Electric Machine Control for Energy Efficient Electric Drive Systems

Doctoral thesis, 2018

Over the past decade, electric vehicles has increasingly become an area of attention for both academia and industry, much due to challenges such as emissions legislation and the environmental impact of the transportation sector. The absence of the broadband noise from the internal combustion engine brings new acoustic challenges for electric propulsion applications. Magnetic noise from electrical machines is of particular interest in automotive applications. It is not only related to the geometrical design of the machine, but also to the selection of control approach and voltage modulation strategy.

This thesis focuses primarily on software-based electric drive system energy efficiency enhancements, supported by extensive experimental testing, incorporating aspects of dynamic performance and acoustic perspectives. The scientific contribution can be summarized in three parts. Firstly, the interdisciplinary research where efficiency enhancements are coupled to acoustic performance. Secondly, the cause and effect of electromagnetic forces as the link between machine design, controls, and perceived acoustic annoyance. Lastly, the findings from the research on optimization-based inverter control and motion sensorless operation.

It is proven that alternative modulation techniques can reduce the inverter losses with up to 15 % without degradation of the perceived acoustic annoyance. Moreover, research on finite control set model predictive current control and moving horizon rotor position estimation is presented. It is shown that the proposed solutions are feasible, and that the associated optimization problems at hand can be solved in real-time while exploiting their respective attractive properties. Furthermore, excellent performance is obtained in comparison to state of the art alternatives, at the expense of increased computational complexity.