基于变分模态分解和高阶统计量的梯级故障诊断研究

自动扶梯是地铁车站内必不可少的大型公共交通设备,一旦发生故障,小则影响运营,大则引发安全事故;梯级作为自动扶梯的重要结构部位,其固定螺栓松动必然会导致自动扶梯运行异常;针对梯级振动信号故障特征难以提取的问题,提出了变分模态分解(VMD)和高阶统计量(HOS)联合来对自动扶梯故障特征提取;该方法首先对原始振动信号进行VMD分解,得到K个固有模态分量(IMF);然后对主IMF分量进行奇异值分解(SVD)降噪,对去噪后的主IMF分量进行重构得到新的信号;最后通过高阶统计量对新的信号故障特征提取,并利用随机森林分类算法对三类不同的振动信号样本进行分类识别,确定梯级振动故障类型;实验结果表明,该方法可以有效地提取故障特征,实现故障诊断与分类。

Research on Cascade Fault Diagnosis Based on Variational Mode Decomposition and Higher Order Statistics

The escalator is an essential large-scale public transportation equipment in a subway station. Once a failure occurs, the operation will be affected if the escalator is small, and it will cause a safety accident. As an important structural part of the escalator, the loosening of its fixing bolts will inevitably lead to abnormal operation of the escalator. Aiming at the problem that it is difficult to extract the fault characteristics of cascade vibration signals, this paper proposes a method of combining variational modal decomposition (VMD) and higher order statistics (HOS) to extract escalator fault characteristics. This method first performs VMD decomposition on the original vibration signal to obtain K intrinsic modal components (IMF); then performs singular value decomposition (SVD) noise reduction on the main IMF component, and reconstructs the denoised main IMF component to obtain a new Finally, the new signal fault characteristics are extracted through high-order statistics, and the random forest classification algorithm is used to classify and identify three different vibration signal samples to determine the type of cascade vibration fault. Experimental results show that this method can effectively extract fault features and realize fault diagnosis and classification.