engleski

An Efficient Algorithm for Multiparametric Quadratic Programming

We present an efficient algorithm for multiparametric quadratic programming (mp-QP). As in [2], Karush Kuhn Tucker (KKT) conditions are used to characterize polyhedral critical regions and corresponding optimal solution. However, here we avoid unnecessary partitioning of the parameter space by using direct exploration strategy. Starting from the initial critical region we explore its neighborhood by crossing each of facets and checking if a feasible neighboring critical region exists. Procedure is then repeated in an iterative fashion with all newly generated regions.

multiparametric quadratic programming (mp-QP); KKT conditions; critical region; iterative procedure

nije evidentirano