Prediction of blast induced ground vibration (BIGV) of quarry mining using hybrid genetic algorithm optimized artificial neural network

摘要

Efficient prediction of Open Pit mining blast induced ground vibration, has an important role in reduction of the environmental complaints. This paper proposed a new hybrid evolutionary artificial neural network (ANN) optimized by genetic algorithm (GA) to predict peak particle velocity (PPV). The proposed GA-ANN suggests a systematic and automated way to find out a proper ANN architecture namely; number of neurons, activation functions, training algorithm and number of epochs. A data set consisting of maximum charge weight per delay, horizontal distance (HD), radial distance (RD) and a new modified radial distance (MRD) between monitoring and blasting station provided at Sungun Copper Mine site in Iran were used to validate the proposed approach. Comparing the performance of the proposed GA-ANN model by statistical indices indicate the superiority of the GA-ANN model against the empirical predictors and neuro-fuzzy inference system. As an important finding, incorporating MRD instead of the conventional distance measures of HD and RD improves the accuracy of the prediction. Finally results signify the efficiency of the proposed GA-ANN approach in finding optimum architecture of ANN while trying to predict PPV. (C) 2019 Elsevier Ltd. All rights reserved.