Uma metodologia para o controle de seguranca dinamica de sistemas de energia eletrica

Abstract

This work presents an algorithm for the security control of electric power systems using control actions like generation reallocation, determined by sensitivity analysis (linearized model) and optimization by neural networks. The model is developed taking into account the dynamic network aspects. The preventive control methodology is developed by means of sensitivity analysis of the security margin related with the mechanical power of the system synchronous machines. The reallocation power in each machine is determined using neural networks. The neural network used in this work is of Hopfield type. These networks are dedicated electric circuits which simulate the constraint set and the objective function of an optimization problem. The advantage of using these networks is the higher speed in getting the solutions when compared to conventional optimization algorithms due to the great convergence rate of the process and the facility of the method parallelization. Then, the objectives are: formulate and investigate these networks implementations in determining. The generation reallocation in digital computers. Aiming to illustrate the proposed methodology an application considering a multi-machine system is presented.