基于粗集理论的一种混合智能故障诊断新方法  

New Approach of Integrated Intelligent Fault Diagnosis Based on Rough Set Theory

作　　者：朱张青[1] 周川[1] 胡维礼[1] 

机构地区：[1]南京理工大学自动化系,南京210094

出　　处：《控制与决策》2006年第2期233-235,240,共4页Control and Decision

基　　金：国家自然科学基金项目(60174019);清华大学智能技术与系统国家重点实验室基金项目

摘　　要：采用基于需求的约简方法,将依赖于系统数学模型和依赖于故障特征信息的两类故障诊断方法相融合,提出一种混合智能故障诊断新方法.以非线性系统故障诊断为例,给出了混合智能故障诊断方法提取系统故障诊断规则的步骤.该方法不仅可以扩大故障诊断的范围,而且能紧密联系工程实际.最后以异步电动机常见故障的诊断为具体实例,说明了该方法的可行性和有效性.The shortcoming of the traditional intelligent fault diagnosis based on rough set theory is analyzed. The reduction algorithm meeting users＇ requirements is adopted, and a new approach of integrated intelligent fault diagnosis combining the method relied on the systems＇ model and the strategy depending on fault information is presented. For fault detection on nonlinear systems, the steps of getting the diagnosis rules are given. The approach can expend the fault diagnosis bounds, and is easy to implement in engineering practice. Finally, an example of AC motor fault diagnosis is studied to show the validity of the approach.

关 键 词：故障诊断 粗糙集 混合智能 非线性系统 

分 类 号：TP271[自动化与计算机技术—检测技术与自动化装置]