基于小波和BP神经网络的风力机齿轮箱故障诊断

摘要

Wind turbine gearbox vibration signal is a non-stationary signal, its time-domain characteristics are very complex. The use of general time-domain analysis or frequency domain analysis method is difficult to analyze the specific fault. Based on the research status at home and abroad, a new method to diagnose the vibration of the wind turbine gearbox using wavelet analysis and neural network is proposed. The use of LabVIEW and Matlab software solves the wavelet packet band energy, and the use of time frequency technology filters the vibration signal of wind turbine failure. The change of frequency band energy is taken as the fault characteristic value, so the intelligent fault diagnosis of fan gear box is realized. From the simulation results, this method can effectively diagnose the fault of the gear box.%风力机齿轮箱振动信号是一种非平稳信号,其时域特性非常复杂,利用一般的时域分析法或者频域分析法分析出其具体故障所在较为困难.基于国内外研究现状,结合小波变换和神经网络方法,提出了一种基于小波变换和BP神经网络结合的风力机齿轮箱故障诊断方法.运用LabVIEW和Matlab软件的小波包求解频带能量,并利用时频分析技术对风力机故障的振动信号进行滤波.将频带能量的变化情况作为风力机齿轮箱故障特征值,进而对其进行智能故障诊断,随后应用神经网络方法进行故障识别.从仿真结果来看,此方法可以有效的诊断齿轮箱故障情况.