NEURAL NETWORK BASED IDENTIFICATION AND CONTROL OF DRIVE SYSTEM - PART I: HOPFIELD NEURAL NETWORK BASED LINEAR SYSTEM PARAMETERS' IDENTIFICATION IN CONSIDER OF SENSORS' CHARACTERISTICS AND ITS APPLICATION IN DC DRIVE SYSTEM PARAMETERS' IDENTIFICATION

摘要

The paper firstly discusses a Hopfield-Neural-Network (HNN) based linear system parameters' identify scheme under the assumption that HNN inputs are the detected system states delayed by sensors. Sufficient condition ofcorrect identification is derived. The scheme is used for identification of load torque (T_L) of DC drive system.Simulation results show that, even under bad workingcondition caused by incorrect setting to the controllers, thederived scheme guarantees correct identify results.