Prediction on strength of backfill body based on RBF neural network

The influence of the admixture on the filling body strength is studied using RBF neural network based on its outstanding merits and the filling body strength of admixture by the trained RBF neural network is predicted and analyzed.The relative error of predicted value and measured value of training samples in RBF neural network model is between 0.00%~1.19%,and the relative error predicted value and measured value of test samples 4. 76%~6.08%.The results show that this prediction model not only can improve the efficiency of experimental work and save manpower and materials but also can have higher accuracy.