The Application Of An Improved Genetic Algorithm In Nonlinear System Identification With Recurrent Neural Network

There are many defects if the BP network is used to identify nonlinear plants, so the recurrent neural network is presented to identify this plant. When optimizing the weights, the improved genetic algorithm is used in RNN, So the astrict of BP arithmetic can be removed. At last, an identification simulation examples for nonlinear dynamic plants validates the efficiency of the networks and the improved GA.