DIRECT TORQUE CONTROL OF INDUCTION MOTOR BASED ON ARTIFICIAL NEURAL NETWORKS WITH ESTIMATE AND REGULATION SPEED USING THE MRAS AND NEURAL PI CONTROLLER

[attachment=16958]

INTRODUCTION
The robustness, the low cost, the performances
and the ease of maintenance make the
asynchronous motor advantageous in many
industrial applications or general public. Joint
progress of the power electronics and numerical
electronics makes possible today to deal with the
axis control with variable speed in low power
applications [1]. Jointly with these technological
projections, the scientific community developed
various command approaches to master in real time
the flux and the torque of the electric machines.
One of the most recent steps in this direction is the
direct torque control-DTC, which provides
excellent properties of regulation without rotation
speed feedback. Proposed by I Takahashi and T
Noguchi and of Depenbrock, this method, appeared
in second half of the eighties, competing with the
methods of vectorial control.


PRINCIPLE OF DTC
The methods of direct torque control DTC
consist in controlling directly the opening or
closing the inverter switches from the computed
values of stator flux and torque. The state’s changes
of the switches are related to the evolution of the
electromagnetic state of the motor. They are no
longer controlled from the references of voltage and
frequency given to the adjusted control of an
inverter with pulse width modulation.


ANN BASED DIRECT TORQUE CONTROL

The neurons artificial network is a model of
calculation with a conception schematically
inspired by the real neurones functioning system.
Formal units, once assembled, help realize complex
information processing. It constitutes an approach
which gives more opportunities to approach the
problems of perception, memory, learning and
analysis under new angles. It is also a very
promising alternative to avoid certain limitations of
the classic numeric methods.


PI BASED NEURAL CONTROLLER

Also, in this section, we thought of replacing the
PI conventional speed regulator by a PI neuronal
with the objective of increasing the response time
period of the system, to optimize the performances
of the closed loop control in case diverse
disturbances would interfere in the regulation loop,
and to adjust the parameters of the regulator to
changes in the reference level