Moving Horizon Estimation And Optimization For Temperature Distribution Of FCCU Riser Reactor

With the rapid development of petroleum industry,fluid catalytic cracking,as a primary petroleum refining method,plays an important role in petrochemical plants.The riser reactor is the main place where the cracking reaction occurs.The temperature distribution of riser reactor plays an important role in fluid catalytic cracking unit(FCCU)since it directly affects the quantity and quality of the product.However,there are only limited thermocouples installed on the FCCU riser and it is impractical to place a great many thermocouples along the reactor.If the temperature distribution of the riser could be accurately estimated and reasonably controlled,not only the reaction intensity can be monitored real time,but also a safe and reliable operation can be ensured and good economic benefits can be obtained.In this paper,we focus on the estimation and control of the temperature distribution of FCCU riser.A state estimator is designed to estimate the dynamic temperature distribution using limited measurements.Based on the estimator,a controller is further proposed to control the temperature distribution.First of all,we divide the reaction system into five virtual components and establish a fivelump dynamical reaction model for the first and the second reaction zone of the riser reactor.Since it is not easy to estimate the temperature by using a nonlinear partial differential equation model,finite element method is used to transfer the model to lumping parameter model.Finally,the numerical results show the accuracy and reliability of the proposed model.Secondly,a neighborhood optimization based distributed moving horizon estimation(NDMHE)strategy is proposed.The riser reactor is divided into several subsystems according to the location of thermocouples,and the temperature distribution of every subsystem is observed by an individual small scale moving horizon estimation(MHE).These MHEs are solved serially.Consequently,the computing time is saved to satisfy the requirement of computing speed for online implementation.In addition,each subsystem considers the performance of itself and its downstream neighbor together to improve the estimation accuracy.The proposed method is applied to a FCCU in JiuJiang Petrochemical Co.Ltd.Both simulation and experimental results show the efficiency of the proposed method.Finally,a thermal energy correction based model predictive control(MPC)is applied to the control of the temperature distribution.Since the temperature can only be measured at a few positions,MHE is incorporated in the control loop to estimate the temperature distribution of riser reactor.Moreover,in order to improve performance,a thermal energy correction term is added to the formulation of MPC.Thermal energy is a key index that reflects the conversion rate of reaction.By adding the thermal energy term,the yield could be ensured even when disturbances occur.Finally,a thermal energy correction based MPC algorithm is formed,and the validity of the proposed algorithm is verified through simulation.