Support apparatus of injection-molding machine

• US 20080102147A1
• Filed: 10/18/2007
• Published: 05/01/2008
• Est. Priority Date: 10/27/2006
• Status: Active Grant

• Current Assignee: Nissei Plastic Industrial Co., Ltd.
• Original Assignee: Nissei Plastic Industrial Co., Ltd.
• Inventors: Iwashita, Eiki, Yamagiwa, Yoshitoshi, Shioiri, Takayoshi
• Granted Patent: US 7,896,636 B2
• Time in Patent Office: Days
• US Class Current: 425/148
• CPC Class Codes: B29C 2945/76006 Pressure
• CPC Class Codes: B29C 2945/76066 Time
• CPC Class Codes: B29C 2945/76083 Position
• CPC Class Codes: B29C 2945/7611 Velocity
• CPC Class Codes: B29C 2945/76381 Injection
• CPC Class Codes: B29C 2945/76384 Holding, dwelling
• CPC Class Codes: B29C 2945/76404 injection-holding
• CPC Class Codes: B29C 2945/76588 shrinkage, dilation, dimensional change, warpage
• CPC Class Codes: B29C 2945/76615 Weight
• CPC Class Codes: B29C 2945/76638 Optical properties
• CPC Class Codes: B29C 2945/76769 Moulded articles
• CPC Class Codes: B29C 2945/76913 Parameter setting
• CPC Class Codes: B29C 2945/76979 Using a neural network
• CPC Class Codes: B29C 45/766 the setting or resetting of moulding conditions, e.g. before starting a cycle
• CPC Class Codes: B29C 45/7686 the ejected articles, e.g. weight control

Abstract

An apparatus for supporting a worker involved in operating an injection-molding machine includes a neural network in which a prediction function is determined based on test molding. The worker enters a portion of a plurality of molding conditions in the form of a fixed value using a first input apparatus, thereby obtaining a relational graph of the remaining portion of the plurality of molding conditions and the predicted quality value in which the quality of a molded article is predicted.

Claims

1. A support apparatus of an injection-molding machine for supporting a worker involved in operating an injection-molding machine, the apparatus comprising:

a neural network that uses as an input item a plurality of molding conditions used when a non-defective article is obtained in test molding, uses as an output item a quality value obtained by measuring the non-defective article, and establishes a prediction function based on the input and output items;

a first input apparatus provided for work of entering fixed values for molding conditions other than selected molding conditions when at least one molding condition has been selected from the plurality of molding conditions;

a computer for acquiring a prediction function determined by the neural network, setting the output items of the prediction function to an unknown number, entering fixed values for a portion of the input items of the prediction function, entering the selected molding conditions for the remaining portion of the input items of the prediction function in the form of variables, and then calculating a predicted quality value, which is a predicted value of the quality value, using such a prediction function;

a graph generator for generating in the form of a graph the relationship between the selected molding conditions and the predicted quality values; and

a display unit for displaying the graph generated by the graph generator.

2. The support apparatus according to claim 1, wherein the support apparatus further comprises a second input apparatus used by a worker to enter a target quality value, which is a target value of the quality, and a graph correction unit for correcting the graph on the basis of the target quality value entered using the second input apparatus.

3. The support apparatus according to claim 1, wherein a plurality of graphs is displayed on the display unit when there is a plurality of the selected molding conditions.

4. The support apparatus according to claim 1, wherein the quality value is the product weight.