The iterative algorithm described in section II can be
used to reduce the torque ripple at a number of harmonics
simultaneously. As outlined in section IV, the results shown
here are for a reduction of the 20th and 24th harmonics.
These harmonics were chosen as they are the primary
contributors to torque ripple and share characteristics with
other torque ripple harmonics.
Figure 3 shows the normalized magnitude of the 20th and
24th harmonics of the torque ripple measured by the torque
sensor for the first 10 steps of the algorithm when three
different types of sensors are used. Part (a) shows the results
when the torque sensor is used to provide data to the
algorithm. At step 1, the two harmonics are measured
without any compensation added to the torque input
( ˆ ( )
1 C h ) = 0 or all harmonics). At step 2, the magnitude of
both harmonics increase as they are now comprised of both
the original torque ripple (Tˆ(h) ) and that caused by an
injected signal ( ˆ ( )
2 C h ) which has an arbitrary but pre
defined magnitude and phase.