Application of feature selection methods for defining critical parameters in thermoplastics injection molding
| Autor: | Ole Vidar Lyngstad, Kristian Martinsen, Olga Ogorodnyk, Mats Larsen |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2019 |
| Předmět: |
0209 industrial biotechnology
Manufacturing process business.industry Computer science media_common.quotation_subject Process (computing) Feature selection 02 engineering and technology Molding (process) 010501 environmental sciences 01 natural sciences 020901 industrial engineering & automation General Earth and Planetary Sciences Quality (business) Information gain Process engineering business 0105 earth and related environmental sciences General Environmental Science media_common |
| Zdroj: | Procedia CIRP |
| Popis: | Thermoplastics injection molding is a manufacturing process used for mass-production of plastic parts. The process includes four main stages during which material used goes through complicated thermo-mechanical changes. In order to make the process more controllable and repeatable it is, at first, necessary to understand which parameters are the most important ones. The following paper describes how application of statistical feature selection methods, such as Information gain and ReliefF, allows to identify which injection molding parameters have a greater influence on the final part quality. The article gives short description of the above-mentioned methods and shows what were results of their application on dataset obtained from 160 machine runs, during which 41 machine and process parameters were logged. © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/) |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|