Research on Jet Loom Data Analysis System Based on Neural Networks

To forecast quickly the operation condition of loom, optimizing operation parameters of loom, and improve the production efficiency of loom. The paper studied operation prediction of loom production based on neural network. Because traditional network method had the defects of slow convergence velocity and low prediction accuracy, BP algorithm was improved by combined algorithms by the merging of impulse item and adaptation of learning rate, network structure and parameters adjustment were used to optimize neural network, and to predict the operation condition of the loom. Research showed that improved BP network has good rate of convergence, the number of training was less and improved the reliability of the algorithm.