基于Bayes估计相空间融合和CM-SVDD的有载分接开关机械故障诊断

A Mechanical Fault Diagnosis of On-load Tap-changers Based on Phase Space Fusion of Bayes Estimation and CM-SVDD

为进一步提高变压器有载分接开关(on-loadtap changer,OLTC)机械故障的识别精度,从OLTC振动信号的混沌动力学特性出发,提出基于Bayes估计相空间融合与改进支持向量数据描述(support vector data description,SVDD)的OLTC机械故障诊断方法。首先,利用多变量相空间重构(multivariate phase space reconstruction,MPSR)对OLTC切换时的多通道振动信号进行高维重构,然后基于Bayes估计与奇异值分解提取高维融合相空间相点的特征指标,据此建立以轮廓域图(contourmap,CM)为决策判据的改进SVDD识别模型,用于OLTC机械故障的有效诊断。对某OLTC模型正常与典型机械故障下切换时振动信号的分析结果表明:基于Bayes估计的OLTC振动信号高维相空间融合方法能获取更为完整的OLTC机械状态信息。相对于现有故障诊断方法,所提出的CM-SVDD诊断模型具有更优的识别精度和较高的计算效率。研究成果可为基于振动分析法的OLTC机械状态诊断技术提供重要依据。

To further increase the identification accuracy of mechanical faults of on-load tap changers(OLTCs),this paper presented a mechanical fault diagnosis model of OLTC based on the combination of phase space fusion of Bayes estimation and the improved support vector data description(SVDD)considering the chaos dynamic features of vibration signals of OLTC.Firstly,the multi-channel vibration signals of OLTC were rebuilt in the high dimension space by the multivariate phase space reconstruction(MPSR)method.And then the Bayes estimation and singular value decomposition(SVD)were applied to extract the features of vibration signals of OLTC in the fused high dimension phase space.Finally,the improved SVDD model was obtained to effectively recognize the mechanical faults of OLTC,where the contour map(CM)was selected as the decision criterion.The calculated results of vibration signals of an OLTC model under normal and fault conditions show that the high dimension phase space fusion based on Bayes estimation is capable of fully describing the mechanical condition information of OLTC.Compared with the existed diagnosis method,the proposed diagnosis model of CM-SVDD has high diagnosis accuracy and calculation efficiency.The results could provide the important reference for the mechanical condition diagnosis of OLTC by vibration analysis methods.