| 摘要 |
PURPOSE:To enhance mechanical characteristics and appearance and eliminate the irregularity of the result, by optimally performing cavity pressure control and mold- opening control for an injection compression molding machine through learning control. CONSTITUTION:In a learning mechanism 2, and error of a mold-opening controlled variable from an optimum mold-opening model (waveform) used as an initial set point is corrected and converged through several trials to determine a virtual set point, which is outputted from the learning mechanism 2. The difference between the virtual set point thus outputted and a mold-opening controlled variable fed back is inputted to a first PD controller 10, then the difference between an output from the controller 10 and an output from a mold-clamping hydraulic servo valve 12 fed back is obtained, and is inputted to a second PD to a second PD controller 11. Based on an output from the second PD controller 11, the servo valve 12 is set into a PD action, whereby a driving mechanism is operated to control the mold-opening quantity. Cavity pressure control is similarly performed by using an optimum cavity pressure model (waveform) as an initial set point. Thus, an optimum molding control can be constantly performed, and a molded product having excellent mechanical characteristics and appearance can be obtained. |
