The seismic nonlinear integrated prediction on the clastic reservoir in the periclinal area around Central Tarim

Based on genetic algorithms (GA) with fuzzy nerve network (ANFIS), the reservoir seismic nonlinear integrated prediction technology is composed of the seismic high resolution nonlinear inversion, the reservoir seismic nonlinear characteristic analysis, the reservoir nonlinear combined prediction and the validity evaluation and so on. This kind of geophysical technical series entirely attribute to the nonlinear system domain, characterized by high resolution and global optimization, distinguished from the conventional reservoir prediction technology, and pushes the seismic reservoir prediction and evaluation to semi-quantitative or quantitative development phase. Since the degree of exploration is low, the structure of the deep buried clastic reservoir is poorly developed and the internal reservoir is heterogeneous, attempt is made to apply the reservoir seismic nonlinear integrated prediction technology to clastic reservoirs in the peri-clinal area around Central Tarim and new progress and better application effect are achieved. Synthetic prediction of reservior indicates that the Donghe sandstone and the lower member of sandstone of Silurian are favourable reservior segments with validity value over 0.7. This method is of reference significance and provides a new approach and technique for the prediction and exploration of oil and gas in the area.