Application Of Fuzzy Neural Network Pid In Atmospheric And Vacuum Distillation Process Simulation

Petroleum refining and chemical products are rich in variety and widely used in various fields.As the "leader" of global manufacturing industry,China needs huge oil production.In order to offset the negative impact of petroleum as a strategic resource,it is imperative to improve the production efficiency of petroleum refining and chemical industry.Atmospheric and vacuum distillation unit is the first process of petroleum refining,from which crude oil is separated into various product oils and raw materials of downstream production lines.The process of atmospheric and vacuum distillation unit is relatively complex,and the correlation between internal variables is strong.As a controlled object,it has serious lag and strong nonlinearity.Traditional PID control method has great limitations.In order to improve the control effect,the intelligent control method is combined with PID control,and the process simulation technology is used to model the atmospheric and vacuum distillation unit,so as to build a dynamic simulation platform and verify the controller effect.First of all,in this paper the atmospheric vacuum distillation device of a plant is taken as the research object.According to the basic data of crude oil feed and the process requirements of the unit,the steady-state modeling of atmospheric and vacuum distillation unit is carried out with the help of process simulation software Aspen Plus.And the simulation data after debugging convergence is within a reasonable error range.Then,based on the steady-state simulation results,the data are supplemented and dynamically converted.The dynamic simulation of atmospheric and vacuum distillation is carried out by Aspen Plus Dynamics software.And the dynamic operation data are derived.The main variables are analyzed,and the control loop of atmospheric and vacuum distillation process is designed.With the help of AM Simulation module of Simulink,the established dynamic model of Aspen Plus Dynamics is called to establish co-simulation.Finally,the fuzzy neural PID algorithm controller is designed.With the help of MATLAB’s powerful computing power,the main circuit of the Atmospheric vacuum distillation unit is controlled by joint simulation.The conventional PID algorithm is compared with the fuzzy PID algorithm.Experimental analysis shows that the dynamic performance of fuzzy neural PID is stronger and more robust than that of traditional PID algorithm.Aiming at the modeling problem of atmospheric and vacuum distillation complex process,the joint simulation platform built by process simulation has a higher degree of fit with the actual production model.And the improvement of the controller has certain reference value for the controller design of actual atmospheric and vacuum distillation unit.