Study On The Quantifying Of Refined Oil Mixtures By Terahertz Time-domain Spectroscopy

The majority of refined oil is transported through pipelines. Due to the cost ofpipes, different kinds of refined oil are often transported in one pipe alternately.Therefore, it is inevitable that refined oil mixtures are generated in the junction of twokinds of refined oil. Quantitative analysis of refined oil mixtures has significance intracking, cutting, and refining the oil mixtures. If unqualified refined oil entered intooil storage tank, the entire tank of oil will be polluted, causing serious economiclosses. A Safe, accurate, and efficient quantifying method has important economicsignificance and strategic significance for reducing many kinds of accidents and lossin the process of oil storage and transportation. In this work, the quantifying of refinedoil mixtures based on terahertz time-domain spectroscopy was studied. The majorcontent shows below.Based on the characteristics of composition of refined oil, the electron conductormodels of different kinds of oil in the terahertz band were established, the keyparameters of electron cloud conductor model and the refractive index formula ofdifferent types of oil were given.The terahertz time-domain spectroscopy analysis system for refined oilidentification was established. A8F light path system structure was used to study fourkinds of refined oil in terahertz frequency, including time domain waveform, specta,absorption coefficient spectrum and refractive index spectrum. The optical parametersof90#,93#,97#gasoline and-10#diesel were analyzed.The characteristics of time domain waveform, specta, absorption coefficientspectrum and refractive index spectrum of6kinds of oil mixtures were analyzed. Toquantify the components of refined oil mixtures, a method is proposed to evaluate thebest frequency band for regression analysis.With the data in this frequency band,dualistic linear regression fitting is used to determine the volume fraction of gasolineand diesel in66different kinds of mixtures. Result shows that, the mixtures ofgasoline and diesel were will fitted and the Adj.R-Square of33samples are above0.9(maxmun is1), absolute error of gasoline volume fraction is under6%.A multiparameter-combined analysis with time domain waveform was proposed.The multiparameter represents the time delay and amplitude of the first transmission dip and peak in the time-domain pulse. The relationship between these fourparameters and the gasoline/diesel content in oil mixtures was thoroughly investigated,and four distinct calibration models for quantifying oil mixtures were built using leastsquare fitting. To enable the development of an informative and accurate calibrationmodel, the four individual models were given proper weights and combined. Theweight was determined by the cosine-optimal method, which aimed to determine themost proper weight under the condition of the cosine of the angle between the fittedcontent vector and the real content vector that reaches the maximum. This method ismore accurate and precise than the method of using terahertz-absorption-coefficientspectra and dualistic linear regression fitting. For the use of quantifyinggasoline-gasoline mixtures, the absolute error of gasoline volume fraction decreasesfrom18%to2%.