Research of pre-stack AVO elastic parameter inversion problem based on hybrid genetic algorithm

摘要

AbstractPre-stack amplitude variation with offset (AVO) elastic parameter inversion technology, combined with genetic algorithm, provides a relatively effective identifying method to oil-gas exploration. However, many problems, such as, fast convergence in algorithm and being easy to fall into local optimization, are brought in traditional genetic algorithm, which leads to an unsatisfied inversion performance. Therefore, this essay proposes a hybrid genetic algorithm which is better in solving pre-stack AVO elastic parameter inversion problem. Taguchi thought is also introduced into this algorithm, which helps to produce better descents and to avoid falling into local optimization, and makes results more robustness. Additionally, as genetic algorithm is poor in local search and inversion of p-wave, s-wave and density, neighborhood search is adopted to optimize density inversion. Inversion accuracy is greatly improved.