Simulation study on probability distribution of sequence evaluation in order-based crude oil blending scheduling problem

Bai Liang,Jiang Yongheng,Huang Dexian

DOI: https://doi.org/10.16866/j.com.app.chem2012.01.001

2012-01-01

Abstract:Crude oil blending and scheduling can help refineries improve the feedstock property,which is important to the smooth operation of oil refining unit,implementation of advanced control and fulfillment of production plan.Under the actual production background,a multi-pipeline crude oil blending and scheduling model is established and a novel solution methodology based on a two-level optimization structure and ordinal optimization is proposed correspondingly Ref.[4-5].In this solution methodology,a blending sequence and flowrate hybrid optimization problem changes into a blending sequence's sampling and evaluation problem,which relates to the breadth and depth of search in algorithm,respectively.How to deal with the balance between them is important to rationally allocate the computation and further improve the optimization efficiency.Therefore,as the basis for measuring search depth and breadth,the distribution of blending sequence evaluation is studied in this paper.In first analyzing the similarities and differences between design evaluation in simulation optimization and sequence evaluation in scheduling of crude oil blending optimization,this paper statistically analyzed the sequence evaluation distribution through Monte-Carlo method and then fitted the statistical results using the Weibull distribution.Finally,the evaluation comparison of two blending sequences based on Weibull distribution description is showed and future direction is pointed out at the end of this paper.