Prediction of peak overpressure of underwater shaped charge based on DACO BP neural network

DOI:	10.11809/bqzbgcxb2023.06.003
Keywords:	underwater shaped charge; peak overpressure prediction; numerical simulation; BP neural network; Mersenne Twister; ant colony; adaptive dynamic ant colony optimization
Abstract:	In order to improve the accuracy and stability of parameter prediction of underwater shaped charge explosion shock waves, this paper proposes a prediction model of peak overpressure of underwater shaped charge through back propagation (BP) neural network optimized by an adaptive dynamic ant colony optimization (DACO) based on the simulation data of charge explosion obtained by AUTODYN numerical simulation software. Mersenne Twister (MT) is used to sort the data randomly to improve the generalization ability of the model to different data distribution. The pure growth strategy of pheromones and the hyperbolic dynamic adaptive adjustment strategy of volatilization coefficients are designed to improve the global optimization ability and convergence speed of the ant colony algorithm. The global optimal solution obtained through DACO algorithm is mapped to the weights and thresholds of BP neural network to improve its prediction accuracy and stability. The experimental results show that the prediction model of peak overpressure of underwater shaped charge based on DACO BP neural network has good effectiveness, stability and credibility.
Published:	2023-06-28
PDF HTML