Research of process parameters of molecular distillation on product purity based on rough sets neural network

Molecular distillation is a complex nonlinear chemical production process of which its mechanism is complex and contains several tight coupling variables, therefore it is difficult to establish precise mathematical model. Based on on product yield factors of the molecular distillation, this paper proposes a product purity prediction model based on rough sets and neural network theory, also the realization process is given. And the specific effects on manufacturing process from molecular distillation parameters are analyzed by simulation, the simulation result proves that the product purity prediction model based on rough sets and neural network theory accelerates the learning speed and improves prediction accuracy.