You need to identify the broad area of engineering you are interested first perhaps.
If you are interested in power electronics I would suggest you think about Multi-point HVDC systems for windfarms etc. and how they might be modelled computationally and/or using physical scale models of some sort (perhaps even using mechanical models of some type).
It appears that generator nodes need to connect to the HVDC network via voltage multiplying DC-DC power converters (see the work of Professor Dragan Jovcic at the University of Aberdeen for example) Are there any alternatives to using power electronics for this?
You need to think about fault conditions and stability. How are faulty notes and faults between notes quickly identified and isolated to keep the rest of the system operational? Should you try to keep the power transmission voltage stable under fault conditions using distributed energy stored by the wind turbines?
It seems to me you will potentially need to define a safety critical digital comms layer to form part of the power transmission system to quickly determine where the fault is and how to best to quickly isolate it. How should this be specified and standardized?
What HVDC voltage is to be used for the standard? How will this effect the rest of the physical model?
When power generators connect or disconnect what are the instabilities that arise and how can this be addressed as part of the overall design?
Would some distributed energy storage be useful to help maintain stability under catastrophic failure conditions for example if transmission lines are suddenly fail or are cut etc?