基于ANFIS和铁谱磨粒分析的泥浆泵故障诊断  

Fault Diagnose of Mud Pump Based on ANFIS and Ferrography Wear Particles Analysis

作　　者：姜雪丽[1] 裴峻峰[1] 卢群辉[2] 朱恩全[2] 

机构地区：[1]江苏工业学院,江苏常州213016 [2]江苏油田,江苏扬州225009

出　　处：《流体机械》2008年第6期46-49,共4页Fluid Machinery

基　　金：江苏省油气储运技术重点实验室开放项目(CY0606);江苏省科技成果转化专项资金项目(BA2005041)

摘　　要：针对润滑油中铁谱磨粒分析所存在的问题,以泥浆泵作为研究对象,将自适应神经网络模糊系统(ANFIS)引用到铁谱磨粒的自动识别中。通过研究铁谱磨粒特征与故障原因的对应关系,建立ANFIS仿真模型,从而实现了铁谱磨粒分析的智能化。该分析模型使铁谱磨粒分析过程突破了人工经验的限制,从而提高了泥浆泵故障诊断的准确率。Use Adaptive Network-based Fuzzy Inference System to accomplish Automatic Recognition for Ferrography Wear Particles in allusion to the problems of ferrography wear particles analysis. Its study object is mud pump. This paper establish an ANFIS model by studying the relations between characteristics of wear particles and failure cause. The model makes FetTography Wear Particles analysis break the restriction of experiences and improve the accuracy of faulty diagnosis of mud pump.

关 键 词：ANFIS 铁谱磨粒分析 泥浆泵 故障诊断 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]