BP neural network based Taguchi method and its application in the preparation of composite photocatalyst

The traditional Taguchi design method has been widely accepted and recognized as an important tool for improving the quality of a product or a process. However, the optimal setting of a design parameter may sometimes be very impractical since it is determined only among the levels included in the parameter design experiment. In this paper, based on the capability of self-learning, self-training and output prediction of BP (Back Propagation) neural network, we combine the Taguchi method with BP neural network to optimize the preparation parameters of composite photocatalyst TiO2-MWCNTs.Firstly, we construct the BP neural network model using the Taguchi method to determine the number of neurons, learning rate and momentum values. Then, the BP neural network is trained using sample data and is applied to optimize the preparation parameters of the photocatalyst. Finally, we evaluate the fitting precision of the BP neural network and verify the validity of this method by comparing the results with the optimized results of classical Taguchi method.