摘要详情

ID / 提交时间

290 / 2025-06-16 17:18:28

标题

Adaptive Decomposition Level Determination Method for Fractional High-density Discrete Wavelet Based on Multi-scale Symbolic Dynamic Entropy

关键字

High-density discrete wavelet transform, Fault diagnosis, Multi-scale symbolic dynamic entropy, Adaptive selection, Wavelet transform、

主题及专题

4、System Monitoring, Fault Diagnosis and Prediction

状态

全文待审

作者

昕张 / 福建省特种设备检验研究院

志农李 / 南昌航空大学

苏苏郑 / 南昌航空大学

姗姗黄 / 南昌航空大学

摘要

Local characteristics at any time and frequency are expected to be manifested well done when the non-stationary signal is analyzed. Therefore, based on high-density discrete wavelet transform (HD-DWT) and fractional time-frequency analysis, the fractional high-density discrete wavelet transform (FH-DWT) is constructed in this paper. However, when using the FH-DWT for fault diagnosis, the diagnosis effectiveness is greatly influenced by the decomposition level. Therefore, how to adaptively determine the decomposition levels in FH-DWT has become an urgent problem. Based on the principle that multi-scale symbolic dynamic entropy (MSDE) can measure the similarity between components at different scales, a method for adaptively determining the optimal decomposition level is proposed in this paper. In the proposed method, the MSDE value in each layer can be used to measure the similarity between low-frequency and mid-frequency components, and the decomposition level can be adaptively determined. The results of simulation show that the FH-DWT based on MSDE is superior to traditional HD-DWT, and can adaptively determine the optimal decomposition level in FH-DWT. Finally, the effectiveness of proposed method is verified by two experimental cases further.