Application of BAS-Elman Neural Network in Prediction of Blasting Vibration Velocity
Cai Chen, Qian Qian, Yunfa Fu
Abstract
In the field of blasting Engineering, an important research topic is to get ac-curate prediction of blasting vibration, which is the prerequisite and guarantee of blasting vibration control. Traditional Sodev’s empirical formula and basic neural network prediction model do not have good performance on ac-curacy and convergence speed. To overcome these shortcomings, an improved Elman neural network prediction model that is based on Beetle Antennae search algorithm (BAS) is proposed. After setting up the Elman neural network model with considering the influential parameter of free surface area, the BAS is used to accelerate the network training by searching good initial parameters. The prediction effect of the BAS-Elman model is compared with those of BP algorithm model, Elman algorithm model and GA-Elman algorithm model. The results show that BAS-Elman neural network model has higher accuracy than other models and can be used to predict the blasting vibration velocity of tunnel engineering.