Research On Precondition AVO Pre-stack Inversion Method Based On Bayesian Theory

AVO pre-stack inversion method for extraction of formation rock elastic parameters of a kind of effective means in the field of oil and gas exploration play an important role, more difficult as the geophysical prospecting, puts forward higher requirements on oil and gas exploration technology, pathological features of AVO inversion problem itself is prone to multiple solutions, early AVO inversion method to maintain the stability of pathological inversion problem, often using two approximation equation inversion, introducing the theory of large error in the calculation process, affect the precision of inversion results, then the development of the three parameters of prestack AVO inversion method can provide more abundant reservoir information, is of great significance for concealment reservoir exploration.Three parameters increase, leading to pathological characteristics of AVO inversion is two inversion is more serious, in order to guarantee the stability and accuracy of the inversion results, need to make the necessary regularization constraint inversion problem, combined with the AVO pre-stack inversion algorithm of bayesian theory to a certain extent, improve the problem, by choosing appropriate prior distribution and the likelihood function of inversion problem, the stability of inversion problem and the balance of the accuracy of the inversion results effectively.Preconditioned regularization theory is against the conventional regularization method, and that by introducing a pre conditional operators to reconstruct the morbid problem, reduce the accumulation of theoretical error in the calculation process, improve the accuracy of the obtained results, the model tests show that pre conditions regularization algorithm is compared with the conventional iterative regularization algorithm is faster, higher stability, under the same number of iterations, the result is far higher than the conventional algorithm, for solving large-scale inversion problem has practical application value.Advantages of regularization method based on the preliminary conditions will embrace this bay leaf, AVO pre-stack method, forming a new inversion algorithm, is used in the objective function of the conditional operator for restructuring, the prior distribution function to choose the cauchy distribution of three variables, hold more constraint information, the distribution function and considering the statistical correlation between parameters, in order to further improve the accuracy of the inversion results, will also be smooth impedance and a certain number of wave impedance point constraints as constraint inversion equation and model test results show that based on bayesian theory condition of prestack AVO inversion algorithm iteration speed, cumulative error is small, the inversion process stability, get the final inversion result accuracy is higher, the actual data application results more reliable.