| 초록 |
Near-infrared (NIR) spectroscopy has been extensively utilized in the field of petrochemical and semiconductor industries over the past twenty years. The major advantages of using NIR spectroscopy in these fields are based on its ability for fast and simultaneous multi-compositional detection (calibration model) of gasoline, naphtha, and etching solutions in an on-line system. However, special cares during the modeling process are strongly required to build up selective and accurate on-line calibration models because of compositional complexity of the petroleum-drive products and semiconductor etching solutions, as well as intrinsic broad spectral features of NIR spectrum. In order to establish the robust calibration process, here we presents two technical approaches, namely spectral range optimization (SRO) and pure component selectivity analysis (PCSA). |
