分级组合式神经网络诊断系统研究  

Study on the Grading and Connecting Neural Network System of Fault Diagnosis of Large Rotating Machinery

作　　者：张小栋[1] 

机构地区：[1]西安交通大学机械工程学院

出　　处：《振动．测试与诊断》1999年第2期111-117,共7页Journal of Vibration,Measurement & Diagnosis

摘　　要：简要介绍了一个新的机械故障诊断模型——模糊神经网络诊断模型。将之应用于大型旋转机械故障诊断中，基于分级策略，创立了分级组合式神经网络诊断系统，并对其主要功能做了较详尽的论述。通过理论和应用分析，指出了这个新系统的优越性。This paper briefly introduces a newly developed diagnostic system based on fuzzy neural network. The system is graded to several diagnostic subsystems according to different vibration signature on which the fault diagnosis in each subsystem is relied. The connection of these subsystems to other supplementary ones forms an organic whole, which can be applied to the fault diagnosis of large rotating machinery. Theory and application analysis show some advantages of the system.

关 键 词：故障诊断 神经网络 转子 人工智能 机械故障 

分 类 号：TH17[机械工程—机械制造及自动化] TP18[自动化与计算机技术—控制理论与控制工程]