2D Iterative Learning Predictive Control For Batch Process Based On Time Delay

In recent years,the batch process helps industrial development to a higher level,which has become an important part of building a modern industrial chain system.At the same time,great achievements have also been made in researches on batch process control.In the control technology of batch processes,the existence of the time delay,input and output constraints and disturbances has a great impact on the stable operation and control performance of the system.Therefore,how to improve the stability of the system with time delay,input and output constraints and disturbances,and improve the control performance of the system has become an important guarantee for the safe operation of the system.The core content of this paper is as follows:In the second chapter,aiming at the state delay,input and output constraints and uncertain disturbances in batch processes,a 2D model predictive iterative learning control method of batch processes with state delay is proposed.Firstly,according to the given batch process,combining with the 2D system theory and iterative learning control law,the original system model is converted to the equivalent 2D-FM closed-loop system model.In the meantime,the optimal performance index with terminal constraints is introduced and the update law is designed to make the objective function reach the minimum upper bound under input and output constraints.The robust constraint set is adopted that the system state is converged to this set round the desired point.Then,the infinite time domain optimization problem is transformed into a convex optimization problem with linear matrix inequality(LMI)input and output constraints,and the sufficient condition is given to ensure the solvability of the model predictive control(MPC)problem in the form of LMI.Finally,through the modeling and simulation of the reactor,it is proved that the above method is feasible.In the third chapter,aiming at the state delay,input and output constraints,uncertainties and multi-stage characteristics of batch processes,a2D model predictive iterative learning control method of batch processes with state delay is proposed.Firstly,the state error and system output tracking error are introduced,the established system model is transformed into an equivalent2D-FM switched system model.According to the transformed equivalent model,a two-dimensional iterative predictive tracking controller is designed by combining the iterative learning control with model predictive control.The corresponding switching signal is designed,and the minimum running time of each stage is obtained by using the average dwell time method.The delay-dependent sufficient condition of the robust hybrid 2D model predictive iterative learning controller is given.Then the optimal performance index with terminal constraints is introduced,the update law is designed,and the optimal upper bound of the performance index under input and output constraints is given.The designed control law not only makes the closed-loop system stable,but also ensures the H_?performance of the closed-loop system.The robust constraint set can overcome the disadvantage that the traditional asymptotically stability cannot converge to the origin when disturbances are involved,and make the system state converge to the expected value.Finally,the superiority of the above strategy is proved by modeling and simulation.In the fourth chapter,aiming at the state delay,input and output constraints,internal and external disturbances,and strong nonlinearity of the system,a 2D model predictive iterative learning control method of nonlinear batch processes with state delay is proposed.Firstly,through the fuzzy control theory,a two-dimensional T-S fuzzy model is established.Then combine with the system state error and output tracking error,the original system state space model is converted to a closed-loop system model in the form of prediction by the Roesser model.In the meantime,the optimization performance index with terminal constraints is introduced and the predictive update law is designed to make the objective function reach the minimum upper bound under input and output constraints.The robust constraint set can make the system state converge to this constraint set,and make it meet the expected value of the system better.Then,the infinite time domain optimization problem is transformed into a convex optimization problem with LMI input and output constraints,and the existence condition of fuzzy predictive update law is given to ensure the solvability of the MPC problem in the form of LMI.Finally,through the simulation result of MATLAB,it is showed that the above scheme is indeed effective.