Research On Modelling And Optimizing Control Of Quality For Drug Fluidized Bed Granulation Process

Drug production is an important part of national economy,as an important part of drug productive process,the drug fluidized bed granulation process owns a lot of problems such as the difficulty in controlling the quality,the lower production efficiency and serious material consumption.The research on modelling and optimizing control of quality for drug fluidized bed granulation process has important theoretical and practical significance.This thesis first introduces the basic principle and the internal mechanism of fluidized bed granulation process,and then analyses its influence factors.Based on the previous efforts,the mechanism model to describe the average particle size of end particles is established,and the effectiveness of mechanism model is validated by the simulation.Then the mechanism model is used to simulate the real production process of fluidized bed granulation and produce the process date,and the process predicted model is developed with the process data.Based on the developed predicted model,the operating variables,including the pulsed frequency of spray,the binder spray rate and the atomizing pressure,are treated as the decision factors and the error between the expected value and the predicted value of average particle size is treated as the optimization objective to establish the optimization problem.The SQP algorithm is utilized to solve the optimization problem.On account of the model mismatch problem,this thesis considers the uncertainty of model error and present optimization control strategy of granulation process based on the robust optimization.The simulation shows that this method improves the feasibility of the optimization results.And then the idea of iterative learning control is used to realize batch-to-batch optimization control of fluidized bed granulation process,and the result of simulation proves its effectiveness.Lastly,this thesis come up with the integrated optimization control method,which combines the batch optimization with batch-to-batch optimization,to enhance the speed of indicator controlling and the ability of restraining the disturbance.The simulation has validated its effectiveness.