Optimum elimination of inadmissible operating states

This work contains a general analysis of an algorithm developed to determine a state of performance which guarantees the elimination of the network's operating states defined as inadmissible and optimizes the costs.The mathematical model formulated required an appropriate linear programming algorithm and a specific program to determine the sensitivity matrix.The experience obtained with the use of this linear method to solve the problem of optimum load flow shows that, compared with non-linear methods, it offers some advantages: on the one hand, the possibility of considering a greater number of constraints owing to the reduction of variables achieved by means of the sensitivity matrix and, on the other hand, the ability to detect incompatibility in the formulated constraints. In addition, this method offers greater guarantees of convergency, for the same quantity of functional constraints, than the nonlinear methods, because it does not use penalty functions.