Predictive on the Shearing Property of Fabrics Based on Flexible Neural Network

Abstract

For the flexible neural network is a kind of network which includes flexible and the invariable parameter S function, it not only adjust connection weight in training process, but also adjust parameter of S function, so it has the higher convergence rate and generalization ability than the BP neural network, which is based on the article use fabric structure parameter as input, fabric shearing property as output, and use bipolarity flexible neural network to predict to the fabric shearing property, the results show that the predictive accuracy of flexible neural network is higher than Bp neural network.