A New Non-linear AVO Inversion Method and Its Application in the Prediction of Carbonate Reservoirs

Abstract

Summary Non-linear method in AVO inversion is always used to correct the errors which are generated by conventional AVO inversion with initial model and estimated wavelet. Particle swarm optimization (PSO) is a common non-linear method used in non-linear AVO inversion, showing advantages including less artificial adjustable parameters, high precision and fast convergence speed. But not all the results of PSO inversion are accurate, especially for reflection coefficient of density. A non-linear AVO inversion method based on chaotic quantum particle swarm optimization (CQPSO) is proposed in this paper to improve accurateness. By comparison with conventional PSO, CQPSO has more efficient capability including shorter operation time, higher efficiency convergence, and global search capability, etc. Due to these available characteristics, CQPSO inversion could be used to obtain elastic properties directly from the synthetic seismogram, and provide more precise results, especially for reflection coefficient of density. After the testing with model data and the case study with seismic data, the results of CQPSO inversion are all coincident with drilling data in reservoir properties and fluid enrichment. These coincidences confirm that the new inversion method is feasible and effective.