Explosive blast prediction using MLP network based training algorithm

Peoples have been studying the blast wave profile resulting from detonations for many years. Through extensive experimentation, they have been able to predict the propagation profile of blast waves given certain parameters. However, previous studies have primarily focused on the central point of initiation for spherical explosive shapes. The purpose of this research is to compare the predictive performance of blast peak overpressure based on the type and shape of the explosive, as well as the point of detonation. To achieve this, the experiment involved detonating 500 grams of PE-4 and Emulex at various distances (ranging from 0.5 m to 4.0 m) and developing a prediction model using a Multilayer Perceptron (MLP) network. Lavenberg Marquardt (LM) training algorithm perform better than Backpropagation (BP) for modelling the Explosive Blast Prediction using Tansig and Logsig training algorithm.