基于变分模态分解和PSO-SVM的起重机齿轮箱故障诊断

起重机齿轮箱的振动信号具有信噪比低、非线性的特点,需要一定的专业知识和经验才能实现故障诊断。为了实现起重机齿轮箱的智能故障诊断,提出了一种基于变分模态分解(Variation?al modal decomposition,VMD)改进小波降噪和粒子群算法(Particle swarm optimization,PSO)优化支持向量机(Support vector machine,SVM)的智能故障诊断方法。首先,利用VMD将振动信号分解,得到不同尺度的本征模态函数(Intrinsic mode function,IMF),将分解的高频分量进行改进小波降噪后和低频分量完成信号重构;然后,提取重构信号的特征参数构建特征向量,使用核主分量分析(Ker?nel principal component analysis,KPCA)对向量降维处理实现特征信息融合;最后,利用PSO优化后的SVM进行故障识别分类。实验验证表明,基于VMD改进小波信号预处理和PSO算法优化SVM的模型具有很高的识别准确率,能够有效、准确地对起重机齿轮箱的故障类型进行识别和分类。

Fault Diagnosis of Crane Gearbox based on Variational Mode Decomposition and PSO-SVM

The vibration signal of crane gearbox has the characteristics of low signal-to-noise ratio and nonlinearity,so it needs some professional knowledge and experience to realize fault diagnosis.In order to real?ize intelligent fault diagnosis of crane gearbox,an intelligent fault diagnosis method based on variational modal decomposition(VMD)improved wavelet denoising and particle swarm optimization(PSO)support vector ma?chine(SVM)is proposed.Firstly,VMD is used to decompose the vibration signal to obtain the intrinsic mode function(IMF)of different scales.The decomposed high frequency component is improved after wavelet denoising and the low frequency component is reconstructed.Then the feature parameters of reconstructed signal are extracted to construct the feature vector,and kernel principal component analysis(KPCA)is used to real?ize the feature information fusion.Finally,the PSO optimized SVM is used for fault identification and classifica?tion.The experimental results show that the SVM model based on VMD improved wavelet signal preprocessing and PSO algorithm has high recognition accuracy and can effectively and accurately identify and classify the fault types of the crane gearbox.