A hybrid optimization approach for solving parameter estimation problems
Abstract
We present a hybrid optimization approach for solving parameter estimation problems that is based on the coupling of the Simultaneous Perturbation Stochastic Approximation (SPSA) and a globalized Newton-Krylov Interior Point (NKIP) algorithms. The approach generates a surrogate model using filtered data from SPSA that allows to use efficiently first order information and applies NKIP algorithm to find an optimal solution. We implement the hybrid optimization algorithm on four test cases and two medium-large scale applications from reservoir simulations, and present the numerical results obtained from this approach. Keywords. Parameter estimation, global optimization, local optimization, interior-point methods, surrogate models, large scale optimization.
Subject Area
Mathematics
Recommended Citation
Quintero, Carlos A, "A hybrid optimization approach for solving parameter estimation problems" (2007). ETD Collection for University of Texas, El Paso. AAI1445698.
https://scholarworks.utep.edu/dissertations/AAI1445698