Machine learning for the sulfur content prediction in the diesel hydrotreatment product

摘要

The article analyzes the data of a two-year monitoring of the operation of a diesel hydrotreating unit. The feed characteristics and the unit operation parameters which are most associated with the depth of hydrodesulfurization are selected. Based on the characteristics, the random forest method was used to construct a model for predicting the sulfur content in hydrotreated diesel fuel.