Apparatus, method and computer program product for screw configuration inference
AssigneeThe Japan Steel Works, Ltd.
InventorsHideki Tomiyama, Yohei Fukuzawa, Takahide Takeuchi
Life of this patent
Estimating the term…
Abstract
The apparatus of the invention includes: an acquisition unit 101 that acquires input information including resin properties and acquires, as output information, a specified value of a physical quantity related to a kneaded resin or a kneading machine; a storage unit 16 that stores a knowledge file 162 containing correlations among the input information, the physical quantity and a plurality of screw configurations; and a search unit 106 and a configuration generator unit 107 together that infers, on the basis of the input information and the knowledge file 162, such a screw configuration that is able to meet the specified value.
Claims8
A screw configuration inference apparatus, comprising:
- a central processing unit (CPU) configured to execute executable units stored on a memory, the executable units comprising:
- an information acquisition unit that acquires input information including resin properties and acquires, as output information, a specified value of a physical quantity related to a kneaded resin or a kneading machine;
- a storage unit that stores knowledge information including correlations among the input information, the physical quantity, and a plurality of screw configurations;
- and
- a configuration inference unit that infers, based on the input information and the knowledge information, such a screw configuration that is able to meet the specified value;
- and
- a configuration generator unit that retrieves, from the storage unit, screw element geometry data corresponding to the screw configuration and renders a drawing of the screw configuration based on the screw element geometry data,
- wherein the knowledge information reflects correlations among process types, operating conditions, configuration data, and analysis results learned based on learning information including analysis results to be obtained by a numerical analysis carried out based on analysis conditions made of configuration data that include process types including resin properties, operation conditions including extrusion amounts, and configuration data including screw shape.
A screw configuration inference method to be carried out by a screw configuration inference apparatus, comprising the steps of:
- acquiring input information including resin properties;
- acquiring, as output information, a specified value of a physical quantity related to a kneaded resin or a kneading machine;
- storing, in a storage unit, knowledge information including correlations among the input information, the physical quantity, and a plurality of screw configurations, wherein the knowledge information reflects correlations among process types, operating conditions, configuration data, and analysis results learned based on learning information including analysis results to be obtained by a numerical analysis carried out based on analysis conditions made of configuration data that include process types including resin properties, operation conditions including extrusion amounts, and configuration data including screw shape;
- inferring, based on the input information and the knowledge information, a screw configuration that is able to meet the specified value;
- and
- render a drawing of the screw configuration based on the screw element geometry data corresponding to the screw configuration retrieved from the storage unit.
A computer program product for screw configuration inference, comprising computer instructions to cause the computer to function as:
- an information acquisition unit that acquires input information including resin properties and acquires, as output information, a specified value of a physical quantity related to a kneaded resin or a kneading machine;
- a storage unit that stores knowledge information including correlations among the input information, the physical quantity, and a plurality of screw configurations, wherein the knowledge information reflects correlations among process types, operating conditions, configuration data, and analysis results learned based on learning information including analysis results to be obtained by a numerical analysis carried out based on analysis conditions made of configuration data that include process types including resin properties, operation conditions including extrusion amounts, and configuration data including screw shape;
- a configuration inference unit that infers, based on the input information and the knowledge information, a screw configuration that is able to meet the specified value;
- and
- a configuration generator unit that retrieves, from the storage unit, screw element geometry data corresponding to the screw configuration and renders a drawing of the screw configuration based on the screw element geometry data.
Dependencies are inferred from the claim text.