Loss measurement (efficiency measurement) is becoming increasingly important as motors become smaller and faster.

Currently, loss can only be obtained by subtracting the mechanical output from the input power, and the selection of the device used will become more important.
We will introduce measurement differences due to differences in power meters, temperature effects, and mechanical loss effects.

As motors become smaller and faster, the PWM frequency must be increased.
Torque ripple suppression has conventionally been achieved by applying current harmonics, but harmonic control becomes difficult at higher speeds.
Therefore, we introduce a method of torque ripple suppression (harmonic current control method) using 5,7 pulse voltage drive and a method of reducing magnet eddy current loss, which is a problem in the high-load region.

When controlling a recent high power density motor to meet high performance requirements, have you ever encountered behaviors that differ from the mathematical models shown in textbooks and papers?
In many cases, you may consider them as noise or errors and avoid them by filter processing because of the inverter drive, but this kind of filtering may have undesirable effects on controllability such as response and stability.

What we treat as noise or errors may be caused by things we ignored when considering the control target.
Even if there was no problem in the past, in order to meet higher performance requirements, it may be better to review the model of the control target that has been used in the past, properly model the things that have been ignored, and reflect them in the control system design.
And, improvements in measurement and computer performance are making this acceptable.
We will introduce examples of how noise and errors can be reduced by reviewing the model and control system design, using signal superimposed sensorless control and parameter identification as examples.