Forecast system for blasting vibration velocity peak based on Matlab and BP neural network
-
摘要: 爆破振动预测是一个复杂的非线性问题，可应用非线性功能强大的BP神经网络技术来解决，但由于其数值计算量大、可操作性不强等特点，在实际工程中应用困难。为了解决该问题，本文中将Matlab程序的强大计算能力与VB的友好界面相结合，利用ActiveX自动化技术和BP神经网络算法，开发得到爆破振速峰值预测系统。该预测系统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参数，以预测爆破振速峰值。通过在北京市昌平线暗挖区间隧道工程中的应用表明：该预测系统在实际工程中使用方便，操作简单，预测精度高，人机交互界面友好。Abstract: In this work we combined the powerful computing capabilities of Matlab programs combined with VB friendly interface and developed the forecast system for blasting vibrational velocity peak using ActiveX automation technologies and BP neural network algorithm. The forecast system can select as an input parameter various factors affecting an engineering project's blasting vibration. The actual application of the system in the construction of the underground cut tunnel in the Beijing-Changping Line shows that this system is simple and convenient to use in practical engineering, accepted for its high precision of prediction, good application effect and friendly human-computer interaction interface.
-
Key words:
- VB /
- Matlab /
- BP neural network model /
- blasting vibration /
- forecast system
-
表 1 现场实测数据
Table 1. Measured data
组号 爆心距/m 高程差/m 总药量/kg 炮眼深度/m 单段最大药量/kg 振速峰值/(cm·s-1) 1 35.68 9.45 72 2.2 1.5 0.61 2 36.15 16.26 72 2.2 1.5 0.94 3 17.30 13.21 60 2.2 1.2 2.90 4 17.87 17.00 60 2.2 1.2 3.50 5 34.02 13.17 66 1.7 1.2 0.89 6 30.31 17.57 66 1.7 1.2 1.41 7 25.28 6.67 78 2.5 1.8 0.92 8 26.91 12.17 78 2.5 1.8 1.17 9 30.14 17.67 78 2.5 1.8 1.61 10 17.15 5.72 80 2.5 1.8 4.11 11 17.64 11.22 80 2.5 1.8 2.88 12 20.68 16.72 80 2.5 1.8 4.66 13 15.43 5.72 84 3.0 1.8 3.66 14 16.78 11.22 84 3.0 1.8 3.54 15 14.25 5.72 54 2.0 1.2 3.07 16 16.44 11.22 54 2.0 1.2 2.41 17 20.91 16.72 54 2.0 1.2 4.79 18 34.70 14.06 72 2.5 1.2 0.71 19 32.00 17.84 72 2.5 1.2 0.98 20 39.01 16.86 72 2.5 1.2 1.27 21 20.05 8.52 72 2.2 1.5 1.48 22 20.24 12.62 72 2.2 1.5 1.77 23 21.74 16.72 72 2.2 1.5 2.26 24 13.59 5.72 66 2.5 1.5 3.59 25 16.50 11.22 66 2.5 1.5 3.90 26 17.15 5.72 80 2.5 1.8 4.34 27 18.64 12.22 80 2.5 1.8 2.93 28 20.68 16.72 80 2.5 1.8 4.61 
下载: 导出CSV
表 2 实测振动速度与预测振动速度对比
Table 2. Measured and predicted results of vibrational velocity
组号 实测振动速度/(cm·s-1) 系统预测振动速度/(cm·s-1) 相对误差/% 萨氏预测振动速度/(cm·s-1) 相对误差/% 21 1.48 1.35 8.78 2.38 60.81 22 1.76 1.96 11.36 2.35 33.52 23 2.26 2.54 12.39 2.13 5.75 24 3.59 4.13 15.04 4.14 15.32 25 3.90 3.25 16.67 3.14 19.49 26 4.34 4.02 7.37 3.25 25.12 27 2.93 2.73 6.48 3.12 6.48 28 4.61 4.42 4.12 2.64 42.73 平均误差/% 10.27 26.15
下载: 导出CSV
-
[1] 吕淑然.爆破震动强度预测的神经网络模型研究[J].金属矿山, 2009(6):40-43. doi: 10.3321/j.issn:1001-1250.2009.06.012Lü Shuran.Study on the neural network model for forecasting the blasting vibration intensity[J].Metal Mine, 2009(6):40-43. doi: 10.3321/j.issn:1001-1250.2009.06.012 [2] 张艺峰, 姚道平, 谢志招.基于BP神经网络的爆破振动峰值及主频预测[J].工程地球物理学报, 2008, 5(2):222-225. doi: 10.3969/j.issn.1672-7940.2008.02.017Zhang Yifeng, Yao Daoping, Xie Zhizhao.The prediction of blasting vibration peak vlue & main frequency by BP neural network[J].Chinese Journal of Engineering Geophysics, 2008, 5(2):222-225. doi: 10.3969/j.issn.1672-7940.2008.02.017 [3] 李存国, 李俊国.马兰庄铁矿生产爆破地震效应神经网络模型[J].中国钼业, 2005, 29(5):20-24. doi: 10.3969/j.issn.1006-2602.2005.05.004Li Cunguo, Li Junguo.A neural network model of producing blasting vibration effects in the Malanzhuang Lronmine[J].China Molybdenum Industry, 2005, 29(5):20-24. doi: 10.3969/j.issn.1006-2602.2005.05.004 [4] 周品.Matlab神经网络设计与应用[M].北京:清华大学出版社, 2013. [5] 张立明.人工神经网络的模型及应用[M].上海复旦大学出版社, 1993. [6] 李永松, 艾凯, 尹健民.基于VB与Matlab的BP神经网络在地应力场分析中的应用[J].长江科学院院报, 2009, 26(6):24-27. doi: 10.3969/j.issn.1001-5485.2009.06.007Li Yongsong, Ai Kai, Yin Jianmin.Application of BP network based on VB and Matlab in analysis of in-situstress field[J].Journal of Yangtze River Scientific Research Institute, 2009, 26(6):24-27. doi: 10.3969/j.issn.1001-5485.2009.06.007 [7] 谭文辉, 璩世杰, 毛市龙, 等.边坡爆破振动高程效应分析[J].岩土工程学报, 2010, 32(4):619-623. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=QK201001261659Tan Wenhui, Qu Shijie, Mao Shilong, et al.Altitude effect of blasting vibration in slopes[J].Chinese Journal of Geotechnical Engineering, 2010, 32(4):619-623. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=QK201001261659 [8] 粟闯, 刘敦文, 向海波, 等.基于BP神经网络不同输入变量的爆破振速预测[J].工业安全与环保, 2012, 38(2):52-56. http://d.old.wanfangdata.com.cn/Periodical/gyaqyfc201202018Su Chuang, Liu Dunwen, Xiang Haibo, et al.Prediction of blasting vibration by BP neural network based on different input variables[J].Industrial Safety and Environmental Protection, 2012, 38(2):52-56. http://d.old.wanfangdata.com.cn/Periodical/gyaqyfc201202018 [9] 赵华兵, 龙源, 刘好全, 等.城市隧道爆破振动速度峰值预测神经网络法应用[J].工程爆破, 2014, 20(5):23-27. doi: 10.3969/j.issn.1006-7051.2014.05.005Zhao Huabing, Long Yuan, Liu Haoquan, et al.Application of neural network to prediction of peak amplitude of blasting vibration velocity of urban tunnel[J].Engineering Blasting, 2014, 20(5):23-27. doi: 10.3969/j.issn.1006-7051.2014.05.005 [10] 谢涛, 何怡刚, 姚建刚, 等.VB调用Matlab的模拟电路故障诊断神经网络方法[J].计算机工程与应用, 2010, 46(10):213-215. doi: 10.3778/j.issn.1002-8331.2010.10.066Xie Tao, He Yigang, Yao Jiangang, et al.Method of analogous circuit fault diagnosis neural network based on VB transfers Matlab[J].Computer Engineering and Applications, 2010, 46(10):213-215. doi: 10.3778/j.issn.1002-8331.2010.10.066 -
图(5) / 表(2)
计量
- 文章访问数: 4676
- HTML全文浏览量: 1369
- PDF下载量: 652
- 被引次数: 0