排序方式： 共有9条查询结果，搜索用时 15 毫秒
1
1.
波形分解是机载激光雷达全波形数据处理的重要基础工作,通过求解波形函数模型的参数,将波形数据利用具体的函数模型拟合出来,实现对全波形及其中各个子波形函数表达。LM(Levenberg-Marquardt)算法及其改进的算法是波形分解中对参数进行拟合求解的常用方法。针对LM算法在参数拟合计算的过程中存在大量迭代和矩阵运算,提出了基于线程块组和线程两级并行粒度的并行计算方案。将串行多次循环迭代求解参数改为单次并行计算取最佳值实现对参数的选择,将矩阵运算进行线程块的协同并行计算,实现了LM算法在通用计算图形处理器上的并行计算。实验证明,在规定阈值条件下,并行LM降低了算法的迭代次数,提高了波形分解LM算法的计算效率,为提高波形分解的处理效率提供了研究思路。 相似文献
2.
Richard McCreary John McGaughey Yves Potvin Dave Ecobichon Marty Hudyma Harald Kanduth Alain Coulombe 《Pure and Applied Geophysics》1992,139(3-4):349-373
Located in northern Québec, the Lac Shortt Mine was a small gold mine consisting of a thin subvertical orebody which was mined in three main phases. High stress and rockbursting conditions were experienced when ore was extracted in the upper zone between the surface and a depth of 500 metres during the first two phases of mining. Severe rockbursts were experienced in late 1989 near the shaft and in the footwall development following a deepening of the mine shaft to a depth of 830 m and partial development of footwall drift access for the third phase of mining (the mining of the lower zone starting at a depth of 830 m moving upward toward a depth of 500 m). A 16-channel Electrolab MP250 microseismic system, with a Queen's University Full-Waveform piggy-back system, was installed underground at the site due to these problems.It was expected that the thinning sill would be subjected to an ever-increasing load as the thickness of the 500 m sill pillar decreased in the face of the mining excavation from below. A monitoring program consisting of the microseismic monitoring system, a range of conventional geomechanics monitoring tools as well as the undertaking of periodic seismic tomography surveys to assess the ongoing state of stress and rock mass condition within the sill was therefore warranted.The anomalously high-magnitude stress field and the brittle rockmass created a situation in which rockmass failure was common and violent. In the creation and thinning of the sill pillar, the location of banded microseismic activity was crucial in tracing rockmass failure and the associated ground control problems. Reliable source-location determination enabled the identification of areas of stress increase. The movement of the rockmass failure front could be followed, and was responsible for stope dilution, footwall and orebody development deterioration, and caving.Source-mechanism analyses gave accurate double-couple solutions for approximately forty percent of these events having at least ten recognizable polarities. Results suggested movement along vertical north-south striking or vertical east-west striking features. Underground observation of damaged access points showed that vertical north-south striking joints were experiencing failure.The microseismic activity, which was consistently concentrated close to the southwest and northeast corners of current production stopes, could be explained by a stress field oriented obliquely to the strike of the orebody, as measured prior to shrinkage of the sill pillar byin situ stress measurements and observed borehole overbreaks. The orientations of theP andT axes for the microseismic activity further confirmed that the stress field oriented obliquely to strike.While an increase in compressional-wave velocity of 2.3 percent, corresponding to a measured stress increase of approximately 10 MPa could be measured by repeated tomographic surveys, it was relatively small and only a factor of two or so above the velocity measured uncertainty. The relative insensitivity of thein situ rock mass modulus to the applied stress is believed to be largely due to the rockmass discontinuities being relatively closed prior to stress increase, as substantiated by the small deformations seen by the extensometer and borehole camera. This situation existed because of the very high pre-mining stress level.The experimental demonstration that the rock could not absorb substantially increased load through the mechanism of discontinuity closure or tightening (which would be reflected in the modulus) may be evidence in itself of potentially burst-prone ground, such as encountered at Lac Shortt. 相似文献
3.
全波形激光雷达的回波中携带了被测目标的距离与特征信息,为了获取这些信息,本文提出了一种回波分解方法。本方法将原始的全波形回波分解为几个独立的高斯脉冲,并得到其函数表达式,从而提取出被测目标的距离等信息。分解过程中,首先,采用可变阈值的经验模态分解滤波法(EMD-soft)对原始波形进行滤波和噪声水平评估;其次,采用一套应对多种波形组成的初始参数估计方法,获取后续拟合所需的初始参数;最后,采用LM(Levenberg-Marquardt)优化算法对回波进行拟合优化,从而获取全波形回波中包含的独立高斯脉冲及其函数表达式。仿真波形的分解实验表明,分解误差在0.1 ns量级,换算成距离误差为15 mm,通过实验室自制的全波形激光雷达实验系统获取的回波的分解实验表明,分解的距离误差小于0.1 m。对比另外两种高斯分解方法对于相同仿真与实验数据的分解结果可以看出,本方法在分解成功率与精度上都有较大的提高。回波分解后的独立高斯脉冲中,除距离外还含有被测目标的反射率、粗糙度、面型等丰富的信息,回波分解方法作为回波分析的基础,将在遥感、测绘等生产与科研领域中发挥非常重要的作用。 相似文献
4.
分别利用近震波形反演求解震源机制解的gCAP方法和全波形反演方法反演了鲁甸MS6.5地震的震源机制解,并比较两种方法的优点与特性。结果发现,数据方位角的覆盖范围会影响解的稳定性,在使用gCAP方法时尤其需要注意;两种方法获取的震源机制解结果较为一致,走向为73°~77°,倾角为72°~85°,滑动角为157°~180°。 相似文献
5.
6.
当采用高斯滤波的方法对全波形激光雷达回波数据去噪处理时,高斯函数宽度和高斯滤波模板长度不同,去噪效果不同。本文通过设定不同的高斯函数宽度和高斯滤波模板长度,对回波信号进行去噪处理。实验结果表明,将激光雷达的发射脉冲半宽作为高斯函数宽度,高斯滤波模板长度设为9,对回波数信号进行去噪处理,滤波后的数据既保留了原有的信息,又得到了很好的平滑处理,取得了较为理想的去噪效果。 相似文献
7.
全波形反演方法是一个有效求解参数重建问题的方法,其本质是一个寻找最优解的优化问题,目前多用局部最优方法求解,如最速下降法、共轭梯度法、高斯-牛顿法、拟牛顿法等.文中给出了常用的优化方法,基于二雏声波方程,将这些方法应用于部分overthrust模型的反演,通过对各方法所得反演模型的精度和计算时间的对比分析,对各个方法的优缺点进行总结,为后续多参数反演或高维方程参数反演提供方法选择上的参考;针对所要求解的反问题,选用的优化方法需要在收敛速率、计算存储量和算法的稳定性之间进行权衡,以得到一个最优的反演结果. 相似文献
8.
LM(Levenberg-Marquardt)算法是全波形机载激光雷达(Li DAR)数据高斯分解中求解模型参数的一种方法。针对其结果在一定程度上依赖初值、雅克比矩阵出现非数值导致无结果等问题,本文提出分组LM算法,以广义高斯混合函数为模型,模型参数初始化后,将参数分组并使用LM算法依次对各组参数进行优化,并生成点云。为验证结果的可靠性,以系统点云为参考,与基于改进的EM(Expectation Maximum)算法全波形分解法做对比。结果表明,本方法不仅得到较高质量的点云,而且得到回波位置和宽度等信息。 相似文献
9.
全波形激光雷达的波形优化分解算法 总被引：1，自引：0，他引：1
随着数据存储能力和处理速度的提高,三维激光扫描系统逐渐具备全波形采集和分析技术。为了从全波形数据中获得脉冲时间、幅度、脉宽以及多回波分布等综合信息,波形分解成为了全波形激光雷达数据处理的关键技术之一。针对LM算法在一定程度上依赖初值,而传统激光雷达数据处理容易遗漏部分重叠的返回波,本文提出了一种改进回波分量初值设定的算法来获取回波脉冲的位置、宽度和强度。针对一套自主研发的全波形记录激光雷达演示系统进行了波形分解试验,定性和定量分析结果验证了该方法的有效性、可靠性和准确性。 相似文献
1