Trend prediction technology of condition maintenance for large water injection units

Abstract

Trend prediction technology is the key technology to achieve condition-based maintenance of mechanical equipment. Large-sized water injection units are key equipment in oilfields. The traditional preventive maintenance is not economical and cannot completely avoid vicious accidents. To ensure the normal operation of units and save maintenance costs, trend prediction technology is studied to achieve condition-based maintenance for water injection units. The main methods of the technology are given, the trend prediction method based on neural network is put forward, and the expert system based on the knowledge is developed. The industrial site verification shows that the proposed trend prediction technology can reflect the operating condition trend change of the water injection units and provide technical means to achieve condition-based predictive maintenance.

Keywords

water injection units / condition-based maintenance / trend prediction

References

[1] Xu X L, Gu Y H, Shi Y C, Asakura T, Zhang S. Virtual instrument system of condition monitoring and fault diagnosis. ISIST'2004, 2004, 1: 1061–1066

[2] Xu X L, Zuo Y B, Zhu C M. A variable-weight neural network combined predicting model to the trend predicting of the condition development of the large-sized rotary sets. WMSCI 2006, 2006, III: 304–307

[3] Chen J C. Machine learning for information retrieval: neural networks, symbolic learning, and genetic algrithms. JASIS, 1995, 46(3): 194–216

[4] Xu X L, Zuo Y B, Wen H Z, Wang X B, Shi Y C. Application of rapid knowledge acquisition method in intelligent diagnosis instrument. ISTAI’2006, 2006, 2: 1034–1037

Acknowledgements

This paper was supported by the Scientific Research Key Program (KZ200910772001) of Beijing Municipal Commission of Education and Funding Project (PHR20090518) for Academic Human Resources Development in Institutions of Higher Learning under the Jurisdiction of Beijing Municipality.

RIGHTS & PERMISSIONS

2014 Higher Education Press and Springer-Verlag Berlin Heidelberg