基于蚁群算法的地震断层识别方法

A seismic fault recognition method based on ant colony optimization

研究了地震断层识别,针对现有断层识别方法准确度较低的问题,提出了一种新的基于蚁群算法的地震断层识别方法。此方法可以准确地定位断层的位置并且将它们提取出来。该方法首先将地震层位视为连通区域,并用连通标注方法将其提取出来;接着找出每个层位在水平方向的端点,然后将整个地震剖面划分成多个矩形块,将每个矩形块最上面的一个水平端点作为蚂蚁的巢穴,最下面的一个水平端点作为食物,在蚂蚁由巢穴出发寻找食物的过程中,将地震剖面视作三维地形,在三维地形中使用地震样点的幅度值作为某一位置的高度;最后在每一个矩形块中所搜寻到的从巢穴到食物的最短路径就是断层的位置。实验结果验证了所提方法的有效性。

The problem of seismic fault recognition is studied,and aiming at the fact that current recognition techniques＇ accuracy is lower,a new seismic fault recognition method based on ant colony optimization is proposed. The method can locate faults＇ position precisely and extract them from seismic sections. Firstly,the method regards seismic horizons as connected components and uses the way of labeling to extract them. Secondly,it finds out the end point in horizontal direction of each seismic horizon. Thirdly,the whole seismic section is divided into several rectangle blocks,taking one top horizontal end point of each rectangle block as the ant nest and one bottom horizontal endpoint of each rectangle block as the food in ant colony optimization. Besides that,the seismic profile is regarded as a three dimension terrain,and the seismic sample amplitude value is used to express the hight of some position.The optimal route in each block is the fault,in which the ant starts from the nest to find foods. The simulation certifies the proposed method＇s effectiveness.