微震震源定位算法优化研究及应用-《计算机技术与发展》

文章信息/Info

Title:: Optimization Research and Application for MicroseismicHypocentral Location

Author(s):: ZHAO De-ming; PANG Rui; WANG Hai-bo; Sinopec Geophysical Research Institute,Nanjing 211103,China

Keywords:: microseismic; hypocentral location; source scanning algorithm; quadratic interpolation; parallel optimization

摘要:: 微地震监测过程中,如何精准、实时地反演出微地震源的位置对生产应用起到了重要作用。国内外学者通过对微地震震源反演原理的研究提出很多反演方法,如纵横波时差法、模拟退火法、多个参数联合反演等,其中震源扫描叠加算法(source-scanning algorithm,SSA)由于其客观性和易用性得到广泛应用。通过研究微地震定位算法,发现震源扫描算法的计算速度有较大提升空间。以提升算法效率和现场应用的实时性需要为主要目的,提出利用二次插值方法减少算法计算量,利用 MapReduce 编程模型的并行化编程方法对变网格震源扫描算法进行了优化。实验结果对比发现,在一个八核CPU 的移动工作站上定位一段压裂的拾取事件剖面的时间可以控制在半个小时以内,速度相比原来的方法提高约 7 倍,达到了现场生产实时监测的要求。

Abstract:: How to reverse micro focal position accurately in real-time plays an important role in the production and application for micro-seismic monitoring process. Scholars at home and abroad put forward many inversion methods through the study of the principle of micro-seismic source inversion, such as time difference method of P-S wave, simulated annealing method, joint inversion of multiple parameters, etc. ,among which source scanning algorithm ( SSA) is widely used due to its objectivity and ease of use. Through the research on micro-seismic source location algorithm found that the computational speed of most algorithms,especially the source scanning algorithm, has a large space to improve. To enhance the efficiency of the algorithm and its real - time demand for application, the quadratic interpolation method is used to reduce the computational complexity of the algorithm,and the parallel programming method of MapReduce programming model is used to optimize the source scanning algorithm in locating method. The experiment proves that on a workstation with eight core CPU the location time of SSA algorithm handles a fracturing data can be controlled in less than half an hour,increased by about seven times compared to the original method,which meets the real-time monitoring requirements for scene? ?production.