Study On Intelligent Control Of Vacuum Freeze Drying Technology

Vacuum freeze-drying equipment in the temperature control system,The control of the shelf temperature determines the level of the freeze-drying process and affects the quality of the freeze-dried material.Shelf temperature control system often use circulating heat medium or electric heating way to sublimation of moisture in the material to provide the required heat,because freeze-drying is a process of heat transfer and mass transfer of porous media,it is affected by the low pressure environment of freeze-drying chamber.It is therefore determined that the shelf temperature system is a large lag,time-varying,non-linear system.In this paper,design the automatic control system based on shelf temperature.Our hospital freeze-drying equipment for the experimental platform and Freeze-drying room shelf temperature for the control object.The parameter self-tuning fuzzy PID controller which combines fuzzy control theory with traditional PID control algorithms is designed.However,in the design of fuzzy controller,the choice of membership function and fuzzy set,fuzzy control rules and the quantization factor and scale factor are mainly based on expert knowledge and experience,so that the controller with a certain degree of subjectivity and randomness.In this paper,the genetic algorithm is used to optimize the membership function,control rules,quantization factor and scale factor of the fuzzy controller to improve the performance of the fuzzy PID controller,which makes the shelf temperature control system better.In addition,vacuum freeze-drying is a process of heat transfer and mass transfer of porous media,and it is difficult for the general mathematical model to describe it accurately.It is very time-consuming and inefficient to study it by using experimental method.In this paper,the BP neural network is used to simulate and predict the freeze drying process,and the genetic algorithm is used to optimize the BP neural network to improve the convergence speed of the BP network,improve the local minimum point and reach the optimal performance probability.The main work of this paper is to use intelligent algorithm to control and optimize the shelf temperature control system and freeze-drying process in vacuum freeze-drying technology.Its content and results have the following aspects:(1)The research progress of vacuum freeze-drying technology is introduced,and the theoretical basis of freeze-drying technology,freeze-drying process and vacuum freeze-drying equipment are briefly analyzed.A set of small vacuum freeze-drying equipment with two-stage compression refrigeration system was developed.The cooling temperature could reach below-80 ℃,and the ultimate vacuum degree could reach 10 Pa or less.The structure was simple and the cost was low.(2)Design and optimization of fuzzy PID controller for shelf temperature control system.In the design process of fuzzy PID controller,because of membership function,fuzzy control rules,quantization factor and scale factor,these parameters need to rely on expert knowledge,will be accompanied by a certain degree of subjectivity and randomness,can not guarantee the control quality requirements.Therefore,the use of MATLAB software to write the program,the use of genetic algorithm to optimize the design of the fuzzy PID controller membership,control rules,scale factor quantization factor parameters.Finally,it is proved that the genetic algorithm has a significant effect on the optimization of the fuzzy controller,so that the optimized controller has the advantages of short adjustment time,small overshoot and strong anti-interference ability.(3)Design and construct GA-BP neural network algorithm to simulate and predict the vacuum freeze-drying process.The BP neural network with single hidden layer was constructed by using the freeze-drying chamber pressure,the heating plate temperature and the material thickness as the input of the neural network,the drying rate and the shrinkage of the material as the output of the neural network.The BP neural network is simulated by MATLAB software,and the weights and thresholds of BP network are optimized by genetic algorithm.The results show that the BP neural network theory based on genetic algorithm is feasible and feasible for the simulation and prediction of freeze drying process,and has high industrial application value.