Application of Genetic Diagonal Recurrent Neural Network to Servo System

The learning rate and the structure of the diagonal recurrent neural network (DRNN) are selected by the experiential method. The core of the genetic algorithm DRNN (DRNN) is the optimization of the DRNN learning rate and structure by the genetic algorithm. This optimized results composes new neural network. At the same time, the detailed step is abstained. Finally, this GA-DRNN is applied to the design of the servo system controller, Simulation results show that compared with controller based on DRNN, the controller based on GA-DRNN possesses higher precision.