基于多智能算法融合模型的多级并行叠前分频波形反演研究

As one of the most active research directions of prestack seismic technology, prestack waveform inversion technique has important application value in the field of complex oil and gas exploration. Research on prestack gathers optimization technology, to improve the quality of the input data of prestack waveform inversion. Research on the high resolution spectrum decomposition technique of prestack gather data to obtain prestack frequency-dependent gather , the high quality prestack frequency-dependent gather will be as the input data of prestack waveform inversion, make full use of thin reservoir response difference of seismic wave in different frequency components, to improve the resolution of the prestack waveform inversion. Further study of multiple intelligence algorithms fusion optimization model, and to replace existed prestack waveform inversion based on simple genetic algorithm for improving the global search capability of the inversion algorithm, which can overcome the inversion trapped in local minimum defects, improve the inversion accuracy, at the same time, to combine with the high-performance multistage parallel computing, further improving the efficiency of the inversion calculation. The research of this project is not only the basic research on prestack seismic field, an important application of artificial intelligence and optimization method, to provide strong technical support for the shallow eastern China continental sedimentary oilfield in thin reservoir identification, onshore and offshore deep deep waters no or few wells of oil and gas exploration breakthrough that not only has important academic value, but also has significant social and economic benefits.

作为叠前地震技术最活跃的研究方向之一，叠前波形反演技术在复杂油气藏勘探领域具有极其重要的应用价值。本课题拟研究叠前道集优化处理技术，改善叠前波形反演输入数据的品质；研究叠前道集高分辨率谱分解技术，获得叠前分频道集数据，以高品质的叠前分频道集作为叠前波形反演的输入数据，充分利用薄储层对地震波不同频率成分的响应差异，来提高叠前波形反演的分辨率；研究多智能算法融合模型，替代现有叠前波形反演中所采用的经典启发式算法，提高反演算法的全局寻优能力，从而克服反演陷入局部极小值的缺陷，提高反演精度；研究反演算法的高性能多级并行计算模式，进一步提高反演的计算效率。该项目的研究不仅是叠前地震领域重要的基础性研究，也是人工智能优化方法的重要应用研究，为我国东部陆相沉积油田中浅层薄储层识别、陆上深层和海上深水域无井或少井区的油气勘探突破提供强有力的技术支撑，不仅有重要的学术价值，而且会产生重大的社会经济效益。

作为叠前地震技术最活跃的研究方向之一，叠前波形反演技术在复杂油气藏勘探领域具有极其重要的应用价值。本课题研究叠前道集优化处理技术，改善叠前波形反演输入数据的品质；研究叠前道集高分辨率谱分解技术，获得叠前分频道集数据，以高品质的叠前分频道集作为叠前波形反演的输入数据，充分利用薄储层对地震波不同频率成分的响应差异，来提高叠前波形反演的分辨率；研究多智能算法融合模型，替代现有叠前波形反演中所采用的经典启发式算法，提高反演算法的全局寻优能力，从而克服反演陷入局部极小值的缺陷，提高反演精度；研究反演算法的高性能多级并行计算模式，进一步提高反演的计算效率。该项目的研究不仅是叠前地震领域重要的基础性研究，也是人工智能优化方法的重要应用研究，为我国东部陆相沉积油田中浅层薄储层识别、陆上深层和海上深水域无井或少井区的油气勘探突破提供强有力的技术支撑，不仅有重要的学术价值，而且会产生重大的社会经济效益。