Font Size: a A A

Research On Fault Diag Nosis For High-Voltage Electrical Machine And Maintenance Technology Applications In Oilfield

Posted on:2019-01-10Degree:MasterType:Thesis
Country:ChinaCandidate:F X ZhangFull Text:PDF
GTID:2381330572969957Subject:Control engineering
Abstract/Summary:PDF Full Text Request
Daqing Oilfield Limited Company Natural Gas Branch is responsible for the majority of the entire oil field of crude oil stability and natural gas purification tasks?The number of 300 to 7200 kilowatts high-voltage motor more than 90 units,These high-voltage motor's load are compressor,The compressor is a key equipment for natural gas processing equipment,and long-term work,So the operation of high-voltage motor is directly related to the natural gas processing device product yield and natural gas processing capacity?The number of High-voltage motor running more than 15 years has reached more than 20 units?Frequently damaged high-voltage motor is affecting 28 sets of oil and gas treatment device maintenance plan arrangements and natural gas distribution?So fault diagnosis for High-Voltage electrical machine is an important job to ensure the safety of equipment,It provides a reliable basis for the development of reasonable and effective spare parts and maintenance plans?The main contents of the paper are as follows:Firstly,the fault mechanism of the asynchronous motor is studied from the stator current and vibration signal of the motor,and the characteristic frequencies under four fault conditions are found,which lays a foundation for fault diagnosis.Secondly the development trend of fault diagnosis technology for electric motors at home and abroad is studied.It is found that fault diagnosis has gradually transitioned from the mathematical model and signal analysis stage to the artificial intelligence stage;the artificial intelligence based fault diagnosis method is gradually transitioned from SVM?KNN?BPNN?Decision Tree?Bayes Classifiers,etc.to deep learning.The third is to study the characteristics and research progress of four neural network-based deep learning methods,expound the advantages of CNN,and design a CNN for the fault diagnosis of the key components of the motor(rolling bearings),successfully diagnosed the fault type and fault degree of the motor,the method have high precision and fast convergence speed,the application works well.Fourthly,the fault diagnosis of the high-voltage motor was successfully carried out in HongYa Associated Natural Gas Processing Plant of Daqing Oilfield Natural Gas Company.At the same time,the field experience of the high-voltage motor maintenance technology in the oilfield was summarized in the past years.In the next step,we will continue to carry out research and discussion on the fault diagnosis method for high-voltage motors,and strive to improve the management level of high-voltage motors in the natural gas branch,and provide technical support for the continued stable production of Daqing Oilfield.
Keywords/Search Tags:oilfield, high-voltage electrical machine, Deep Learning, CNN, fault diagnosis, maintenance
PDF Full Text Request
Related items