AUTHOR=Cao Yu , Cheng Xu , Zhang Qiong 

TITLE=An improved method for fault diagnosis of rolling bearings of power generation equipment in a smart microgrid

JOURNAL=Frontiers in Energy Research

VOLUME=Volume 10 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/energy-research/articles/10.3389/fenrg.2022.1006215

DOI=10.3389/fenrg.2022.1006215

ISSN=2296-598X

ABSTRACT=In the construction of smart microgrid in petrochemical enterprises, generating unit is an import part. Rolling bearing is a key component of generator, and its health directly affects the safe operation of the whole generating unit, the correct fault diagnosis of it not only can improve the stability of smart microgrid, but also can reduce the loss of the factory. This paper propose an improved fault diagnosis method based on VMD and CNN. VMD is used to remove random noise in the original signal and CNN is used to extract feature from the vibration signal processed by VMD. Since the modal number and penalty parameter of VMD is hard to choose and have a profound impact on the decomposition results, DE is used to be the optimization method and envelope entropy is used to be the fitness function to optimize the parameters of VMD. As it is difficult to ensure the most fitness hyper parameters of the CNN, this paper proposes a method by using the DE algorithm to get the suitable hyper parameters for the CNN, and then use the CNN to diagnose fault. The results of the bearing vibration data from the Case Western Reserve University (CWRU) show that the combination of VMD and CNN can improve the convergence speed more than 10% and the accuracy to over 99.6%.