Research On Production Optimization Algorithm And Software Development In Crude Oil Atmospheric And Vacuum Unit

Crude oil is an important non-renewable strategic resource.Atmospheric and vacuum units are the core units in the production of petrochemical enterprises.In order to ensure the stable and efficient operation of the device and obtain higher economic benefits,the refinery usually optimizes its production process.In the process of crude oil purchase,in order to save costs,a variety of crude oils are usually selected for mixing and refining in proportion,that is,crude oil blending.The selection of crude oil varieties and blending ratios are important,but the crude oil components are complex,and the amount is calculated by manual experience Large and time-consuming and labor-intensive,current research on crude oil blending also focuses on finding suitable alternative crude oils,and less consideration of economic issues such as crude oil purchase costs and product benefits.In the crude oil refining process,in order to reduce losses and increase revenue,the refinery will produce sideline products with a certain yield according to the plan.Among them,it is necessary to accurately control the distillation temperature of each sideline.The current enterprises to solve these problems need to be based on process simulation software.By artificially setting the initial temperature value,obtaining the sideline yield at the current temperature,and repeatedly adjusting the temperature parameters until the target yield is reached,the process is cumbersome and has large errors,so an automated algorithm is urgently needed to solve it.In this subject,a new optimization model is created and solved for the crude oil blending problem.Aiming at the optimization problem of distillation temperature parameters under the target yield of the side line,an automatic temperature checking algorithm is proposed.Finally,a crude oil distillation simulation calculation software was developed to provide information solutions for refineries.The main contents include:(1)A non-linear multi-objective optimization model with the smallest deviation of similarity between the blended crude oil and the target crude oil and the greatest economic benefit is proposed.The optimal solution set Pareto frontier is obtained using NSGA-? based on the reference point selection mechanism.The experimental results are compared with MOPSO.Compared with the algorithm,it has a better optimization effect,and the obtained crude oil blending scheme can provide a reasonable decision for the refinery.(2)An automatic temperature check algorithm is proposed,modeled by numerical parameter optimization distillation temperature,thereby obtaining temperature-side separation lines according to types of crude oil,the yield of the target line and each side of the distillation scheme,experimental verification of the algorithm fully meet the requirements of the minimum error by automating Calculate to solve practical problems for the refinery.(3)Designed and implemented a set of crude oil distillation simulation calculation software.The software integrates functions such as crude oil data management and simulated distillation calculations,and provides crude oil blending and automatic temperature checking and optimization,making the refinery more standardized and automated in data processing.This software has strong versatility and can be widely used in the petrochemical industry.