Physical-Model-Based Adaptive Control of Injection Molding Process

基于物理模型的注塑过程自适应控制

• 批准号: 9800349
• 负责人: Kuo Wang
• 金额: --
• 依托单位: Cornell University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1998
• 资助国家: 美国
• 起止时间: 1998-05-01 至 2000-04-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9800349&HistoricalAwards=false
• 关键词: Physical; Model; Based; Adaptive; Control

DMI-9800349 Wang Considerable effort has been made in the past to analyze and optimize the process control in injection molding at the level of machine parameters, process variables or even part quality, such as weight or thickness. Numerous methods have been tried, but results all depend on specific material/mold configuration and on a particular molding machine, i.e., the method is not generic. No definitive correlation between any process variable and the part quality has been established. Accordingly, system disturbances (e.g., uncontrollable material variation, machine-to-machine characteristics, etc.) still cannot be handled adaptively to ensure part quality. The objective of the research is to solve these problems by taking a novel three-pronged approach: (1) Determine the initial machine-setup condition based on rigorous process simulation rather than trial-and-error or ad-hoc experience; (2) Develop and incorporate a physical model based on molding dynamics and incorporating basic material properties in the control loop for part quality to take care of any possible system disturbance; and (3) Develop smart sensors and associated intelligence to monitor and control part quality on-line and adaptively without the need for elaborate and impractical instrumentation (e.g., flush-mounted pressure transducers in the cavity). Successful completion of this project will have great impact on the injection molding industry in terms of ability to control product quality automatically and adaptively.

电话：[phone] 在过去，人们已经做出了相当大的努力来分析和优化注塑成型中的过程控制，这些过程控制在机器参数、过程变量甚至部件质量（例如重量或厚度）的水平上。 已经尝试了许多方法，但结果都取决于特定的材料/模具配置和特定的成型机，即，该方法不是通用的。 任何工艺变量和零件质量之间没有明确的相关性。 因此，系统干扰（例如，不可控的材料变化、机器到机器的特性等）仍然不能自适应地处理以确保部件质量。本研究的目的是通过一种新的三管齐下的方法来解决这些问题：（1）基于严格的工艺模拟而不是试错或临时经验来确定初始机器设置条件：（2）开发并整合基于成型动力学的物理模型，并将基本材料特性纳入零件质量控制回路，以照顾任何可能的系统干扰;以及（3）开发智能传感器和相关的智能以在线和自适应地监测和控制部件质量，而不需要复杂和不切实际的仪器（例如，空腔中的齐平安装的压力传感器）。 该项目的成功完成将对注塑行业自动和自适应控制产品质量的能力产生重大影响。