|
| Anomaly Detection of Natural Gas Dehydration Unit Based on Grey Correlation Analysis of Process Parameters |
| Received:March 05, 2019 Revised:May 25, 2019 |
| View Full Text View/Add Comment Download reader |
| DOI：10.7643/ issn.1672-9242.2019.05.005 |
| KeyWord:anomaly detection grey correlation degree parameter clustering dehydration unit |
| Author | Institution |
| PENG Bo |
1. Chongqing Gas District of Southwest Oil and Gasfield Company-, Chongqing , China |
| ZHANG Bo |
1. Chongqing Gas District of Southwest Oil and Gasfield Company-, Chongqing , China |
| TAN Jian |
1. Chongqing Gas District of Southwest Oil and Gasfield Company-, Chongqing , China |
| TAN Zhi-bin |
2. State Key Laboratory of Mechanical Transmissions, Chongqing University, Chongqing , China |
| LIANG Tian-you |
2. State Key Laboratory of Mechanical Transmissions, Chongqing University, Chongqing , China |
| YIN Ai-jun |
2. State Key Laboratory of Mechanical Transmissions, Chongqing University, Chongqing , China |
|
| Hits: |
| Download times: |
| Abstract: |
| Objective To establish an abnormal detection model of equipment based on grey relational degree to quickly identify abnormal equipment to solve the high cost of equipment maintenance caused by the common maintenance methods of petrochemical equipment, and ensure the reliable operation of equipment. Methods The grey relational degree analysis method was used to calculate the relational degree among the production monitoring parameters of the supercharger station after data cleaning. And a relational degree matrix among the parameters was established by the relational degree calculated to realize clustering of parameters. The abnormal state of the equipment corresponding to the same monitoring parameters in the clustering results was identified by the method of grey relational degree change among the parameters. Results Similar monitoring parameters were highly correlated in most time periods. And the abnormal state of equipment was predicted when the correlation was abnormal. Conclusion Compared with the methods of threshold judgment of monitoring parameters, the prediction model based on grey correlation analysis method has higher prediction accuracy, realizes fast and effective identification of abnormal equipment. It ensures the reliable operation of equipment, and reduces the cost of equipment maintenance. |
| Close |
|
|
|