DOIONLINE 
DOIONLINE Publish In |
International Journal of Advance Computational Engineering and Networking (IJACEN)-IJACEN |
 Journal Home Volume Issue |
||||||||
Issue |
Volume-3, Issue-4 ( Apr, 2015 ) | |||||||||
Paper Title |
A New Robust Design Modeling Method Based On Generalized Regression Neural Network | |||||||||
Author Name |
Tuan-Ho Le, Sangmunshin, Hwa-Il Kim, Sittichai Kaewkuekool | |||||||||
Affilition |
Department of Industrial & Management Systems Engineering, Dong-A University, Busan, 604-714, Republic of Korea 3 Department of Industrial Health, Catholic University of Pusan, Busan 609-757, Republic of Korea 4 Department of Production Technology Education, King Mongkut's University of Technology Thonburi, Bangkok 10140, Thailand | |||||||||
Pages |
19-23 | |||||||||
Abstract |
Over the recent two decades, robust design has been popularly utilized as a powerful method to improve the quality of products in the offline stage of manufacturing processes. Most estimation methods to define the functional relationships between the input factors and output responses are based on the response surface methodology (RSM) which requires several assumptions. Unfortunately, these assumptions are not always hold in the practical industrial problems. Based on the nature of artificial neural networks (ANNs), these relationships can be conducted without any assumption. Therefore, the primarymotivation of this paper is to propose the ANNs as an alternative modeling method in robust design. First of all, an ANN framework-based robust design modeling method is proposed. Secondly, the generalized regression neural network-based modeling method is proposed to estimatethefunctions between the input and output variables.Finally, a comparative study between the proposed neural network-based estimation method and the conventional least squares method based on RSM is conducted in the numerical example. The final results show the efficiency of the proposed neural network-based modeling method in robust design. Keywords- Robust Design, Response Surface Methodology, Generalized Regressionneural Network, Estimation | |||||||||
| View Paper | ||||||||||
