摘要
为提高冲击地压危险预测模型的准确性及精度,选用ROCK 600-50型岩石三轴流变仪与声发射系统对砂岩进行常规三轴实验。采集岩石的三轴抗压强度、声发射信号强度、声发射绝对能量等多种信号作为评估指标,建立SSA优化BP神经网络对所选数据进行优化和处理,进而设计出以模糊物元分析为基础的冲击地压危险预测模型,最后建立模糊综合评价模型。通过仿真分析验证,结果表明,砂岩试件的各项指标与岩石破裂过程所经历的四种阶段呈对应关系。根据上述选取的特征指标将冲击地压划分成4个危险等级,可为冲击地压预测提供更加精准的依据。
In order to improve the accuracy and precision of rock burst hazard prediction model, ROCK 600-50 rock triaxial rheological apparatus and acoustic emission system were selected to carry out the conventional triaxial experiment on the sandstone with bursting liability. Triaxial compressive strength, acoustic emission signal strength, absolute energy of acoustic emission,and other signals in the rock failure process were collected as evaluation indexes. The SSA-optimized BP neural network was established to optimize the selected data, and then the rockburst hazard prediction model based on fuzzy matter-element analysis is designed, and finally the fuzzy comprehensive evaluation model was established. Simulation suggest that the indexes of sandstone specimens correspond to the four stages of rock fracture process. According to the characteristic indexes selected above, the rock burst is divided into four risk levels. It can provide more accurate basis for rock burst prediction.
作者
梁燕华
毛诗允
刘刚
LIANG Yanhua;MAO Shiyun;LIU Gang(School of Electrical and Control Engineering Heilongjiang University of Science and Technology,Harbin 150022,China;School of Mining Engineering Heilongjiang University of Science and Technology,Harbin 150022,China)
出处
《电声技术》
2022年第10期124-127,131,共5页
Audio Engineering
基金
基于模糊物元分析的冲击地压危险等级评价(LH2019E084)。
关键词
冲击地压
模糊物元分析
声发射
神经网络
rock burst
fuzzy matter-element analysis
acoustic emission
neural network
