There are many influential factors of molecular distillation which is a complex process and coupling relationship with each other. Considering the production efficiency and product quality in actual production, it' s obviously important to choose various suitable parameters and it is difficult to grasp accurately by traditional orthogonal testing method. In this paper, genetic neural network model is used to establish the "black box model" between input and output and optimize the parameters of the system. The inputs are temperature, vacuum and feed rate and the outputs are the content and yield of essential oil in schisandra chinensis. The optimization model for the production technology and index is built up in MATLAB 7.0 by genetic neural network then the optimum operative parameters are obtained to conduct production.%影响分子蒸馏系统工艺的因素及过程较为复杂,并且各因素之间具有耦合关系,在实际生产中需要兼顾产品质量及生产效率,因此,选择合适的工艺参数值尤为重要,而传统的正交试验方法对这种具有耦合关系的参数系统难以准确把握,本文应用遗传神经网络,建立输入输出的黑箱模型,对参数进行优化.其中,以蒸馏温度,真空度、进料速率作为输入量,以精油的含量和得率作为输出量,利用遗传神经网络原理,在MATLAB 7.0平台上建立起生产工艺参数与产品指标之间的优化模型,得到最佳工艺参数值,以此指导生产.
展开▼
