The preconditioned biconjugate gradient algorithm applied to geophysical electromagnetic modeling

Solution of the large system of equations that results from a finite-element approximation in electromagnetic modelling over a mesh strongly refined is most commonly done iteratively, because direct solution requires excessive amounts of memory and computation. The conjugate gradient – type methods used in combination with preconditioning – are among the most effective iterative procedures for solving large sparse nonsingular systems of linear equations Ax = b.