From looking at books and academic papers, it seems to me most people tune internal voltage controllers in dc-dc converters assuming other paralleled converters in the network are tightly controlled already and hence tune the controller as if it was in a standalone network (maybe with a smaller bandwidth and irrespective of the other converters). The first controller might be tuned with a higher bandwidth to the second converter controller and so on. Am I mistaken? Any wise words/links in the right direction would be great, cheers.
Hi Daniel,
From what I know, you seem to be partly correct ! It is true that the tuning for individual converters would assume the voltage on the common bus to be "stable".
But I doubt if the wisdom behind such a tuning strategy is the existence of some hierarchically superior controllers ! Rather I would say that the common bus is assumed to have high capacity, or "inertia" towards voltage dynamics.
BTW, the type of dynamics expected on the AC side (if there is one !) of the specific DC-DC converter would also decide the DC side control.
(I am interested to know about other possible control strategies, if someone could please write about them.)
-Sanjay
Thanks for that Sanjay.
In terms of hierarchical control, a secondary controller can be employed to offset the voltage deviation caused by droop control. This secondary controller can be centralised, but suffers from single point failure and requires high bandwidth communications, which isn't ideal In a microgrid. Distributed secondary controllers are increasingly being used, emulating centralised control performance but with low bandwidth communications. Either way, the bandwidth of these controllers is a lot slower than the internal converter controllers' (therefore dictates the stability of the system more so), but these are tuned based on the internal converter dynamics and control, so it is important to get the internal tuning right.
In terms of internal converter control, I'm thinking people might be neglecting the dynamic interaction between the converters, as you plug in more converters to the network, though you might get away with it for a while, the right half plane zero inherent in boost etc converters is reducing, and therefore maybe at some point the system will eventually go bang (unstable) if these dynamics aren't some way accounted for in the internal control tuning.
Just a thought. Be interesting to hear what you and others think. Cheers.
Daniel
Yes Daniel,
I agree with you that one can make it all hierarchical by employing secondary controllers. It further may be a good idea in case of occurrence of faults.
But is it being done currently ? Or are you investigating a new possibility ??
Cheers !
The second paragraph is just a thought. I can't find any literature that explicitly describes the way internal controllers are tuned when paralleled with other converters in a microgrid for example since it's not the focal point for power electronics folk. I'm trying to understand how it's generally done. I'm guessing that controllers are tuned assuming other converters are tightly controlled, which I think you confirmed. Therefore they might not be accounting for the tuning of other controllers and a network with line resistances; this could be detrimental in a heterogeneous network. I have seen in the past where wind turbines were controlled irrespective of other turbines, with the second turbine maybe tuned slower than the first and so on, this was OK for 10 turbines, but once an 11th was connected to the point of common coupling, the system became unstable. Just wondering if dc-dc converters are indeed internally tuned irrespective of the tuning of other converters, then the same could happen. Currently investigating this.
And I will second you by saying that it is a good novel thought !
Will be interested to know how you evolve the idea, so do please keep me posted. And of course, anything I could help with by way of discussion is always there for you.
Good luck...
-Sanjay
- Badges
- Science topic