摘要

In this paper, an adaptive trust region filter algorithm for unconstrained optimization is presented. The global convergence results of the algorithm are established and the numerical results show that the new algorithm is efficient in practical computation.