In model predictive control, building the correct
model and solving the optimal problem are two jobs that
always require lots of time. These are also two issues that
many scientists are interested in studying when applying
model-driven reporting control to certain objects. With a
TRMS object, we can build a white box model, a gray box
model or a black box model.
Some authors have built TRMS model published in
, , , . We have studied the solving methods of
optimal problem in model predictive control in , , .
In , we built a white box model of TRMS object
according to Newton method. Studying the effects of the
interchannel effects of the white box model TRMS. In ,
we used Gradient descent back-propagation, and some of
its conjugate algorithms to identify the TRMS model. In
this paper, we applicate the neural network in order to
identify black box model of twin rotor MIMO system
based on mean squared error method, using these results
compare to the real model so as to choose a suitable
algorithm and provide the ability to apply that model in
simulation and object control.
Keywords : Black Box Model, Neural Network, Yaw Angle, Pitch Angle, Identify, Mean Squared Error.