【摘要】 特征提取是机械故障诊断中最关键的技术之一，本文以机械设备中齿轮、轴承等主要零件为对象，以多种故障的对比实验为基础，用22种不同的时、频域特征提取算法对实验中采集的加速度信号进行分析，比较了提取的各特征量对不同工况和故障状态的敏感度，为进一步的诊断决策奠定了基础。更多还原

【Abstract】 Feature extraction is one of the most important techniques in the fault disgnosis of mechanical systems. For the cases of gears and bearings, based on a variety of contrastive experirnents, 22 methods of feature extraction in the time and frequency domain are adopted in the paper, where acceleration signals collected in the experiment are analyzed. Difference in susceptibility of the features under various situations and fault conditions is compared. They are taken as the basis of decision in fault diagnosis.更多还原