Back-propagation neural net using Taguchi Quality

被引：0

作者：

Lee, MC ^{[1
]}

Young, MS ^{[1
]}

Hung, B ^{[1
]}

Ma, RP ^{[1
]}

Uang, CM ^{[1
]}

机构：

[1] Natl Chang Kung Univ, Dept Elect Engn, Tainan, Taiwan

来源：

PHOTOREFRACTIVE FIBER AND CRYSTAL DEVICES: MATERIALS, OPTICAL PROPERTIES, AND APPLICATIONS V | 1999年 / 3801卷

关键词：

Taguchi; type 1 structure identification; back-propagation; neural network; orthogonal array;

D O I：

10.1117/12.363943

中图分类号：

T [工业技术];

学科分类号：

08 ;

摘要：

In this paper, the Taguchi Quality Engineering Method (TQEM) is applied to choose the required flight parameters in a back propagation neural algorithm for the analysis of aircraft wing loading. The screening out unwanted parameters is an important procedure in strain measurement, as using too many flight parameters increases the complexity of the network structure and also the difficulty of the learning scheme. The demonstrated results show that TQEM is easier and more reliable than traditional methods for choosing the required parameters for system identification.

引用

页码：107 / 113

页数：7