水电机组故障诊断系统信号特征的提取

信号特征提取是水电机组故障诊断系统的关键，水电机组的故障主要是振动故障，而且机组的振动表现出复杂性、多样性以及渐进性。对这样的信号特征提取，普通特征提取方法的准确性就不高。小波分析将信号分解到不同层以及在不同层上将信号分解为不同频段，然后在这些不同层不同频段上提取特定频率的一段信号。因此本利用小波分析的这一特性，根据机组不同振源的特性提取了机组振动的诊断特征向量。

Characteristic Pickup of Hydroelectric Set Fault Diagnose System

The signal characteristic pickup is the key to hydroelectric set fault diagnoses system.The main fault of waterpower set is the vibration fault; further more the vibration of set is complexity, diversity and gradualness. The ordinary ways of characteristic pickup are of low efficiency to such signal characteristic pickup. Wavelet analysis decomposes the signal to the different layers and decomposes the signal to different frequency sects, then pickup the special signal in different layers and different frequency sects.Therefore, using this characteristic of wavelet analysis the diagnose eigenvector of hydroelectric set vibration was pick-uped on base of the nature of different vibaration source.