SYSTEM IDENTIFICATION IN HYDRAULIC SERVO SYSTEM WITH DIAGONAL RECURRENT NEURAL NETWORKS

摘要

This paper points out that the Diagonal Recurrent Neural Networks (DRNN) can deal with the dynamical system more effectively. We use this neural networks to identify the hydraulic servo system dynamical performance. The adjustment of weight is the algorithm that take time varied into account. The simulation results and experiments testified that this method could rapidly and exactly get the dynamical performance.