Torque ripple minimization in switched reluctance drives using self-learning techniques

The nonlinear torque-production mechanisms in the doubly salient, switched reluctance motor drive are both current and position dependent. It is shown that the shape of the static torque-angle-current characteristics of this drive can be fully determined by a series of measurements performed with the drive in a self-learning mode, without the need for an external loading device. These measurements consist of static tests, in which the torques produced by currents in different phases are balanced, and dynamic measurements, in which the relative currents required to produce the same torque at different positions are ascertained. The controller can then achieve very smooth low-speed performance by determining the current required to obtain the optimum torque contribution from each phase, at each rotor position.