Design method of product agile customization based on dynamic multi-artificial neural network models

An approach of agile customization design based on dynamic multi-artificial neural network (M-ANN) models is proposed. The method of artificial neural network is applied to establish the matching relation between customerpsilas requirements and corresponding product structure. The renewable dynamic M-ANN models are formed from which the appropriate network can be selected and the adaptive design scheme could be obtained agilely according to different requirements. The roll plate machine is taken as the example to discuss the establishing process of the network and the validity of this approach.