Studies on modified limited-memory BFGS method in full waveform inversion

Full waveform inversion (FWI) is a non-linear optimization problem based on full-wavefield modeling to obtain quantitative information of subsurface structure by minimizing the difference between observed seismic data and predicted wavefield. The limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) method an effective quasi-Newton in FWI due its high efficiency with low calculation storage requirements. Like other conventional methods, approximation Hessian matrix L-BFGS satisfies equation, which only exploits gradient model while available objective function value neglected. modified equation considers gradient, model, together. Theoretical analysis reveals that superior as it achieves higher-order accuracy approximating matrix. can be obtained using modify method. This modification improves little increase for each iteration. We incorporate into FWI, numerical results show has higher convergence rate, better results, stronger anti-noise ability than