基于BP神经网络的爆破振速峰值预测

摘要

The ack Propagation (BP) neural network model for prediction of blasting vibration peak particle velocity was established by analyzing the main factors of influencing blasting vibration. The monitored vibration data of a open pit iron mine was compared with the results of the BP neural network model and convention empirical formula. Results show that simple application of the value K and a according to the Sardolfski's formula may cause errors in PPV prediction mainly due to different seismic wave behavior in diferent cases. By contrast, high accuracy and reliability can be achieved by the BP network model.%在分析爆破振动主要影响因素的基础上,建立了对爆破振动质点速度峰值进行预测的BP神经网络模型.将某露天铁矿的爆破振动监测数据分别与模型预测结果和应用传统经验公式进行预测的结果比较,发现由于每次爆破的地震波传播不同,直接应用萨氏公式的K、α值进行预测会引起较大误差,相比之下,BP网络模型的预测结果更为准确和可靠.