基于随机共振和经验模态分解的水力发电机组振动故障诊断

针对实际水力发电机组故障诊断中微弱信号难以检测引起故障诊断准确率低的难题,提出了一种基于随机共振(stochastic resonance,SR)和经验模态分解(Empirical Mode Decomposition,EMD)的微弱信号检测方法。首先,采用随机共振对振动信号进行降噪处理,提高信号的信噪比;继而对随机共振的双稳输出信号进行EMD分解,并采用能量法进行故障特征向量的提取,最后将其作为基于遗传算法优化支持向量机(GA-SVM)故障诊断模型的输入,实现故障模式的识别与诊断。仿真结果表明,该方法能够准确识别机组的异常情况,具有较高的故障诊断精度。

Vibration fault diagnosis of hydroelectric generating unit by using stochastic resonance and Empirical Mode Decomposition

Aiming at the low accuracy problems caused by the difficulty of weak signals detection in fault diagnosis for actual hydroelectric generating unit, this paper presents a weak signal detection method based on stochastic resonance (SR) and Empirical Mode Decomposition (EMD). This method first reduces noise signal of a vibration signal using stochastic resonance to enhance its stochastic resonance, then uses EMD to decompose its output signal and energy method to extract its feature vectors. Taking the feature vectors as input, a genetic algorithm optimization and support vector machine model is able to achieve identification and diagnosis of the signal faults. The simulation results show that this method can accurately identify the unit''s abnormal situation with high accuracy in fault diagnosis.