방적
Improving Cotton Spinning Quality Using Fuzzy Sets.
- 출판일1999.11
- 저자
- 서지사항
- 등록일
2016.11.02
- 조회수
218
The Textile Engineering Department at the University of Minho in Portugal conducted experiments in the use of neural and fuzzy systems in cotton spinning. These experiments aimed to improve cotton spinning quality by the application of neuro-fuzzy systems, which were trained by a learning algorithm derived from neural network theory and can be defined as a special three layer feedforward neural network. A neuro-fuzzy systems can be interpreted at all times as a system of fuzzy rules and approximates an n-dimensional function that is partially given by the training data. For applications in cotton spinning, the TEXPERT NEUROFUZZY CLASSER was used with a carded, combed, and open end yarn data set. The data set contained 524 cases distributed into the three yarn classes. Approximately half were used for training and approximately half were used for testing. 5 refs.