Method for constructing model for rapid analysis of crude oil properties based on near-infrared spectrum by using Gaussian process regression model

Abstract

The invention discloses a method for constructing a model for rapid analysis of crude oil properties based on a near-infrared spectrum by using a Gaussian process regression model. A method for constructing a model for predicting the crude oil properties comprises the following steps of: (1) determining the property data of a crude oil sample; (2) determining a near-infrared spectrum of the crudeoil sample; (3) preprocessing the near-infrared spectrum obtained in step (2) by subtracting a straight line to eliminate the background interference and the baseline drift; (4) performing principal component analysis (PCA) on the spectrum obtained in step (3), and storing a load vector with the contribution rate of the feature value greater than 95% as P<pca>; (5) using the P<pca> to reduce the dimension of the training subset composed of randomly selected samples, obtaining a sub-model by the Gaussian process regression (GPR); (6) selecting the training set by the Gaussian process regression(GPR) after the dimensional reduction of the near-infrared spectrum of the sample to be tested by the P<pca>; and (7) determining one or more wavenumber segments according to the training set, and using the partial least squares (PLS) method to establish a prediction model for crude oil properties.

Description

technical field [0001] The invention relates to a method and application for constructing a near-infrared rapid analysis of crude oil properties by using a Gaussian process regression model. Background technique [0002] Crude oil evaluation plays an important role in crude oil mining, trade, processing, etc. my country has been carrying out related work on crude oil evaluation. At present, there is a complete set of standard evaluation methods, but these methods often take a long time to analyze and require samples. The amount is large and the analysis cost is high, which can no longer meet the needs of practical applications. Near-infrared analysis technology is currently one of the most promising and widely used rapid analysis methods. In recent years, the application of optical fiber in the field of near-infrared spectroscopy has brought near-infrared spectroscopy from the laboratory to the field. Near-infrared spectroscopy has the advantages of insensitivity to electrom...

AI Technical Summary

Problems solved by technology

For the process industry, the continuity of production often requires the model to be able to track the on-site working conditions in real time; and when the model deviates greatly from the current working conditions and the prediction accuracy cannot meet the needs of online detection, the model can be updated in a timely and effective manner

Images