研究”(项目编号：NO.2007BAF13B04)和国家自然科学基金项目“注射成形过程中聚合物分层结晶行为的超声在线测量”(项目编号：NO.51105334)的主要研究内容,并结合模腔内熔体信息无损检测方法的预研规划而提出,旨在通过本课题的研究为精密注射成型过程中模腔内熔体流动速度与压力的无损检测与控制技术产业化应用提供关键技术解决方案。论文对注射成型过程进行了详细的介绍,并对过程中涉及到的多项变量进行分析、归类,进而阐述了本文的研究对象与制品质量间的关系,实现了注射系统的数学建模,建立了过程变量与设备变量之间的联系,讨论了优化工艺参数对注塑制品质量的重要影响,并通过Moldflow仿真进行验证,为实验研究打下了坚实的理论基础。针对充模过程中模腔内熔体前端速度的无损检测,引入了高斯过程软测量方法。该方法基于统计分析理论,将充模过程中的注射单元作为待辨识系统,利用高速摄像机及可视化模腔系统所采集到的模腔内熔体前端流动速度作为训练数据集的主导变量,将注射成型过程变量检测平台所采集到的螺杆推进速度、油缸压力以及模腔截面积作为辅助变量,对辨识模型进行训练与优化,进而将新的辅助变量作为优化后的模型输入,并将模型输出与实际值相比较。不同工况下的多组试验结果表明了此方法的有效性。为了将此方法在线应用到注射成型过程中,本文设计开发了相应的注射过程运动控制器,将完成离线辨识与优化的软测量模型写入徽处理器,实现了熔体前端速度的在线检测与控制。实验结果证明,应用该方法注射成型的制品密度分布更加均匀,机械性能得到明显提高。针对保压过程中模腔压力的检测,提出了基于超声回波信号的模腔压力高斯软测量方法。此方法首先根据塑料的PVT特性建立起熔体密度与压力的定性关系,并基于注射过程中熔体密度变化引起超声回波信号幅值变化这一特性,以模腔压力为主导变量,以超声回波信号幅值、模具温度以及油缸压力为辅助变量建立了高斯软测量模型。通过模具内的嵌入式模腔压力传感器获取压力信号,以及注射过程变量检测平台所采集到的各辅助变量值作为训练数据集对模型进行训练。并最终在不同保压工艺参数、不同检测位置以及不同注射材料等多种工况下进行实验,实验结果证明在材料相同时软测量模型预测输出值可以很好的与实际输出值相吻合。此方法为模腔压力的在线超声检测与工艺参数优化后续研究奠定了坚实的基础,并将对制品质量的提高起到至关重要的作用。实现关键过程变量的检测给注射工艺的直接优化提供了基础,然而由于注射成型是典型的周期过程,批量生产中注塑制品的重复精度是另一项重要的评价要素,因此本论文在对注射过程进行深入理论分析、大量实际操作的基础上针对伺服驱动型注塑机非线性、强干扰以及大滞后的问题,将具有强鲁棒性的模糊控制方法与可以快速消除静差的传统PI控制方法结合,构成了模糊PI控制策略。为了检验新型控制算法在注射过程中的控制性能,本文同时实现了传统的积分分离PID控制。实验结果表明,模糊PI控制算法在注射成型中的保压压力控制过程中可以消除超调,抑制静差,缩短滞后时间,提高制品重量的重复精度,各个方面的试验结果都优于传统的积分分离式PID控制。更多还原

【Abstract】 Injection molding is the most widely used methods in manufacturing precise and complicated plastic products for industrial, commercial and medical demands. Detection and control in injection molding process have become one of the most active research areas for their significant influences to the product quality. The flow front velocity and mold cavity pressure play key role in the injection process. However, the NDT methods for these variables pressure are not fully developed. This research is supported by the National Key technology R&D Program (NO.2007BAF13B04) and Project (NO.51105334) supported by National Natural Science Foundation of China under the Ministry of Science and Technology of the People’s Republic of China. This thesis provides solution of in-mold detection and control technology.In order to state the relation between the research objectives and product quality, this study introduced the injection molding method, analysed and classified the multi variables during the process, realized the mathematical model, establish the relations between process variables and machine variables, indicated that the process parameters are important to product quality, and laid the theoretical foundation for the experiments.Based on the Gaussian process (GP) soft-sensor, this study provides a NDT method for the in-mold flow front velocity. The injection unit was set as identification system. The flow velocity was measured by visualization cavity and high speed camera, while the screw speed, cylinder and cavity aera were acquired by other sensors. Then the soft-sensor model was optimized by those training data. Experimental data show that the soft-sensor is able to predict the flow front velocity effectively. In order to apply this optimized soft-sensor in the injection molding process online, a GP predictor is implemented into the controller. Results show that the products manufactured with this method have more uniform distribution of density and the mechanical properties have been improved.In terms of mold cavity pressure detection, this study presentes an ultrasonic-based Gaussian process soft-sensor. The ultrasonic waves are applied in injection molding process according to the PVT characteristics of the plastics. The correlation between the reflection signal and the mold cavity pressure is highly nonlinear. Being trained by the regression model, a Bayes-based GP soft-sensor is used to predict the mold cavity pressure. Experimental data from different process parameters showed that such a developed ultrasonic-based GP soft-sensor is able to predict the mold cavity pressure well during injection molding process. For comparisons, this paper also realizes the neural network (NN) soft-sensor. Results showed that GP soft-sensor has higher precision and better flexibility. Such a developed technology will provide helpful reference for the design of the NDT method of mold cavity pressure in injection molding process.Non destructive testing of flow front velocity and mold cavity pressure will optimize the process parameters efficiently, which provides a foundation for improving the product quality. However, injection molding is a periodic process, and the repeatability precision is also an important factor. Therefore, this study also realizes a fuzzy PI (FPI) controller for flow front velocity and packing pressure control. The novel controller is designed for solving the problems of large overshoot, static error, and long delay time on servo motor-driven injection molding machines. According to the nonlinear, severe interferences, and long delay time characteristics in injection process, this paper integrates predictive grey system, robust fuzzy ratiocination, and PI control. Using these mathematical model and control algorithms, a FPI controller is implemented into an MCU using C programming techniques. The integral discrete PID controller and solid fuzzy controller were realized for contrast experiments. The experimental results show that FPI controller had better performance on reducing overshoot and static error, increasing both response rate and repeatable accuracy. Such a developed technology would provide helpful references for designing the controller of energy-saving servo motor-driven injection molding equipment.更多还原