To find an effective solution for testing the quality of finished gas products or component gases??researchers at the Zhejiang University applied small wave conversion and spectrum integration technologies in spectrum data pre-handling, in an attempt to reduce the impacts of fluorescent background interference and high-frequency noises on the accuracy of analysis. In developing a near-infrared based online analyzing system for gas quality, researchers noticeably raised the accuracy of analysis, using vector mathematics in non-linear quantitative analysis of near-infrared spectrums, which avoids the limitation of the Partial Least 2 Squares (PLS) that is frequently used in quantitative analysis of near-infrared spectrums. Researchers also produced a near-infrared spectrum based fast approach to identify gas specifications. Using a classification model built on the near-infrared information of primary gas elements, the new method makes a fast classification for unknown gas specimens. 

Based on a range of findings achieved, the research team developed a new generation instrument that can test gas quality at a fast speed but for a lower cost for lab applications. The new system has found successful applications at Hangzhou Oil Refinery and Qingjiang Oil Refinery, both affiliated to SINOPEC.  In the meantime, researchers rolled out the systems for automatic cleaning of near-infrared probes, and for automatic sample collecting, which enhances the consecutive running capacity of the online system. The prototype near-infrared online analyzing system has been successfully incorporated in the consecutive reforming unit at the Shanghai Gaoqiao Refinery, a part of SINOPEC.