Finding an optimal well-log correlation sequence using coherence-weighted graphs
Yunzhi Shi, Xinming Wu, Sergey B. Fomel · 2017
We develop a workflow for sequential correlation of multiple well logs following an optimal path that preserves maximum coherency between neighboring log traces. In order to find an optimal sequence along multiple correlation pairs, we formulate this problem as solving the traveling salesman problem (TSP). The ”distances” between wells are defined as geometric distances weighted by log coherences measured by local similarity attribute, thus formulating a non-Euclidean graph. We adopt genetic algorithm (GA) to find the solution with near optimization in an efficiently feasible computation. Numerical examples demonstrate that the proposed method can achieve effective and efficient correlation of multiple well logs. Presentation Date: Tuesday, September 26, 2017 Start Time: 3:55 PM Location: 340A Presentation Type: ORAL