The Intelligent Inspecting System Of NIRS Instrument

The applications of Near-infrared Spectroscopy (NIRS) analysis technology have been extended from quality analysis of agriculture product and food to QA/QC and on-line analysis in the petroleum chemistry, pharmaceutical, polymer, tobacco, microorganism ferment and textile. As this technique becomes more and more important ,it gradually becomes the main quality control measures in these fields .It is a new technology which provides precise qualitative and quantitative analysis according to the absorptive character of the sample's chemical constituent on NIR spectral region. Near-infrared Spectroscopy analysis technology is efficient, rapid, nondestructive, environmental friendly and can be run at low costs, however, there are some inherent weaknesses as follows:1) NIRS instrument is easily affected by the condition of environment; such as temperature, humidity, light intensity and so on.2) NIRS analysis technology is an indirect technique which can be used by mathematical model based on chemometrics and has a requirement for the professional user.3) NIRS instrument is so precise highly that the change of every instrument model can affect the measured value of the spectrum and the maintenance costs of after-sales is high.In order to resolve these problems, the author has designed and developed an intelligent testing system and applied in domestic-made NIR instrument , which can inspect the operation parameters of NIR instrument such as temperature, light intensity ,voltage, and etc in real time, can also realize many functions including warning, fault diagnosis, data processing, model transferring and forecasting and so on.The main work in the paper is as follows.1) The hardware design of intelligent testing system: including models selection of controlling apparatus, structure component and connection.2) The software design of system: developing communication, monitoring and control software by using Delphi 7.0 as the development tool, and achieving the anticipating aims on the whole;3) Using fault diagnosis according to the spectral image of the standard sample, employing some effective data processing methods such as Shenk's patented algorithm and Wavelet analysis, transferring model and forecasting NIR spectrum.The experimental results are presented that the intelligent testing system can not only effectively detect and record the operating condition in real- time, which is in favor of the stable and good operation for the NIR apparatus, but also provide abundant reference information for the maintenance of the instrument. It provided the data processing method which can adjust the primitive spectrum and the mathematical model could be used by general user to predicted the spectrochemistry index.