Model Predictive Control Of Vacuum Pressure Swing Adsorption Process

The research systematicly conducted the model predictive controller design of a two-bed seven steps VPSA process for N₂/CH₄ separation.In order to accomplish the optimization and control study of VPSA process,the mathematic model library for VPSA process is selected and established.For the sake of verify the accuracy of simulation,the comparison of simulation and experiment is presented.After optimization work,the VPSA process can achieve the performance of 80%purity and87%recovery.Then the disturbances have been introduced in the optimal operation state,and the requirement of controller is explicit.So that a detailed study of the model predictive controller design for the VPSA system is presented.In this research,two methods have been employed to develop the MPC controller,except the universal model identification tool,the mixing-pool model was adopted in MPC controller design.To verify the effect of controllers,three disturbances scenarios were performed and closed-loop simulations were conducted.The results not only demonstrate the effect of controller that the purity of product can be promised under various disturbances,but also determined the competence of the MPC controllers and proved the method of MPC controller design in this paper.