Multi-source spectrum light fuel oil quality detection system

Abstract

The invention discloses a multi-source spectrum light fuel oil quality detection system. The system comprises a near-infrared light source module, and the near-infrared light source module comprises a heat dissipation base, a halogen lamp holder, a halogen lamp and a reflection cup; a mid-infrared light source module comprises a reflector, an IR light source and a heat dissipation base; an interferometer module is a solid angle mirror interferometer and mainly comprises a laser, a beam splitter, a compensator, a solid angle mirror and a fixed plane mirror; an ATR crystal module adopts zinc selenide as an attenuated total reflection crystal material to realize total reflection; and an integrated flow cell is mainly composed of an SMA adapter, an optical fiber sleeve, a collimating lens, a flow cell main body, a liquid inlet, a near-infrared window, a focus lens, a liquid outlet, a sealing groove, an ATR crystal and an intermediate infrared spectrometer shell. Data fusion processing is carried out by applying a multivariate analysis method, the problems that modeling and maintenance are difficult, analysis errors are large, quality inspection of part of oil products cannot be carried out, instrument consistency is poor, models between instruments cannot be shared and the like are solved, the comprehensive advantages are better, and the result is more accurate.

Description

technical field [0001] The invention relates to the technical field of light fuel oil detection, in particular to a multi-source spectral light fuel oil quality detection system. Background technique [0002] Near-infrared light: Near-infrared light is an electromagnetic wave between visible light and mid-infrared light. According to the definition of ASTM (American Society for Testing and Materials Testing), it is an electromagnetic wave with a wavelength in the range of 700-2500nm. It is customary to divide the near-infrared region into Near-infrared short-wave (700-1100nm) and near-infrared long-wave (1100-2500nm) two regions. Mid-infrared light: Mid-infrared light is defined by ASTM (American Society for Testing and Materials Testing) as an electromagnetic wave with a wavelength in the range of 2500-25000nm. Multi-source spectrum: near-infrared spectrum + mid-infrared spectrum. Light fuel oil: gasoline, diesel, aviation kerosene, jet fuel, etc. [0003] At present, ne...

AI Technical Summary

Problems solved by technology

[0003] At present, near-infrared spectroscopy or mid-infrared spectroscopy combined with chemometric methods can realize rapid detection of fuel quality. So far, some key technologies have not yet been overcome, which limits the large-scale practical application of this technology in fuel oil detection. The main problems: construction The modeling and maintenance work is too difficult, the analysis error is large, and the properties of some oil products cannot be tested, the consistency of the instruments is poor, and the models between instruments cannot be shared

[0007] 3) One property and one model, multiple properties need to establish multiple models, and the modeling workload is heavy

[0008] 4) There are many properties and spectra with strong non-linearity, which cannot be detected by the PLS method, such as insoluble matter, induction period, wear scar diameter, actual colloid, etc., that is, the types of fuel property models that can be detected by PLS are limited

[0010] (2) Models cannot be shared between instruments of the same type: the cost of modeling is very high and the cycle is very long. However, slight differences in instrument manufacturing lead to differences in the near-infrared spectra collected by different instruments for the same sample

When the model established using the spectrum of one instrument is applied to another instrument, the error of the prediction result becomes larger, and in more cases, it cannot be used

Therefore, the difference between instruments and the standardization of spectra have become the bottleneck restricting the application of near-infrared spectroscopy in military fuels.

[0011] (3) Poor instrument consistency: On the one hand, the principle of the analysis method has high requirements on the technical specifications of the spectroscopic instrument, and the spectral baseline noise needs to be suppressed at a level of a few ten thousandths, and the consistency requirements between instruments are relatively high

In addition, the penetration capabilities of near-infrared light and infrared light are also very different, and the sample cells need to use different window materials

Images