Pre-stack amplitude-versus-offset (AVO) inversion is important for determining the physical parameters of reservoirs.The conventional AVO inversion method,based on approximate Zoeppritz equations,has high computation efficiency and practicability.However,the method assumes a small difference in elastic parameters between the upper and lower medium.Additionally,it has certain application limitations when used simultaneously with multi-parameter inversion in the presence of a changing background medium.To address these shortcomings,we propose a method of AVO inversion based on exact Zoeppritz equations and an improved artificial bee colony algorithm.Using Bayesian inference,the objective inversion function is derived based on exact Zoeppritz equations.The improved artificial bee colony algorithm is incorporated into AVO inversion to achieve stable P-wave velocity,S-wave velocity,and density inversion.Firstly,using the particle swarm algorithm for reference,the adjustment coefficient is introduced,and the binomial crossover is combined with the artificial bee colony algorithm to form an algorithm guided by the global optimal solution.Secondly,the relationship between the fitness function and objective function is established,and AVO inversion is combined with the bee colony algorithm.Synthetic and field data from the east central uplift belt of Dongying Sag in the Jiyang Depression are used to assess the effectiveness and feasibility of this combined method.Results show that it overcomes the limitations of the conventional,linear AVO inversion method and has good stability and convergence precision.This inversion method can be used to better estimate the physical parameters and characterize the seismic distribution characteristics of reservoirs.
论文详情