针对传统旋转机械单通道故障诊断的信息不完整以及缺少故障样本等问题,提出了基于全信息小波包和支持向量机的旋转机械故障诊断方法。运用小波包频道能量分解技术提取了全信息能量特征向量,以此作为支持向量机多故障分类器的故障样本,经训练的分类器作为故障智能分类器可对设备工作状态进行自动识别和诊断。实验研究表明:基于全信息小波包和支持向量机的故障诊断方法能准确、有效地对旋转机械的工作状态和故障类型进行分类,显著提高了故障诊断的准确率。 Aiming at the problems of incomplete information and missing samples in traditional single-channel fault diagnosis of rotating machinery, a rotating machinery fault diagnosis method based on full information wavelet packet and support vector machine is proposed. The whole information energy eigenvector is extracted by wavelet packet energy decomposition, which can be used as a fault sample of SVM multi-fault classifier. The trained classifier can be used as fault intelligent classifier to automatically recognize and diagnose the working status of equipment. Experimental results show that the fault diagnosis method based on all-information wavelet packet and support vector machine can classify the working status and fault types of rotating machinery accurately and effectively, which improves the accuracy of fault diagnosis remarkably.