Predictive modeling of gasoline octane loss based on XGboost algorithm and multiple linear regression analysis

Yannan Xie; Kang Ji; Mengxiang Chen; Jiangli Zhang

doi:10.1117/12.2641507

4 August 2022 Predictive modeling of gasoline octane loss based on XGboost algorithm and multiple linear regression analysis

Yannan Xie, Kang Ji, Mengxiang Chen, Jiangli Zhang

Author Affiliations +

Proceedings Volume 12306, Second International Conference on Digital Signal and Computer Communications (DSCC 2022); 123061T (2022) https://doi.org/10.1117/12.2641507
Event: Second International Conference on Digital Signal and Computer Communications (DSCC 2022), 2022, Changchun, China

Abstract

For 325 data samples collected from a catalytic cracking gasoline refining unit, first, a selection method of octane number modeling variables is proposed by using a combination of recursive elimination method, regression analysis, and correlation analysis. Second, the main features are evaluated by support vector regression (SVR) and ridge regression, respectively. Error comparison selects 26 main variables as modeling operation variables, Then, a learning device based on Boosting method was proposed to build an octane number (RON) loss prediction model, and finally, uses the model interpretability variance, the root means square error (RMSE) and the number of Bad Cases are three indicators to evaluate the learning ability and overall performance of the octane number (RON) loss prediction model. The sensitivity analysis of the model is carried out to verify the octane number loss. Reasonableness of the prediction model.

Citation Download Citation

Yannan Xie, Kang Ji, Mengxiang Chen, and Jiangli Zhang "Predictive modeling of gasoline octane loss based on XGboost algorithm and multiple linear regression analysis", Proc. SPIE 12306, Second International Conference on Digital Signal and Computer Communications (DSCC 2022), 123061T (4 August 2022); https://doi.org/10.1117/12.2641507

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available