摘要
针对当前爆破振动安全评价实践现状及不足,建立了基于BP神经网络理论的爆破振动模型。该模型可根据各工程实际情况选取影响爆破振动的主要因素作为输入层参数,可同时预测爆破振动峰值和主振频率。工程实例表明:基于BP神经网络的爆破振动模型比常规经验公式更简单实用,预测精度高,更符合《爆破安全规程》(GB722-2003)的评价要求。
In view of the shortage of current blasting vibration safety appraisal, the paper establishes model of blasting vibration based on the BP neural network theory. According to the actual project situation, this network model can take the primary factors which influ- ence blasting vibration as the input level parameter and it can simultaneously forecast the peak value ＆ main frequency. Comparing the new model with convention empirical formula through the project practice, the first is simpler and more accurate. Furthermore, it can easily meet the requirements of ＂Demolition Safety Regulation＂ （GB722 -2003）.
出处
《工程地球物理学报》
2008年第2期222-226,共5页
Chinese Journal of Engineering Geophysics
关键词
爆破振动
BP神经网络
预测
blasting vibration
BP neural network model
prediction
