露天采矿爆破振动对民房破坏的旋转森林预测模型

针对爆破振动对民房破坏问题,应用因子分析及旋转森林理论,选取爆破振动幅值、主频率、主频率持续时间、灰缝强度、砖墙面积率、房屋高度、屋盖形式、圈梁构造柱、施工质量、场地条件为影响因素。建立基于因子分析的露天采矿爆破振动对民房破坏的旋转森林预测模型。利用现场实测32组爆破数据作为学习样本进行训练。采用回代估计方法进行检验,误判率为1/32。用另外12组爆破试验数据作为测试样本进行预测。结果表明,就爆破振动对民房破坏而言,灰缝强度、房屋高度、屋盖形式、圈梁构造柱及施工质量与其相关程度较高。用模型所得预测结果的误判率为1/6。

Rotation Forest Analysis Model and Its Application to Predicting Residential House's Damage against Blasting Vibration of Open Pit Mining

For the sake of studying the residential house's damage against blasting vibration,ten indexes,i.e.,blasting vibration amplitude,dominant frequency,dominant frequency duration,mortar seam intensity,the rate of brick walls,height of housing,roof form,the structural column of ring girth,the quality of construction and site conditions,were taken as analysis variables of the problem,based on the principles of factor analysis and rotation forest theory.A model based on the principles of factor analysis,was built for prediction of residential house's damage against blasting vibration of open pit mining.Training was carried out using 32 sets of measured blasting data as learning sample.The model was verified,the false rate was 1/32.Another 12 sets of measured data were tested as forecast samples.The results show that five indexes,i.e.,mortar seam intensity,height of housing,roof form,the structural column of ring girth,the quality of construction and site conditions,correlate significantly with the damage,and that the false rate was 1/6.