Font Size: a A A

Abnormal Sample Detection & Improvement In Online Near-infrared Spectroscopy Analyzer

Posted on:2007-07-20Degree:MasterType:Thesis
Country:ChinaCandidate:Y Z HanFull Text:PDF
GTID:2121360182490410Subject:Pattern Recognition and Intelligent Systems
Abstract/Summary:PDF Full Text Request
Gasoline octane number is one of the most important quality characters of petroleum products, which represent anti-blast ability. Traditional laboratory analysis methods are usually not suitable because of their cost, long delay, complication of operation and maintenance. Near-infrared spectroscopy (NIR) is a non-destructive and rapid analysis method, which is preferable to rapidly determine the properties of petroleum products online. This thesis researched the application techniques of NIR and developed a new online gasoline octane number analyzer. The main contributions of this thesis are as follows:1. Introduce the construction of the Online Near-infrared Spectroscopy Analysis System, and then introduce the application of the online NIR analysis tech. in-country and oversea.2. Because spectrogram pretreatment could only be used weakening the yawp affection which comes from sampling distance, difference of sample color, spectrogram excursion, detection yawp and circle yawp, but could not distinguish the abnormity sample, we use Principal Component Analysis(PCA) & Mode class to find abnormity sample. It is confirmed in online NIR analysis, and guaranteed the data flatness and exactness.3. Based on the old system, we made a sampling system which could be cleaned itself, and designed new software system of server and client. The new system could be much easier to be maintained. In this system, client PC collects sample spectroscopy, and sends it to server PC;then server PC carries out pretreatment, mode prediction and database management. After some time using in a factory, it is proved that our system could be fit for online gasoline octane number analysis.
Keywords/Search Tags:online, near-infrared spectroscopy, spectroscopy pretreatment, principal component analysis, mode class, gasoline octane number
PDF Full Text Request
Related items