Dynamic security border assessment of electric power system

The algorithm based on inversion of an artificial neural network with search of a unique element has been developed for estimation of the border of dynamic security. According to algorithm, the most nearest point to a current mode on border is found, and then its distance to the border is defined, allowing to judge security of a current mode. In case of need, the operator makes corresponding operating influences, withdrawing the system into more safe operating mode. The method of choice of an optimum set of entrance parameters of model of an estimation of security has been defined. Results of experimental check of work of algorithm on the example of the Komi power system are given.