| 用Voronoi多胞体确定故障诊断分类器的训练集 |
| |
| 引用本文： | 解培中,张志涌.用Voronoi多胞体确定故障诊断分类器的训练集[J].南京邮电大学学报(自然科学版),1998(Z1). |
| |
| 作者姓名： | 解培中 张志涌 |
| |
| 作者单位： | 南京邮电学院电子工程系 |
| |
| 摘 要： | 提出了利用Ｖｏｒｏｎｏｉ多胞体来选择神经网络故障诊断分类器训练集的新方法。该方法由３部分组成，即首项为正的归一化处理（首正归一）、边界样本选择及置信判决。利用该方法不仅可以大大减少训练样本数，提高训练速度，而且通过调节故障类别分割面位置和形状，为提高故障诊断准确率提供了可能的途径。
|
| 关 键 词： | 故障诊断，神经网络，电路 |
Determination of the Training Sets for Fault Diagnosis Classifiers by Using the Voronoi Cells |
| |
| Xie Peizhong,Zhang Zhiyong.Determination of the Training Sets for Fault Diagnosis Classifiers by Using the Voronoi Cells[J].Journal of Nanjing University of Posts and Telecommunications,1998(Z1). |
| |
| Authors: | Xie Peizhong Zhang Zhiyong |
| |
| Institution: | Xie Peizhong Zhang Zhiyong Department of Electronic Engineering,Nanjing Institute of Posts and Telecommunications,210003,Nanjing,PRC |
| |
| Abstract: | Based on Voronoi cells, a new approach to conform the training sets of a neural network is proposed. The approach is composed of three parts: normalization, near bound selection and confidence determination. The approach can be used not only to reduce the number of training samples and increase the training speed drastically, but also to provide a way to improve the accuracy of fault diagnosis by adjusting the position and shape of the splitting planes. |
| |
| Keywords: | Fault diagnosis Neural networks Circuits |
| 本文献已被 CNKI 等数据库收录！ |
