Neural network simulation for industrial crystalization

摘要

Fast simulation of the process is very useful for the process optimization and control. The neural network is a powerfu ltool for fast simualtion of the complex processes. Idnustrial crystallization includes the mass and heat transfer with complicated nucleation and growth processes and is affected by many factors, mainly by kinetic parameters, mixing conditions and operation ocnditions. Simulating and controllign the crystallization process in an industrial crystallizer become more complicate.d Many simualtion programs have been developed for simualtion of various crystallization process. However only a few work were tried for simualting the crystallization process using neural network in some special cases. The aim of this work is to develop a neural network for simulating continuous process in an industrial crystallizer. It can be used as a basic tool for optimizing and controllign the continuous crystallization.