Electrical systems are pretty interesting. They generate, transfer, and consume electric energy for our own needs in the industrial, transportation, and residential sectors. Considering electric energy generation, and excluding solar-powered plants which are still a minority, the rest of the electricity is obtained with generators, which are electric machines. Looking at the energy consumption side, almost half of the global electricity is consumed by electric motors, which are also electric machines. So all in all, electric machines are (and have been for a long time) a key element in our electrical systems.
If we look at the future, the industrial and transportation sectors are pushing towards even higher degrees of electrification, which will eventually end up in the addition of more electric machines and drives in industrial plants and electric vehicles of any type and size. At the same time, increased computational capabilities in microprocessors and the refinement of control algorithms enable further opportunities to reduce energy consumption and improve various performance metrics of electric drives. Electrical machines and drives are therefore a vital part of the future electrical ecosystem in the global quest for a sustainable society.
In the electric machines and drives (EMD) team, we search, analyze, and develop concepts and technologies aimed at enabling energy savings, improving dynamics, and reducing operational and maintenance costs of electric machinery, through their design and control. We perform theoretical analysis as well as experimental evaluation in our laboratory. We collaborate with a wide range of industrial and academic actors and are always interested in establishing new, collaborative initiatives.
Our perception of the future
We are not pretending to challenge the views of several experts in the field, but we are entitled to some personal views :-). Here are the main points that fuel our work.
Combining electric machine design and control
The days of electric machine design separated from electric drive development are soon to be over. Reality tells us that modern applications like renewable energy generation, electric traction, robotics, and others cannot survive without power electronics and control. Much is still to be gained in designing an electric machine+converter system as a whole, rather than the optimization of the single components.
Electric drives are very intelligent devices
Electric drives are extremely powerful devices in terms of computational capabilities, and they become better and better as the microprocessor/FPGA technology improves. Despite that, the typical control structures for electric machines do not typically require much computation. This means that there might be a lot of unused computation, which could be exploited for other functionalities, like real-time health condition monitoring of electric machines and drives, as well as for improving the real-time estimation of electrical parameters.
Estimating parameters is an evergreen
We are extreme fan(atic)s of automatic procedures for the estimation of electrical (and why not, mechanical) parameters in electric drive systems. The simple reason is that control algorithms require a series of information on the models which is typically not available in machine nameplate data. Someone, or rather something, has to retrieve that information for the control engineer.
It is very self-fulfilling when the converter can automatically extract all the relevant control parameters by performing autonomous injection of voltages in the electric machine and post-processing of the measured currents. All of this happens just by software, based on a few elementary nameplate data. Mind you, it must run without relying on the information on the speed and position of the rotor, as required by sensorless control algorithms.
As new electric machine topologies and designs are proposed and built all over the world, the challenge of making the automatic estimation procedures as general as possible, while maintaining their reliability, is still evident and inspirational.
Multi-phase machines are coming (and may have already passed you)
We lean towards the benefits of using multi-phase machines (where the number of phases is greater than three) in applications where it makes sense, and there are many. There is a lot to discuss, but what we like the most is:
1) Winding factors are generally better than corresponding three-phase machine designs, with space harmonics located at higher frequencies and with lower amplitude. This turns out into less current/torque ripples and fewer electric losses.
2) There is a possibility of implementing true fault tolerance, which means running at a reduced number of phases without being forced to stop the whole drive system. Be aware that this second point requires a combined research effort of electric machine design, converter design, health condition monitoring, control reconfiguration, and so on… separate scientific efforts are just not enough.
On reducing the costs…
Throughout the years, we have been listening to many research works and presentations aiming at “reducing the costs” of an electric machine, an electric drive, or both. What is your definition of cost? Ours is very close to this one. We do not consider any other definition worthwhile.
If you are interested in more specific activities running in the EMD team, have a look at our “Search and not research” page!