基于matlab的爆破震动分析及震动强度的智能预测-blasting vibration analysis and intelligent prediction of vibration intensity based on matlab.docx

摘要大规模的或频繁的爆破作业带来了一系列的严重问题包括爆破震动效应的危害，对 周围建筑设施和人员构成安全上的隐患，引起了人们的高度重视。本文以澳柯玛香海花 园小高层地下车库爆破开挖振动安全监测项目为依托，结合大量现场实测数据，对爆破 振动强度的各影响因素进行了分析、排序，对爆破地震波进行了时频分析，对爆破振动 强度进行了预测，主要表现在：首先，理论上分析了爆破地震波的产生机理、传播特点、危害及影响爆破地震强度 的各基本参数；然后运用灰色关联理论对影响爆破振动强度的各因素进行了排序，得出 在对爆破震动速度峰值、震动主频和震动持续时间的综合影响方面的顺序为：最大段药 量、距离和高度差，其中没有最优因素，最大段药量为准优因素。其次，通过对现场实测数据的最小二乘回归分析，得出了地下车库爆破开挖时经验 公式预测方法中的 K 、? 值，得出了开挖爆破地震波传播规律；运用小波分析理论，通 过 Matlab 编程实现了对实测数据的小波包分析，得出了地下车库开挖爆破地震波的频谱 特性和能量特性。最后，在理论上叙述了 BP 神经网络的特点，说明了用其预测爆破振动强度的可行 性，建立了爆破振动速度峰值预测的 BP 神经网络模型，并将其应用于工程实践，通过 与实测数据和回归预测结果的比较，得出 BP 神经网络预测方法优于回归预测方法。关键词：爆破地震波灰色关联理论最优因素回归分析小波包分析BP 神经网络模型ABSTRACTA series of serious incident included the hazard of the blasting seismic effects, caused by the large-scale or the frequent blasting operation, is caused people’s attention because they heavily threaten the around building facility and people. In this paper, based on the safety monitoring of underground garage excavation blasting in the AUCMA Xianghai garden little higher layer, combined with the large of field measured data, the various factorsinfluencing the blasting vibration intensity is analyzed and ordered, and the blasting seismic wave is analyzed in time and frequency, and the blasting vibration intensity is predicted. The main work is:Firstly, the mechanism of production, transmission characteristic and hazard of the blasting seismic wave, the various factors influencing the blasting vibration intensity are analyzed in the theory; then the various factor influencing the blasting vibration intensity are ordered by the way of grey correlation analysis, and the synthetically influencing factor order to the velocity peak value, the vibration dominant frequency, the time of vibrated duration is the maximal segment charge, the distance of measuring point and blasting source, the difference of the measuring point and the blas