Torque Ripple Minimization in a Switched
Reluctance Drive by Neuro-fuzzy
Compensation

Simple power electronic drive circuit and fault tolerance of converter are specific advantages of SRM drives, but excessive torque ripple has limited its use to special applications. A new method for shaping the motor currents so that the ripple can be reduced is presented by this paper. Here, a compensating signal is added to the output of a PI
controller. This is added in a current regulated speed control loop.

Introduction
Switched reluctance machines have strong non-linear characteristics. Dynamic control of SR drives using artificial intelligence is hence important. where some degreen of torque ripple is tolerated such as in industries, fuzzy logic control has been implemented successfully. This is also significant in areas where smooth control is required at low speeds such as servo applications.

TORQUE PULSATION
it is not possible to obtain a ripple-free output speed with just pi-like control. it is more convenient to compensate for the torque pulsations through phase current waveshaping for operations at low speed.

PROPOSED METHOD
In this system, the output signal produced by the compensator is added to the PI controller's output signal, which, in steady atate, is ideally constant but producing significant ripple. after the addition, a current Icomp results . This Icomp is used as a
compensated reference signal for the current-controlled SR
drive converter. A neuro-fuzzy learning algorithm is used to adjusted the compensating signal iteratively. Here the training error recieved from the internal variables of the SR drive system is usef for ietrative control. Acceleration or speed ripple can be used as better variables for training.

Report:
[attachment=4367]