Font Size: a A A

Application Of Partial Least Squares Regression In Oilfield Cementing Quality Prediction

Posted on:2015-10-19Degree:MasterType:Thesis
Country:ChinaCandidate:J Y LiFull Text:PDF
GTID:2180330431994925Subject:Applied Mathematics
Abstract/Summary:PDF Full Text Request
The cementing quality is a comprehensive reflection of cementing and drilling design,construction level, which is also one of the key links of drilling engineering. By analyzing thefactors which influence the cementing quality, we can control the process of cementing,and can also make a better design and construction of cementing, which also have a very goodhelp for cementing effect. With the development of oil field development technology andhigher requirements of the quality of cementing, the cementing quality is critical for oil andgas wells normal production, life of oil and gas wells, exploration and development of theoverall effectiveness of influence. Through a variety of factors which affects the quality ofcementing analysis, multivariate quality of research carried out to improve the cementingquality is very important.Based on a large amount of historical data, this study firstly analyzed the factorsaffecting the cementing quality and the influencing factors of influence on cementing quality.Then introduce multiple linear regression equation, extraction method of data tablecomponent, the basic principle and development of partial least squares regression. Accordingto the introduction to the theoretical foundation, established the multivariate linear model,linear partial least squares regression, partial least squares path model and partial least squaresrecursive model. The data were trained and predicted by these models, the predicted valueobtained by cementing quality model, according to the predictive value to adjust theparameters can be controlled, so that optimal cementing quality. And the cementing operationparameters optimized design research, which can improve the quality of cementing. Beingable to carry out smoothly in later production, requires higher quality of cementing operations.The current study of improve cementing quality is only one-sided; there is no comprehensiveand detailed approach to improve the cementing quality. With the development of cementingindustry, people study cementing quality from more areas, and bring up more and moremethods in order to be able to succeed in cementing operation.
Keywords/Search Tags:Cementing Quality, Regression Analysis, Partial Least Squares, ParticleSwarm Optimization
PDF Full Text Request
Related items