CAJ | 학술논문

以福厦高速公路扩建工程中的大帽山小净距大跨度隧道爆破开挖为背景,应用BP神经网络,建立了小净距隧道掘进爆破振动和速度峰值的预报模型,并成功地应用于该隧道进、出口段掘进爆破振动控制工程实践.分析比较结果表明,BP神经网络预报方法的预测结果与实测值的平均误差为7.16%～10.61%,较传统预报方法的预测精度提高20%以上.研究成果为小净距大跨度隧道爆破振动控制提供了新的预测方法,具有理论及实用参考价值.
이복하고속공로확건공정중적대모산소정거대과도수도폭파개알위배경,응용BP신경망락,건립료소정거수도굴진폭파진동화속도봉치적예보모형,병성공지응용우해수도진、출구단굴진폭파진동공제공정실천.분석비교결과표명,BP신경망락예보방법적예측결과여실측치적평균오차위7.16%～10.61%,교전통예보방법적예측정도제고20%이상.연구성과위소정거대과도수도폭파진동공제제공료신적예측방법,구유이론급실용삼고개치.
Taking the project of Damaoshan great span and neighboring tunnel driving blasting in Fuxia highway expansion project as the background,used the back-propagation neural network,a prediction model of peak resultant velocity of tunnelling blasting vibration is established,and successfully applied to the engineering practice of controll blasting vibration in the import and export segments of the Damaoshan tunnel.Comparative analysis of the results show that the between back-propagation neural network method's forecasting result and actual value is 7.16%～10.61%,and 20% more accurate than traditional forecasting methods.