摘要
以山西省某穿越古长城的高速公路隧道爆破开挖工程为实例,根据支持向量机学习原理,建立支持向量机预测模型,以孔径、孔深、孔距、排距、单段最大装药量、总装药量和爆源距作为模型的输入参数,分别预测质点的径向、切向和垂直方向的爆破峰值振动速度及频率,并将预测值与实测值进行对比,以检验模型的精确度。结果表明,支持向量机预测模型对爆破峰值振动速度与频率的预测具有收敛快、精度高等特点,平均误差分别为11.04%、10.16%。利用该模型可以较准确地对爆破振动参数进行预测,在后续的爆破施工作业中,结合预测结果可以更好地对古长城采取有效的保护措施。
Taking the blasting excavation of a highway tunnel engineering in Shanxi Province passing through the ancient Great Wall as an example,according to the learning principle of support vector machine(SVM)the prediction model of SVM was established,with the aperture,hole depth,hole spacing,row spacing,maximum charge of single section,total charge and blasting distance as input parameters to the model,respectively to predict the particle radial,tangential and vertical direction of the peak vibration velocity and frequency,compared the predicted value with measured value to test the accuracy of the model.The results show that the prediction model of SVM has the characteristics of fast convergence and high accuracy for the prediction of the peak vibration velocity and frequency of blasting vibration,with an average error of 11.04%and 10.16%respectively.The model can be used to more accurately predict blasting vibration parameters,in the subsequent blasting operation,combined with the prediction results,the Great Wall can be effectively protected.
作者
江东平
李龙福
朱磊
李明
JIANG Dong-ping;LI Long-fu;ZHU Lei;LI Ming(Maanshan Institue of Mining Research Blasting Engineering Co.,Ltd.,Maanshan 243000,Anhui,China)
出处
《工程爆破》
CSCD
2020年第2期75-79,86,共6页
Engineering Blasting
关键词
隧道工程
古长城
爆破振动
支持向量机
预测模型
tunnel engineering
the ancient Great Wall
blast vibration
SVM
prediction model
