Artificial Bee Colony(ABC) Algorithm On 2D Nuclear Magnetic Resonance

A wealth of formation parameters can be calculated by the inversion of 2-D NMR logging data and the nuclear magnetic parameters such as relaxation time which are based on the inversion calculation of multi-parameter spin-echo string measured.Compared with T2inversion,the most important is that Two-dimensional inversion has better performance in clearly reflecting the overlapping information by extending into two-dimensional direction.In this dissertation,by adding a regularization parameter to the inversion algorithm which includes TSVD,regularization and optimization,we transform the question into a optimization function.The dissertation realized 2-D?T₁-T₂?inversion through referring to other existed algorithms such as combining truncated singular value decomposition method,regularized inversion algorithm and improved artificial bee colony algorithm.Furthermore,the algorithm was verified and evaluated through the derivation of the principle and the analysis of the algorithm,which united with the corresponding numerical simulation experiment.In the end,the dissertation shows that the improved algorithm has high performance in reliability and practicability.