Abstract:

When it comes to blasting in soil, in general, and blasting in clay medium under water, in particular, typical parameters for destruction effect of explosion known as splashed funnel based on water depth, the depth of buried explosives in clay medium and the explosive mass. This relation is multidimensional and multivariable. Using traditional integation addressing empirical data still has limits in presenting a general law in entire domain relating to such relation. Therefore, based on the empirical results collected from the previuos study, this paper will concentrate to make a machine learning algorithm building a regression model, finding the general empirical law about the dependence relation of the radius of splashed funnel in clay medium under water, based on splashed funnel and the water depth, the depth of buried explosives in clay medium and the radius of explosives charges. The efficiency of the model will be evaluated with correlation coefficent R2 between the calculated values of funnels and its real values in experiments. Consequently, the model reached high accuracies which can be applicable in the reality.