Application of Neuro-Fuzzy System and Genetic Algorithms in Maching-parameters Optimization

• Main Authors: CHUANG HSIN YUAN, 莊信源
• Other Authors: LIN CHANG-PIN
• Format: Others
• Language: zh-TW
• Published: 2000
• Online Access: http://ndltd.ncl.edu.tw/handle/49810931046208826087

碩士 === 國立海洋大學 === 機械與輪機工程學系 === 88 === Ball-type end milling cutters are usually used in molding、automotive and aircraft industries in which components with free-form surfaces are manufactured . Surface roughness is the critical quality index of these products . Therefore , choosing how to control these manufacturing processes and getting better understanding on the operation parameters have become the keys on reducing surface roughness . In this thesis we focus on establishing a methodology to solve these problems . Using the data through our milling experiments and existing software , a Neuro-Fuzzy system was built that can be used to predict the system behavior and to find the optimal operation parameters . Genetic Algorithm was chosen in our optimization process due to its simple and multiple-point searching characteristics which helps in controlling the manufacturing process more effectively .