摘要
爆破振动一般采用萨道夫斯基公式进行回归分析,求得K,α值,需要大量的振动速度监测数据来控制误差。对比了传统的最小二乘法的线性回归分析方法和构造方程-求导-迭代的非线性回归分析方法,发现在数据量较少的情况下,非线性回归分析方法取得了较为真实且误差更小的结果,并以缅甸莱比塘铜矿采区内古建筑物周边爆破振动控制实例为背景,利用非线性回归分析方法处理已测振动监测数据,得到适用于此古建筑物周边地质条件的K=151、α=1.702以及误差r=29.2%,以此指导爆破设计、施工及振动预测,有效的保护了古建筑物的安全。
Blasting vibration is generally used Sodev formula for regression analysis. Acquired the value of K and a was need lots of vibration monitoring data to control error. Linear regression analysis method of traditional least squares and nonlinear regression analysis method of construction equation-derivative-iterative was compared. It was found that a more real and smaller error result was acquired by nonlinear regression analysis method under the less data. Based on the engineering example of control blasting vibration in the surrounding of ancient buildings in Myanmar Letpadaung copper mine, the nonlinear regres-sion analysis method was used to process the obtained vibration monitoring data and K = 151, a = l. 702, r = 2 9 . 2% were ac-quired. And the results are suitable for the surrounding geological conditions of the ancient building. It could guide the blasting design, construction and vibration prediction. And the safety of the ancient buildings were effectively protected.
出处
《工程爆破》
2017年第4期59-62,共4页
Engineering Blasting
关键词
爆破振动
古建筑物
萨道夫斯基公式
线性回归分析
非线性回归分析
振动监测
blasting vibration
ancient buildings
Sodev formula
linear regression analysis
nonlinear regression analysis
vi-bration monitoring