The linear traveltime interpolation(LTI) method is based on linear assumption so that its precision depends on grid division, and the rays transported from the reverse direction are not considered when we recalculate traveltime row by row or column by column,and the computation is unstable in the complex models. Here, we put forward two improvements for the LTI method. First, calculating the minimum travel-time of all grid nodes by all-round cycle strategy during ray tracing. Secondly, inserting secondary nodes in the original grid boundary. Model tests show that the all-round cycle strategy of LTI considers the races from all direction, so it's more stable for complex models. In the case of same grid division, comparing with the traditional LTI, the LTI method with secondary inserted nodes has higher precision that be improved at least an order of magnitude. At the same time, the calculation speed of our method is faster and time consuming decreases more obviously when the grids become smaller. The calculation speed of our method can be n -10n times than the traditional one's.%旅行时线性插值(LTI)射线追踪算法是基于线性假设的,在向前处理过程中仅用按行或按列扫描的方法来计算节点旅行时没有考虑逆向传播射线,导致其计算精度与网格剖分大小有关,在处理复杂介质时会使得追踪出来的射线路径不一定满足最短旅行时.因此,笔者提出了两点改进措施:在向前处理时需采用全方位循环的方法来计算节点最小旅行时；在网格边界加入次生节点.模型试算结果表明:采用全方位循环的LTI法考虑了来自各个方向的射线,可提高其对复杂模型的适应能力；在节点间距相同的情况下,网格边界插入次生节点的LTI法较传统的LTI法计算精度至少可以提高一个数量级,同时,计算速度也更快；随着节点间距剖分的越精细,计算耗时下降也越明显,计算速度较传统的方法可提高n～10n倍.
展开▼
