A Combination Of Interior Point Method And SQP Method

Currently, SQP methods and interior point methods are the most popular method for solving nonlinear constrained optimization problems. These methods have advantages and disadvantages. From the numerical results and stability of computation, SQP methods for solving nonlinear constrained opti-mization problems are the most effective methods. But in every iteration, storage of Hessian and changes in active set have great impact on storage and computation of SQP methods. And the interior point methods also appear morbid cases near optimal solution. This thesis presents an algorithm based on identification of ac-tive set, combining with advantages of interior-point methods and SQP methods. We use interior point methods which perform better in early stages of the algorith-m, and turn to SQP method when we get a satisfying identification of active set, so there is a certain decrease in the number of iterations. The well performance of this algorithm in Matlab illustrates the effectiveness of this combination.