A review: data driven-based fault diagnosis and RUL prediction of petroleum machinery and equipment
Daan Ji, Chuang Wang, Jiahui Li, Hongli Dong
Abstract
In this paper, an up-to-date overview is provided on the data driven-based fault diagnosis (FD) and remaining useful life (RUL) prediction problems of the petroleum machinery and equipment (PME). First, the FD and RUL prediction of five key components including bearings, gears, motors, pumps and pipelines are discussed by adopting mathematical statistics and shallow learning. Then, four kinds of widely-used DL models, i.e. deep neural networks, deep belief networks, convolution neural networks and recurrent neural networks, are surveyed, and the applications in the field of PME are highlighted. Finally, the possible challenges are proposed and some corresponding research directions in the future are presented.