STATE ESTIMATION BASED NEURAL NETWORK OBSERVER IN ELECTRIC DRIVE CONTROL SYSTEM OF OPTICAL-MECHANICAL COMPLEXES
Proposes an application of neural state observer using recurrent neural networks to estimate electromechanical variable coordinates in the tracking electric drive of the optical-mechanical complex control system. The mathematical description of the electric drive of optical-mechanical complex in the form of a two-mass electromechanical system with elastic connections is developed. These elastic vibrations of the two-mass system are damped by using feedback signals from the estimated elastic moment and load velocity from neural network observer. The architectures of dynamic recurrent neural networks according to the Elman scheme are analyzed in the form of state space model, which allows it to approximate a wide class of nonlinear dynamic systems. During computer simulation in the MATLAB/Simulink software environment, the comparison of the root-mean-square error between different learning algorithms for Elman's recurrent neural networks was carried out to study their accuracy estimates coordinates in a closed loop control system of optical-mechanical complex.
Authors: M. P. Belov, N. Van Lanh
Direction: Electrical Engineering
Keywords: Optical-mechanical complex, recurrent neural networks Elman, neural state observer, two-mass elastic system
View full article