We explain a new penalty method recently introduced in the literature for solv-ing constrained optimization problems. In this method, the penalty parameter is adjusted dynamically at every iteration to ensure su?cient progress in linear feasi-bility. A trust region is used to assist in the determination of the penalty parameter, but not in the step computation. It is shown that the algorithm has global conver-gence. We implement the algorithm and test the program on a number of di?cult optimization problems. The numerical results con?rm the e?ectiveness of the algo-rithm
We explain a new penalty method recently introduced in the literature for solv-ing constrained optimization problems. In this method, the penalty parameter is adjusted dynamically at every iteration to ensure su?cient progress in linear feasi-bility. A trust region is used to assist in the determination of the penalty parameter, but not in the step computation. It is shown that the algorithm has global conver-gence. We implement the algorithm and test the program on a number of di?cult optimization problems. The numerical results con?rm the e?ectiveness of the algo-rithm