A trust-region method by active-set strategy for general nonlinear optimization

Abstract

The paper explores a trust-region active-set algorithm for general nonlinear optimization with nonlinear equality and inequality constraints. In this algorithm, an active-set strategy is used together with trust-region methods to compute the trial step. L 1 penalty functions are employed to obtain the global convergence. The global convergence of this algorithm is proved under standard conditions. The numerical tests show the efficiency of the proposed algorithm.