关键词: 复杂城市环境, 爆炸荷载, 神经网络, 快速预测

Abstract: The staggered layout of urban buildings increases the complexity of the propagation path of explosion shock wave, which increases the difficulty of evaluating the damage effect comprehensively and accurately. Numerical simulation methods based on computational fluid dynamics can accurately calculate the blast loading, but the calculation amount is large and the calculation time is long. In order to realize the rapid prediction of blast loading in complex urban environment, a fast blast loading prediction method based on neural network is proposed. The influence of the number of training samples on the prediction accuracy and the effect of regional division on the prediction performance of the model are analyzed. In order to meet the data requirements of neural network model training, explosion simulation software is used to perform mesh sensitivity analysis and generate a dataset of 80 explosion examples while considering simulation speed and accuracy. In order to determine the appropriate model structure, fully connected neural networks with different layers are constructed for comparative experiment and analysis. Through comparative experiments, the effects of the number of training samples, the division of region and the construction of dual models on the prediction accuracy of the model are analyzed. The results showed that the prediction error of the proposed method is less than 10% on 16 groups of test data that have not been seen before except the training data，and the inference time only takes 2 seconds. It has a balanced and good prediction ability for various ranges of peak overpressure and provides a new approach and perspective for realizing the rapid prediction of explosion load of urban buildings.

Key words: complex urban environment, blast loading, neural networks, rapid prediction