我所李鹏老师课题组完成的论文《Sound Based Degradation Status Recognition for Railway Point Machines Based on Soft-Threshold Wavelet Denoising, WPD, and ReliefF》于2024-01在《IEEE Transactions on Instrumentation and Measurement》正式发表。论文摘要如下：
Abstract: Railway point machines (RPMs) are one of the most important equipment for the safe operation of railway systems. Different from the existing current curve-based methods, aiming at degradation status recognition for ZDJ9 RPMs, this article first presents a sound signal-based degradation status recognition method considering the advantages of easy-to-acquire and non-contact. First, soft-threshold wavelet denoising method is utilized for data preprocessing, which is a key step for improving recognition accuracy, performing better than hard-threshold wavelet denoising method. Second, to comprehensively acquire degradation information, a degradation information extraction method combining wavelet packet decomposition (WPD), time- and frequency-domain statistical features is developed, which can realize fast and effective degradation features extraction. Then efficient ReliefF is adopted for feature dimension reduction, which can eliminate lots of redundant feature points. Finally, for pattern recognition issue of small set, support vector machine (SVM) is used for degradation status recognition. The degradation status recognition accuracy of the slide plate of ZDJ9 RPMs reaches 91.67%. The superiority of the presented method is verified by some experiment comparisons. The presented method can provide support for the on-site staff to realize repair according to condition.
引用信息：Yongkui Sun, Yuan Cao, Peng Li, Shuai Su. Sound Based Degradation Status Recognition for Railway Point Machines Based on Soft-Threshold Wavelet Denoising, WPD, and ReliefF. IEEE Transactions on Instrumentation and Measurement, 2024-01. DOI: 10.1109/tim.2023.3334370