射出成型製程為一項受環境、設備、材料與時間因子所影響之製造程序。在產品進入量產階段，常因忽略某些因子的變化而產出許多不良品，驟增生產成本。 本研究針對射出成型產品在量產的情況下，利用少量之生產資料，以路徑分析模式發展一針對環境、設備、材料與時間因子與最終品質之多線性關係模式，藉以預測最終產品品質。透過本研究之品質預測模式，製造者能有效決定是否進行投料生產或停機調教機台，如此得以大幅減少製造不良品之成本損失。 Injection molding is a process that is influenced by various factors, such as materials, equipment, environment, and time. Usually injection molding process is assumed stable when the production is in a stage of mass production. Unfortunately, some unseen factor changes may deteriorate the process and result in unwanted defectives. By applying path analysis, this research develops multiple linear equations between the above-mentioned factors and end-product quality using limited on-line data when the production is in the stage of mass production. The results show that the equations can accurately estimate the end-product quality. Based on the predictions by the equations, manufacturers can easily make a “do production” or “stop for checking equipment” decision in production. Due to the accuracy on the equations, defectives can be eliminated.
