Research Article
Time-Frequency Analysis of Vibration Signal Distribution of Rotating Machinery Based on Machine Learning and EMD Decomposition
@INPROCEEDINGS{10.1007/978-3-030-94182-6_9, author={Xiao-zheng Wan and Song Zhang and Ji-ming Zhang and Hui Chai and Huan-yu Zhao}, title={Time-Frequency Analysis of Vibration Signal Distribution of Rotating Machinery Based on Machine Learning and EMD Decomposition}, proceedings={IoT and Big Data Technologies for Health Care. Second EAI International Conference, IoTCare 2021, Virtual Event, October 18-19, 2021, Proceedings, Part II}, proceedings_a={IOTCARE PART 2}, year={2022}, month={6}, keywords={Rotating machinery EMD decomposition Machine learning Time-frequency characteristics}, doi={10.1007/978-3-030-94182-6_9} }- Xiao-zheng Wan
Song Zhang
Ji-ming Zhang
Hui Chai
Huan-yu Zhao
Year: 2022
Time-Frequency Analysis of Vibration Signal Distribution of Rotating Machinery Based on Machine Learning and EMD Decomposition
IOTCARE PART 2
Springer
DOI: 10.1007/978-3-030-94182-6_9
Abstract
The conventional time-frequency analysis method of the vibration signal distribution of rotating machinery has a low correlation coefficient between the time-frequency features and the vibration signal, which leads to a low accuracy in identifying fault types of rotating machinery. For this reason, this paper proposes a time-frequency analysis method for the vibration signal distribution of rotating machinery based on machine learning and EMD decomposition. After collecting the vibration signal of the rotating machinery, it is decomposed by EMD, the collected signal is decomposed into a set of steady-state and linear data series, and the decomposed signal is filtered and denoised. Then set the time window to extract the local time domain feature components of the vibration signal, and then use the artificial neural network to identify the fault type of the rotating machinery. Experimental results show that this method improves the correlation coefficient between time-frequency characteristics and vibration signals, and improves the accuracy of identifying fault types in rotating machinery.
