| 复杂产品多故障诊断中的核模糊聚类方法 |
| |
| 引用本文： | 李天恩,何桢.复杂产品多故障诊断中的核模糊聚类方法[J].系统工程理论与实践,2013,33(1):181-186. |
| |
| 作者姓名： | 李天恩 何桢 |
| |
| 作者单位： | 1. 天津大学 管理经济学部, 天津 300072;
2. 中国航天标准化与产品保证研究院, 北京 100071 |
| |
| 基金项目： | 国家自然科学基金(70931004) |
| |
| 摘 要： | 多故障作为标准单故障的组合, 很多文献对多故障的诊断都提出行之有效的解决策略, 但却忽视单故障模式之间的相互关系, 而影响到多故障诊断效率, 尤其对于故障繁多的复杂产品. 为了克服该缺点, 引入KFCM-F算法和核化聚类有效性指标KVK, 提出两阶段聚类框架, 数据仿真试验证明该框架能有效发现单故障之间的潜在关系, 从而达到压缩故障模式以期提高诊断效率的目的.
|
| 关 键 词： | 复杂产品 多故障诊断 核模糊聚类算法 核化聚类有效性指标 |
| 收稿时间： | 2010-09-21 |
Kernel-based fuzzy clustering algorithm for multi-faults diagnosis of complex products |
| |
| LI Tian-en,HE Zhen.Kernel-based fuzzy clustering algorithm for multi-faults diagnosis of complex products[J].Systems Engineering —Theory & Practice,2013,33(1):181-186. |
| |
| Authors: | LI Tian-en HE Zhen |
| |
| Institution: | 1. School of Management, Tianjin University, Tianjin 300072, China;
2. China Academy Aerospaces Standardization and Product Assurance, Beijing 100071, China |
| |
| Abstract: | Many literatures proposed effective solutions to multi-faults diagnosis as the combination of standard single faults. However, ignoring interrelation among single faults is to impair its efficiency, especially for complex products. To overcome these disadvantages, two-stage clustering frame was proposed by using KFCM-F algorithm and kernel-based cluster validity index KVK. The simulation results validate the effectiveness of this frame to find latent interrelation among single faults and reduce the number of fault pattern for improving diagnostic efficiency. |
| |
| Keywords: | complex products multi-faults diagnosis kernel-based fuzzy clustering algorithm kernel-based cluster validity index |
|
| 点击此处可从《系统工程理论与实践》浏览原始摘要信息 |
| 点击此处可从《系统工程理论与实践》下载免费的PDF全文 |
|
