摘 要 异步电机以其结构简单，价格低廉被应用于工农业生产的各个领域。当异步电机发生故障之时不仅损害电机本体，而且有可能引发重大生产安全事故。为保证电机及其所驱动负载的安全运行，就需要对电机状态进行监测做出早期的故障诊断。尤其随着“节能减排”政策的大力提倡与实行，基于变频调速的异步电机系统被广泛的应用于工业生产领域。 电机转子断条是异步电机的常见故障之一，目前的研究主要集中于工频电源供电下的异步电机转子断条故障诊断的研究，而针对变频器供电的异步电机转子断条故障诊断的研究还很少。所以本文将对变频调速系统中的异步电机，即异步电机在变频器供电情况下转子断条故障进行诊断研究。 本文将从异步电机在工频电源供电情况下发生转子断条故障入手，分析其故障特征和相应的故障诊断方法。然后探讨将转子断条的常规检测方法定子电流频谱分析的方法（简称MCSA）应用于变频调速系统的可行性。即是否异步电机在变频器供电情况下发生断条时也存在频率为 的分量，并且通过判断频谱中是否存在频率为 的故障特征分量来检测变频调速系统的异步电机转子断条故障。 MCSA故障检测方法应用于变频调速的异步电机转子断条故障的检测之中，将针对频率为 的故障特征分量容易被变频器供电频率分量所淹没及变频器输出波形含有高次谐波分量，不易做出故障诊断的问题。分别提出基于数字低通滤波、自适应滤波与细化的快速傅里叶变换结合的转子断条故障诊断方法和基于数字低通滤波、希尔伯特变换与细化的快速傅里叶变换（简称ZFFT或ZOOM-FFT）结合的转子断条故障诊断方法。 为验证方法的可行性，分别进行仿真与实物的实验验证。仿真通过MATLAB/Simulink建立恒压频比变频调速系统的异步电机转子故障仿真模型进行仿真验证。实验室则搭建了电机故障模拟平台，信号采集系统平台分别进行故障模拟与信号采集工作，对所提两种方法进行实验验证。经过仿真和实验验证，两种方法对特征频率和故障特征分量的提取，达到预期效果。可以将这两种方法应用于基于变频调速的异步电机转子断条故障的诊断中。

ABSTRACT Asynchronous motors are widely used in industrial and agricultural production as the main energy conversion device. When asynchronous motor break down,not only harming motor itslef, may cause serious production safety accidents.It is necessary to monitor the state of the asynchronous motors and to diagnose the asynchronous motors incipientfault for keeping the asynchronous motors and the loads safe operation.Especially with the policy of the “energy conservation and emission reduction” is advocated,the asynchronous motor power by inverter in variable frequency systems are more widely used. Broken-bars fault is one the common faults to asynchronous motors,at present,the study focus on the broken-bar fault diagnostics of asynchronous motors power by power network.But the research in the broken-bar fault diagnostics of asynchronous motors power by inverter in variable frequency drives system is very few still.So this paper will focus on this areas to do some research on detecting broken-bars fault in invert-fed asynchronous motors. This paper will begin to focus on the broken-bar fault diagnostics of asynchronous motors power by power network.throught analyzing the broken-bar fault feature and common methods of fault diagnosis in this case.Then we will discuss the fault diagnosis method of motor current signature analysis（MCSA）,it is common methods of broken-bar fault diagnostics of asynchronous motors.We will try to explore the possibility of the application of this common method in the broken-bar fault diagnostics of asynchronous motors power by inverter in variable frequency drives system.We will find whether or not broken-bars fault frequency exists,once broken rotor bar fault occurred in asynchronous motors power by inverter in the variable frequency systems.Adopting the frequency spectral method to analyze whether or not broken-bars fault frequency exists to detect broken-bars fault in invert-fed asynchronous motors. If application of the fault diagnosis method of MCSA to detect broken-bars fault in invert-fed asynchronous motors,we are going to face a problem.the components of rotor broken-bar fault characteristic frequencies will affected bycomponents of power supply frequency in the stator current spectrum.Sometime components of rotor broken-bar fault characteristic frequencies may be lose.It is hard to detect broken-bars fault.So this paper presents two different novel detection schemes for rotor bar breaking fault.One schemes is Based on digital lowpass filter techniques,self-adaptived filter techniques,ZOOM fast Fourier transform.Two schemes is Based on digital lowpass filter techniques, Hilbert transform, ZOOM fast Fourier transform. We will do the Simulation and experimental verification to validate the feasibility of the two detection schemes.V/F inverter-fed drive systems with rotor bar breaking fault is completed by MATLAB/Simulink.Then we built an experimental motor fault diagnosis platform to simulate moter faults in the laboratory.In order to acquire sample signals,we design a signal acquisition system. And the simulation and experimental results demonstrate that the two detection schemes is effective.We can use the two detection schemes to detect broken-bars fault in invert-fed asynchronous motors.