Development of an ANN-Based Closed-Form Equation for the Prediction of Airblast Overpressure Induced by Construction Rock Excavation Blasting in Urban Areas

Blasting has been proven to be the most cost-effective method for rock excavation known man. The cost-effectiveness advantage of blasting is overshadowed by its unpleasant environmental problems, particularly at construction sites close human settlements and public utilities. Therefore, efforts are required develop closed-form equations that can accurately predict problems associated with blasting. This study proposes an ANN-based explicit equation forecasting airblast overpressure (AOp) multiple in South Korea. Nine important factors affect AOp generation were used model. First, a stand-alone ANN was initiated, hyperparameters optimum structure tuned using two novel robust metaheuristic algorithms: slime mould algorithm (SMA) multi-verse optimization (MVO). To appraise predictive accuracy developed soft computing models, multilinear regression (MLR) generalized empirical predictor comparison. analysis showed SMA-ANN MVO-ANN models predicted highest compared other models. hybrid transformed into aid easy when planning round sites. model validated practical engineering applications comprehensive relative importance input parameters performed. relevance shows mass rating (RMR), charge per delay (Q), monitoring distance (DIS) have impacts on AOp.