Method and device for predicting cetane number of diesel oil

The invention discloses a method and device for predicting the cetane number of diesel oil, and relates to the technical field of oil and gas technologies, chemical engineering and computers, and the method comprises the following steps: obtaining the molecular composition information of diesel oil to be processed, the molecular composition information comprising diesel oil molecules and the molecular content of the diesel oil molecules; processing the property information of the diesel oil molecules to obtain molecular characteristic information for representing the properties of the diesel oil molecules; inputting the molecular characteristic information into a pre-constructed structural property correlation model to obtain a molecular cetane number predicted value of the diesel oil molecules; the structural property correlation model is obtained by training a neural network model according to sample molecule composition information of diesel sample data and a corresponding cetane number measured value; and