Research On Fault Diagnosis Technology Of Rotating Equipment Based On Digital Twin

In petroleum production,such as centrifugal pumps,centrifugal compressors and other rotating equipment account for a high proportion.They have the characteristics of complex structure and uncertain relationship among components.In addition,the equipment will fail because of performance degradation during long-term operation at high speed and heavy load.Once a failure occurs,it will inevitably cause additional economic losses due to shutdown of production and maintenance,and cause casualties in serious cases.Fault diagnosis and trend prediction technology can provide guidance for the maintenance and health management of rotating equipment,so as to ensure the safe operation of equipment,reduce management costs and improve operation efficiency.As a new concept based on information technology,digital twin provides a new development idea for industrial design and production,but the research results of equipment health management oriented to digital twin are very scarce.In this paper,a framework for fault diagnosis and trend prediction of rotating system driven by digital twin is proposed innovatively.A real-time mapping method of digital twin model is proposed to solve the problem of real-time updating of model in digital twin technology.Finally,the quantitative fault diagnosis of rotating system is realized by using the mapping model.The specific research contents are as follows:(1)Traditional methods of fault diagnosis and prediction of equipment are mostly based on signal processing or expert system.Although these methods can ensure the accuracy of fault qualitative identification,there are some problems such as unclear fault mechanism analysis,difficulty in fault quantitative and location analysis.In this paper,a framework for fault diagnosis and trend prediction of equipment driven by digital twin is proposed.It innovatively introduces the concept of digital twin into fault diagnosis and trend prediction of equipment,including key technologies,components and implementation process.The framework provides a paradigm for the new generation of equipment health management based on information technology.It deeply integrates the physical mechanism-driven and data-driven methods of equipment health management.It greatly improves the understanding of equipment failure mechanism of managers and the accuracy of existing fault diagnosis and prediction methods.(2)At present,the research of digital twin technology is mostly at the stage of conceptual and framework research,and the theoretical research on real-time mapping of digital twin technology is very few.In this paper,a real-time mapping method of digital twin model based on response surface is proposed.This method can update the geometric model,static parameters and dynamic parameters of digital twin model in real time.The experimental results of real-time digital twin mapping for typical rotating systems show that the updated twin model can meet the requirements of real-time digital twin mapping.(3)Traditional fault diagnosis methods based on signal processing can only qualitatively identify the fault types of equipment,and can not achieve quantitative and location analysis of the fault,which will reduce the awareness of field equipment managers about the operation status of equipment.For this reason,taking unbalanced rotor fault as an example,this paper proposes an accurate diagnosis method based on twin model for unbalanced fault of rotating system.It takes the updated twin model as the object,parameterizes the system fault,and estimates unbalanced quantity and unbalanced fault by fitting the measured signal and simulation signal.The experimental results show that this method can accurately realize the quantitative diagnosis of unbalanced faults in rotating system.