Study On Uncertainty Control Method For Catalytic Cracking Reaction-regeneration System

Fluid catalytic cracking reaction-regeneration(FCCR-RG)system is an indispensable part of the industrial chemical process,especially in the refinery which cracks heavy feedstock oil into light product oil.Therefore,modeling and optimizing control of the FCCR-RG system is of great significance.However,due to the complex composition of the feedstock,the large cracking reaction network,and the complex catalyst regeneration process,the modeling of the system becomes very difficult.At the same time,because of the complexity of the reaction device and various disturbances in the industrial field,it is very difficult to tune and optimize the controller parameters of the system.Therefore,studying FCCR-RG system modeling and control has practical significance.The research content of this thesis mainly focuses on the modeling and optimization control of FCCR-RG system.The main research contents can be summarized as follows:(1)Research on mechanism model of FCCR-RG system.In this thesis,based on the reference pseudo-components,the feedstock oil is re-described.the reaction network of feedstock oil cracking is obtained,and the riser mechanism model is constructed.The regenerator mechanism model is constructed based on the charring dynamics mode.By comparing the data with the actual factory,the FCCR-RG system mechanism model is verified to obtain a mechanism model that can reflect the real system,and also provides a reliable model for the identification algorithm(2)Research on interval parameter identification method for multivariable systems.In FCCR-RG system,the catalyst flow rate from the regenerator to the riser and the flue gas flow rate from the top of the regenerator are selected as control variables,and the temperature at the outlet of the riser and the internal pressure of the regenerator are selected as controlled variables to establish a 2×2 multivariable system.At the same time,considering the complexity of the system and the influence of various disturbances in the actual production on the model parameter identification,a model parameter identification method based on affine bat algorithm is proposed.Taking the minimum squared error as the optimization objective function and the FCCR-RG system model parameters as the optimal parameters,the model parameter identification based on affine bat algorithm is realized based on MATLAB,and the parameter interval of multivariable system is obtained,which provides reliable transfer function model for controller design(3)Interval parameter single variable system controller design.For parameter uncertain models,the model needs to be analyzed firstly.Considering that the boundary model equivalent substitution method can effectively reduce the amount of calculation,this thesis mainly discusses the parameter uncertainty model based on the boundary model equivalent substitution method.In this thesis,the controller of parameter uncertain models are tuned in two directions.First,the robust performance of the model is limited,and the value range of the controller parameters is determined by a graphical method.The second way is to limit the disturbance rejection performance of the model and calculate the value range of the controller parameters.Through the discussion of single variable system with uncertain parameters,the design idea of controller tuning for multivariable system with uncertain parameters is provided.(4)Interval parameter multivariable system controller design.For the strong coupling relationship in the reaction regeneration system,the multivariable system needs to be decoupled first.Considering that the equivalent open-loop transfer function(EOTF)method can be well adapted to the tuning method of the controller in the single-variable system while decoupling.Therefore,the EOTF method is chosen to study the parameter uncertain multivariable system.In this thesis,a graphical tuning method for controller parameter interval is proposed for parameter uncertain multivariable systems.Based on EOTF method and Kharitonov theorem,the boundary model is used to substitute the parameter uncertain multivariable system.At the same time,the margin tester is introduced to limit the robust performance of the system,and the value space of the multivariable system controller parameters is obtainedThe work done in this thesis is oriented to the actual chemical process,which provides a theoretical basis for the model establishment of the FCCR-RG system,and also provides an effective design method for the optimal control of the actual complex FCCR-RG system.