Research On Switching Time Of Economic Optimization And Optimal Control In Economic Model Predictive Control

In recent years,with the further improvement of the requirements of the chemical production process,economic model predictive control(EMPC)has emerged.At present,there are three types of popular methods in EMPC research,namely infinite-horizon EMPC,terminal constrained EMPC and mode switching EMPC.This article focuses on the research of mode switching EMPC.The traditional mode switching EMPC strategy uses the switching time as the switching point between the optimization mode and the control mode,and sets it as a preset parameter.There are three problems with this strategy.First,the economic performance and control performance of the system are not optimal.The estimated switching time based on the experience cannot be guaranteed to be the optimal switching point of the system.The second problem is the poor anti-interference ability.In the actual production process,disturbance may occur at any time.The pre-fixed switching time can not be adjusted according to the disturbance in time.Last but not least,in the long period,the economic performance is poor when the control target is reachable.To solve the above problems,this paper proposes an EMPC algorithm for online estimation of switching time for a kind of nonlinear affine systems.The algorithm is extended to the long-period control process.First,we determine the auxiliary controller in the control mode,analyze the design,stability area,and their respective advantages and disadvantages of the Sontag controller and MPC controller.Finally select the Sontag controller as the auxiliary controller for subsequent research.Second,design the EMPC algorithm to estimate the switching time online within a limited time.The algorithm can be divide two parts:the first part is to estimate the switching time online,change the switching time from constant into variable.Then,the switching time is estimated online according to the parameters such as detected state value,terminal time and so on.The second part is the optimization control part.The switching time obtained in the first part is used as a mode switching condition into the optimization control problem.In order to ensure the accuracy of the switching time estimation,the correction link is designed.The optimal control law is determined through the correction link.Finally,under the long-period,the economic performance and control performance of the system used mode switching EMPC algorithm are analyzed.The limited-time online switching time EMPC algorithm is extended to the long-period process.The cyclic EMPC control strategy is proposed to achieve long-period process optimization and control.This article solves the problems of traditional mode switching EMPC.In a limited time,the EMPC algorithm that estimates the switching time online ensures that the economic performance is optimal.Meanwhile,the control target unreachable is avoided.Furthermore,under the long period,the optimized control of the process is realized through the cyclic control strategy.The optimized mode and the control mode are flexibly switched,and at the same time,the occurrence of disturbances can be dealt with in time.The simulation results show the effectiveness of the methods.