The Study On Process Control Of Molecular Distillation Based On Adaptive Dynamic Programming

Molecular distillation is a new and highly efficient liquid separation and purification technology with the advantages of low distillation temperature,high vacuum,short material heating time,and irreversible separation process.These advantages make it suitable for the separation of substances with high thermal sensitivity,high boiling point,and easy oxidation,and are widely used in the food industry,fine chemicals and medical industries.The molecular distillation process is a complex and nonlinear system with multivariable and strong coupling characteristics.The strong coupling of different variables will directly affect the stability of the molecular distillation process,thus affecting the index of distillation products.At present,the process modeling and control optimization algorithms for molecular distillation are still not deep enough,and actual production process control relies on human experience.Therefore,it is of great significance to establish a reliable molecular distillation process model and control it effectively and efficiently to improve the separation efficiency of molecular distillation and promote the automation of production.First,the mechanism model and process flow of the molecular distillation process are analyzed to determine the main process parameters and product indicators of the molecular distillation process.According to the experimental data of the molecular distillation process,the relationship between process parameters and product indicators was studied.Then,in order to improve the product quality in the molecular distillation process,it is necessary to ensure that the distillation process is performed under a high degree of vacuum.The vacuum degree control model of the molecular distillation process was established and a variable universe fuzzy PID controller was designed.The simulation results show that the control algorithm has the advantages of shorter adjustment time,smaller overshoot,stronger robustness than the conventional PID control algorithm,and can meet the control requirements of the molecular distillation process.Finally,aiming at the optimization of the control strategy of molecular distillation process,the ADP controller based on heuristic dynamic programming structure is designed and the corresponding model network is established.The goal is to minimize the quadratic cost function composed of the quantity of molecular distillation control and state quantity.Determine the optimal control strategy.The simulation results show that the heuristic dynamic programming control algorithm can make the molecular distillation process achieve the desired control optimization effect.