A possible starting point at least from power systems perspective is power quality. Dealing with uncertainty and decision makers in a volatile environment is challenging.

I'd say the following topics:

METAHEURISTICS

1) Application of meta-heuristic based optimization algorithms to tune the PI controllers in Voltage source converters

2) Coordination of virtual inertia to improve the frequency stability using meta-heuristics.

3) Scheduling of Electric Vehicles to improve power quality using nature-inspired optimization algorithms.

REINFORCEMENT LEARNING

4) Application of reinforcement learning to design online volt/var controllers.

5) Use of deep reinforcement learning to develop better performance controllers to improve power system.

MACHINE LEARNING ALGORITHMS SUCH AS ANN, RANDOM FOREST, CNN,..

6) Load forecasting, stability assessment

Besides these, there are many other applications. I hope this list is useful for you.

Good luck

Lot of applications in AI in power electronics like MPPT tracking, adaptive controller design and optimization technique etc....

Thanks. It would be great to provide any resources/papers based on the applications that involves ML techniques such as supervised / Unsupervised learning, etc. Load forecasting and power quality is a good platform to start with.

I think it would be very interesting to have IA for power integrity purpose to present the almost best Power Distribution Network in every situation.

For example an ML trained using 3D solving which optimize component placement.

3D solving is long and fastidious to set up so most of decision are taken using rules of thumb which work but do not give always good result and need a lot of experience (EMC is not intuitive).

In consequence in that field AI may propose good optimization of that "rule of thumb" moreover it can be used by not experiment people.

I think before optimizing architecture, PDN optimization is a major issue.

I found a good paper that depicts Power Electronics 2.0. Do have a look.

In power system fault detection is a great area for machine learning applications along with the application of fault detection in electrical machine. As far as power electronics goes.. its finds application of electric drives, there is a book by bimal k bose.. called modern electric drives where you will find the application. The space vector pwm of inverter has also been dealt with in one of his papers.

Smart insulsted micro grids could be an intersting application of AI.

Article Artificial Intelligence Techniques for the Smart Grid

I studied Neuro-Fuzzy Control for Drive and Motion Control in 1995. There is updated information about that topic now. That could be a good place to start.

So far, Artificial Intelligence (AI) applications to power electronics and related areas are still limited, in part because some power electronics researchers do not trust the power performance of AI, and in part because they may not envision opportunities for its application.

A variety of application techniques for various power electronic problems, ranging from optimal design, control, reliability, energy management, to smart grids, and examining examples that have been successfully implemented. Finally, it highlights AI research trends and possible future applications to power electronics.

James Eric Hagerman Thanks for the novel start.

Hardiansyah Hardiansyah Machine learning learns from large data resources. This is necessary for model building. Is this process applicable to control design for power electronics simulations or equivalent applications as it requires more computing power to deal with big data?

The following documents show exemples of its application.

Artificial Intelligence Techniques for Smart Grid Applications

María-José Santofimia-Romero, Xavier del Toro-García, J. López-López

Published 2011 Engineering