爆破地震峰值预报神经网络研究

STUDY OF NEURAL NETWORK PREDICTION ON PEAK PARTICLE AMPLITUDE OF BLASTING GROUND VIBRATION

在分析某核电站一期地面爆破振动监测结果的基础上，用神经网络理论建立了爆破振动加速度峰值的预报模型，提出了神经网络模型预报加速度峰值的方法。将神经网络模型预报的结果与传统方法（经验公式法）预报的结果相比，前者的预报结果有明显的改善。

On the basis of analysis of datum from monitoring the ground vibration of bench blasting near a running nuclear power plant,a neural network model has been developed to predict the peak particle acceleration amplitude of blasting vibration.Comparing the results calculated in empirical formula with the results from neural network simulation,the later results are obviously improved.