Combining Adjoint Calculations and Quasi-Newton Methods for Automatic HistoryMatching

摘要

This work describes the combination of adjoint methods forderivative calculation with a trust-region Quasi-Newtonoptimization algorithm for solving history matching problems.The techniques employed allowed for the resolution ofproblems with a larger number of parameters, when comparedto other approaches, based on derivative free optimization orthe forward method for derivative calculation.Applications to both synthetic and real cases will beshown. A substantial improvement of the match was obtainedin all examples. Problems with more than 250 parameters weresolved efficiently. The ability to handle problems with a largenumber of parameters is exploited to discuss the possibility ofaccessing the uncertainty in the history matching processthrough exausting the space of possible models.