微震P波到时拾取的PAI-k-MFV算法改进及应用

发布时间：2018-07-04 09:31

本文选题：微震 + P波到时拾取　；参考：《煤炭学报》2017年10期

【摘要】：为了充分发挥PAI-k-MFV算法参数自适应的优势,同时提高P波到时正确拾取率和拾取精度,通过引入STA/LTA算法和改进MFV准则,对PAI-k-MFV算法进行改进。首先,根据预先设置的时窗长度和置信概率的取值范围及步长,采用PAI-k算法拾取得到一组P波到时集;然后,运用STA/LTA算法初步计算P波到时区间,并将到时集中不在该区间内的到时元素剔除,得到更新后的到时集;最后,在新到时集的基础上,根据改进的最大频遇值(MMFV)准则最终确定P波到时。改进算法继承了PAI-k-MFV算法的参数自适应优点,无需人为调整太多参数,同时能够有效克服原MFV准则容易失效和拾取点滞后真实到时点等问题。对比其他算法,改进算法在拾取背景噪声段有强噪声的微震信号和弱微震信号P波到时具有明显优势。利用改进算法和PAI-k-MFV算法拾取冬瓜山铜矿115个实测低信噪比微震信号的P波到时,以人工拾取结果为参考标准,误差在5ms以内(对应30个采样点)视为正确拾取,结果显示:改进算法正确拾取率为96.52%,正确拾取的到时误差绝对值均值为1.1 ms,标准差为1.0 ms;较PAI-k-MFV算法,正确拾取率和拾取精度都大幅度提高。
[Abstract]:In order to give full play to the advantage of parameter adaptation of PAI-k-MFV algorithm and improve the correct pickup rate and precision of P-wave arrival, the PAI-k-MFV algorithm is improved by introducing the STA-LTA algorithm and improving MFV criterion. First of all, according to the range and step size of pre-set window length and confidence probability, a group of P-wave time sets are picked up by PAI-k algorithm, and then the P-wave arrival interval is preliminarily calculated by using the STA-LTA algorithm. Finally, on the basis of the new arrival set, the P-wave arrival time is finally determined according to the improved maximum frequency encounter (MMFV) criterion. The improved algorithm inherits the advantages of parameter adaptation of PAI-k-MFV algorithm, and it does not need to adjust too many parameters artificially, and it can effectively overcome the problems of the original MFV criterion which is easy to fail and the problem of picking point lag real time point. Compared with other algorithms, the improved algorithm has obvious advantages in picking up strong noise microseismic signal and weak microseismic signal P wave arrival time. The improved algorithm and PAI-k-MFV algorithm are used to pick up the P-wave time of 115 measured low-signal-to-noise ratio microseismic signals in Dongguashan Copper Mine. The result of manual pickup is taken as the reference standard, and the error within 5ms (corresponding to 30 sampling points) is regarded as correct pick-up. The results show that the correct pick-up rate of the improved algorithm is 96.52, the absolute value of the correct pick-up error is 1.1 Ms, the standard deviation is 1.0 Ms, and the correct pickup rate and precision are greatly improved compared with the PAI-k-MFV algorithm.
【作者单位】：中南大学资源与安全工程学院;中南大学数字矿山研究中心;
【基金】：国家重点研发计划资助项目(2017YEC0602905) 中南大学研究生自主探索创新资助项目(2017zzts568)
【分类号】：TD326;TP301.6

【相似文献】