Font Size: a A A

A New Fractionation Process Of Delayed Coking Unit And Research On Model-free Adaptive Control

Posted on:2022-01-01Degree:MasterType:Thesis
Country:ChinaCandidate:W J HuangFull Text:PDF
GTID:2481306569480314Subject:Chemical Engineering
Abstract/Summary:PDF Full Text Request
Delayed coking unit is an important straight-run residue thermal cracking unit,which occupies an extremely important position in the petroleum refining process.In view of the high energy consumption of the fractionation system and the characteristics of its product gasoline and diesel mixed into the downstream hydrorefining unit,this paper presents a new fractionation process in which the main fractionation column can improve the overlap degree of gasoline and diesel fractions,and the absorption and stabilization system can strengthen the re-absorption and weaken the pre-absorption:1)shut down the side-stream diesel stripping column of the fractionation column and the process of returning the rich diesel to the column;2)the diesel fuel and stabilized gasoline produced by the unit are used as absorbent for the reabsorption column to greatly reduce the flow rate of supplemental absorbent in the upstream absorption column,that is,circulating stable gasoline,reduce the processing load of the absorption stabilization system,and the energy consumption of the desorption column and the stabilization column;3)a gas-liquid separator is added after the reabsorption column,which uses the allowance of the rich gas compressor to feed the gas into the compressor,reducing the consumption of the compressor,and the separated liquid is sent to the hydrofining unit.The simulation results show that the energy consumption of the absorption and stabilization system of the new process is reduced by more than 30%compared with the absorption and stabilization system of the existing process.Then,the model-free adaptive control study is carried out on the new process.The method is:1)improve the existing MFAC control method,including(1)propose the expected output theoretical value(yt*),and change the original expected output value(y*)from a single value to interval value(y*,yt*);(2)in the initial stage of control,yt*is used instead of y*for guidance to improve the weight of the expected output to the control algorithm and speed up the process;(3)at the end of control,according to the principle that the calculated value should be as close to the target value as possible within the empirical interval,the genetic algorithm is used to calculate the optimal pseudo partial derivative(?)(k)and the optimal expected output theoretical value yto*,which replaces the original pseudo partial derivative(?)(k)and the expected output theoretical value yt*for control.The calculation results of SISO and MIMO in a propylene distillation column of a gas separation unit show that the number of iterations of the improved MFAC method is reduced by 81.6%and 35.3%,respectively.2)on this basis,the MFAC method based on dynamic data is proposed:(1)establish the dynamic model of the controlled object;(2)establish the operation window of the MFAC method based on dynamic data and software integration technology;(3)realize the dynamic control of the controlled object through the operation window.Finally,the research results of a new process absorption and stabilization system for a 2 million tons/year delayed coking unit fractionation show that:after the application of the improved MFAC method in the composition control scheme,the time required for all products to be qualified is reduced by 55 hours,or 67.9%,compared with that before the improvement;on the other hand,after the application of the improved MFAC method in the temperature control scheme,the time required for all products to be qualified is reduced by 7 hours,or 28.0%,compared with that before the improvement,indicating that the improved MFAC method has a better control effect on the absorption and stabilization system of the new process.
Keywords/Search Tags:Delayed Coking, Model-Free Adaptive Control, Expected Output, Dynamic Simulation, Software Integration
PDF Full Text Request
Related items