| Title | Multiple Fault Diagnosis for Sucker Rod Pumping Systems Based on Matter Element Analysis with F-statistics |
| Publication Type | Conference Paper |
| Year of Publication | 2019 |
| Authors | Han, Ying, Li, Kun, Ge, Fawei |
| Conference Name | 2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS) |
| Keywords | cyber physical systems, down-hole working conditions, dynamometer cards, dynamometers, F-statistics, fault diagnosis, fault diagnosis problem, faults classification, fuzzy systems, human factors, Knowledge discovery, matter element analysis, matter-element analysis, Metrics, multiple fault diagnosis, Multiple Faults, oil technology, oilfield production, Petroleum, process control, pubcrawl, pumps, Resiliency, rods (structures), Statistics, sucker rod pumping systems, sucker rod pumping wells |
| Abstract | Dynamometer cards can reflect different down-hole working conditions of sucker rod pumping wells. It has great significances to realize multiple fault diagnosis for actual oilfield production. In this paper, the extension theory is used to build a matter-element model to describe the fault diagnosis problem of the sucker rod pumping wells. The correlation function is used to calculate the correlation degree between the diagnostic fault and many standard fault types. The diagnosed sample and many possible fault types are divided into different combinations according to the correlation degree; the F-statistics of each combination is calculated and the "unbiased transformation" is used to find the mean of interval vectors. Larger F-statistics means greater differences within the faults classification; and the minimum F-statistics reflects the real multiple fault types. Case study shows the effectiveness of the proposed method. |
| DOI | 10.1109/DDCLS.2019.8908910 |
| Citation Key | han_multiple_2019 |
Multiple Fault Diagnosis for Sucker Rod Pumping Systems Based on Matter Element Analysis with F-statistics