慢度-时间相关法与遗传算法结合提取阵列声波时差  

Estimating Slowness of Array Sonic Data with Slowness Time Coherence and Genetic Algorithm

作　　者：宋延杰[1] 何英伟[1] 石颖[1] 王雪萍[1] 

机构地区：[1]大庆石油学院,黑龙江大庆163318

出　　处：《测井技术》2006年第2期122-125,共4页Well Logging Technology

摘　　要：基于遗传算法的全局寻优特点和慢度-时间相关法提取时差的原理,提出了慢度-时间相关法与遗传算法结合的阵列声波测井资料处理方法。详细阐述了慢度-时间相关法与遗传算法结合求取声波时差方法的实现过程与关键操作步骤。慢度-时间相关法与遗传算法结合可将慢度-时间相关法中反映波形相似程度的相关函数作为遗传算法的目标函数,应用遗传算法寻找其最大值,得到与最大目标函数值对应的首波到达时间和慢度值。该方法不受首波到达时间步长与慢度步长大小的影响,并且只需要做部分计算就可以获得给定的时差、首波到达时间范围内的全局最优解。用该方法处理了大庆油田地区2口交叉式多极子阵列声波测井资料,并将处理结果与商业化软件eXpress处理结果进行了对比,两者具有较好的一致性,满足现场解释精度要求。Based on the global optimization characteristics of genetic algorithm and the principle of slowness calculating with slowness time coherence, a new integrated processing technique of array acoustic data is proposed with slowness time coherence and genetic algorithm in the paper. Detailed description of the integrated method and its key steps are given. The correlation function reflecting waveform similarity in slowness time coherence is used as the object function of genetic algorithm, then, maximum of object function is found with genetic algorithm, finally the wave arrival time and slowness values corresponding to the maximum value of object function are obtained. The integrated method is not affected by the steps of wave arrival time and slowness, and global optimization solution is found with less computation in the range of given slowness and wave arrival time. Crossed multiple array acoustic data processing of two wells in Daqing oilfield is made, and the processing result is compared with that of commercial software eXpress, and good agreement has been achieved. It shows the proposed method can be applied to slowness processing of array acoustic data with good accuracy.

关 键 词：遗传算法 慢度-时间相关法 相关函数 适应度 阵列声波时差 

分 类 号：P631.52[天文地球—地质矿产勘探]