The Research of Explosive Performances Based on Artificial Neural Network

In order to learn the application and development of the explosive performance prediction based on the artificial neural network method,analyzed the applications of BP neural network in explosive performance prediction.To make the prediction error in the acceptable range,selecting the suitable amount of reference as input,and then building BP network learning the sample and training the sample,setting up a network,training and optimizing the network,adjusting the number of hidden nodes and iteration times.Artificial neural network would have more accurate prediction of explosives performance,and have important reference value through researching the explosive charge density,blasting heat,blasting speed and the impact resistance sensitivity.