For the study on the relationshl＇p between the dynamic phenomena in the mining such as mine earthquakes, outburst and faults slide, firstly, double shear friction experiments of sandstone were made, and its slide cri...For the study on the relationshl＇p between the dynamic phenomena in the mining such as mine earthquakes, outburst and faults slide, firstly, double shear friction experiments of sandstone were made, and its slide criterion was suggested considering the viewing of engineering. Secondly, in order to study the stability of underground rock and zone of tectonic stress field, based on the analysis on distribution characteristic of initial rock stress measurements, the geology structural model was built and tectonic stress field was made a back-analysis by applying finite element method. The calculating results fit with the analysis result of earthquakes mechanism and the distribution characteristic of the measurements. The high stress regional centers station locates discontinuous zone of I level faults and is corresponding to underground earthquakes scene. From then it is certain that tectonic stress is the major origin and necessary condition of mine earthquakes. The instability slide of the faults is the main manifest and the mining activity is the leading factor. Beipiao fault has a dominate effect on other sub faults and tectonic stress area and is dynamical fountain of dynamic phenomena in the Beipiao Mines.展开更多
当三相异步电动机发生机械振动时,主回路中接触不良的电气接触点在振动作用下会产生串联型故障电弧,进而影响电路安全甚至引发电气火灾。而振动条件会加剧了故障电弧信号的复杂性,因此本文以回路电流信号为研究对象,提出了一种振动条件...当三相异步电动机发生机械振动时,主回路中接触不良的电气接触点在振动作用下会产生串联型故障电弧,进而影响电路安全甚至引发电气火灾。而振动条件会加剧了故障电弧信号的复杂性,因此本文以回路电流信号为研究对象,提出了一种振动条件下的高实时性串联型故障电弧检测方法。首先通过构建滑动记忆矩阵对实验电流数据进行动态保存,其次通过正交方向改进局部三值化模式(orthogonality direction local ternary pattern,OD-LTP)提取滑动记忆矩阵的纹理特征,最终将统计的OD-LTP图像的灰度分布直方图幅值作为特征向量,通过基于沙猫群优化(sand cat swarm optimization,SCSO)的支持向量机(support vector machine,SVM)建立振动串联型故障电弧检测模型。本文通过对比不同矩阵参数,得到最佳的滑动记忆矩阵尺寸,最终所提方法对故障电弧识别的准确率达到99.2%。通过对不同工况、不同特征提取方法对比分析,表明本文提出方法不仅适用于不同工况运行的工业电机变频器系统,其相对于其他特征提取方法也具有较高的实时性。展开更多
基金the Project of China National"973"Program(2005CB221501)National Natural Science Foundation of China(50474010)Research Fund for the Doctoral Program of Higher Education(20040147005)
文摘For the study on the relationshl＇p between the dynamic phenomena in the mining such as mine earthquakes, outburst and faults slide, firstly, double shear friction experiments of sandstone were made, and its slide criterion was suggested considering the viewing of engineering. Secondly, in order to study the stability of underground rock and zone of tectonic stress field, based on the analysis on distribution characteristic of initial rock stress measurements, the geology structural model was built and tectonic stress field was made a back-analysis by applying finite element method. The calculating results fit with the analysis result of earthquakes mechanism and the distribution characteristic of the measurements. The high stress regional centers station locates discontinuous zone of I level faults and is corresponding to underground earthquakes scene. From then it is certain that tectonic stress is the major origin and necessary condition of mine earthquakes. The instability slide of the faults is the main manifest and the mining activity is the leading factor. Beipiao fault has a dominate effect on other sub faults and tectonic stress area and is dynamical fountain of dynamic phenomena in the Beipiao Mines.
文摘当三相异步电动机发生机械振动时,主回路中接触不良的电气接触点在振动作用下会产生串联型故障电弧,进而影响电路安全甚至引发电气火灾。而振动条件会加剧了故障电弧信号的复杂性,因此本文以回路电流信号为研究对象,提出了一种振动条件下的高实时性串联型故障电弧检测方法。首先通过构建滑动记忆矩阵对实验电流数据进行动态保存,其次通过正交方向改进局部三值化模式(orthogonality direction local ternary pattern,OD-LTP)提取滑动记忆矩阵的纹理特征,最终将统计的OD-LTP图像的灰度分布直方图幅值作为特征向量,通过基于沙猫群优化(sand cat swarm optimization,SCSO)的支持向量机(support vector machine,SVM)建立振动串联型故障电弧检测模型。本文通过对比不同矩阵参数,得到最佳的滑动记忆矩阵尺寸,最终所提方法对故障电弧识别的准确率达到99.2%。通过对不同工况、不同特征提取方法对比分析,表明本文提出方法不仅适用于不同工况运行的工业电机变频器系统,其相对于其他特征提取方法也具有较高的实时性。
