A neural network identifier of synchronous machines trained by object oriented genetic algorithm and back propagation

An object oriented genetic algorithm was analyzed and designed by the object oriented methods in this paper. The object oriented genetic algorithm and backpropagation algorithm were combined together to design an evolutionary neural network identifier of saturated synchronous machines. The application of the object oriented genetic algorithm in the designing and training has demonstrated that this algorithm has a good generality and can be expanded conveniently by users. Results obtained from time-domain simulations were used to validate the trained neural network identifier. The capability of the neural network identifier to capture the nonlinear characteristics of the saturated synchronous machines was validated by the good agreement of the results of the identifier with the simulation results