A new Algorithm for Equality- and Bound-Constrained DFO

Kurzfassung

This poster will present a new algorithm for equality- and bound-constrained nonlinear optimization without derivatives. A trust-region SQP framework is used to handle the constraints. The algorithm can be viewed as an extension of the algorithm BCDFO as it also applies the technique of self-correcting geometry and an active-set strategy to handle bound constraints. We present numerical results on a test set of equality-constrained problems from the CUTEr problem collection.