H.Mpharrami - Assitant prof.,Civil Eng., Tarbiat modares

This paper review a general Kuhn-Tucker-based Optimality Criteria methods for numerical solution of nonlinear optimization problems. Then it addresses the major deficiency of the algorithm which is lack of a robust solution procedure for identification of active constraints computations of its corresponding Lagrange multipliers. This problems that was supposed be inherent with optimality criteria method has been treated in different ways in the literature . In this paper it has been shown that the methods that have been used in the literature so far not free of error and mistake. Then a procedure that not have does not have precious drawbacks is proposed to remedy the problem and make the OC method a powerful optimization tool. The proposed procedure consists of inverting a phase one LP problem to a quadratic programming (QP) sub-problem from which Lagrange multipliers can be evaluated . One example has been considered for demonstrating the robustness of the proposed procedure.

Optimality Criteria , Largrange multioliers , QP subproblem