Closed lhbueno closed 7 years ago
The limitation of apparent power is not physical. The limit of the converter is the current limit of the semiconductors. So in case of an undervoltage the current could become a lot larger in the simulation than in a real application where the current is limited by control. I thus do not want to implement your actual proposal.
The actual implementation of power converters in the PowerConverters library relies on the idea, that the DC voltage is adjusted without any limit. The AC active power has to be equal to the DC power due to power conservation in the converter. Consequently there is no control over the DC power in the converter. What current limit shall then be applied?
I propose to implement a physical limit of a maximum AC current. As long as the phase angle is big enough you can reduce the phase angle to maintain (active) power conservation. If phase angle control is not sufficient any more as DC power is too big, you have to trigger an assert to terminate the simulation as you have no other control over the current which is physically meaningful. In case you want to make an additional proposal I suggest that you encapsulate the limitation of the current to a separate block to make the actual converter model more structured, as in the current proposal it is quite difficult to understand what the actual functionality is.
Please consider the example provided in https://gist.github.com/christiankral/ca370b3dd910d5ad0fff46dab12a84e0 which is created with your pull request version:
I do not want to insist on the implementation of the current limiter by control. However, I corrected some issues of the previously model and I divided into a different block.
The current limiter is calculated based on the parameters that are given to the converter. Is_max = (Smax VmpRef ImpRef / VRef) * sqrt(2/3), where Smax is given in pu.
The active current (Id) is limited by Is_max. The reactive current (Iq) is limited by sqrt( (Is_max^2) - (Id_lim^2) ).
Therefore the AC active power will be equal to the DC power (as long as the DC power does not overcome the set AC apparent power) and the AC reactive power will be delivered as much as the AC apparent power permit (reducing the phase angle if necessary)
You can find this changes at: https://gist.github.com/luizhrbueno/8baa3bd647043f58a15c8984f0536e73
or in the branch: https://github.com/luizhrbueno/PhotoVoltaics/tree/Implement_AC_current_limitation_of_multi_phase_converter
In this implementation power balance is not maintained when Smax (and Imax respectively) is reached. I am sorry, but powerAC shows non physical behavor. I am thus not going to implement your proposal.
On 2017-03-07 12:24, Luiz Bueno wrote:
I do not want to insist on the implementation of the current limiter by control. However, I corrected some issues of the previously model and I divided into a different block.
The current limiter is calculated based on the parameters that are given to the converter. Is_max = (Smax VmpRef ImpRef / VRef) * sqrt(2/3), where Smax is given in pu.
The active current (Id) is limited by Is_max. The reactive current (Iq) is limited by sqrt( (Is_max^2) - (Id_lim^2) ).
Therefore the AC active power will be equal to the DC power (as long as the DC power does not overcome the set AC apparent power) and the AC reactive power will be delivered as much as the AC apparent power permit (reducing the phase angle if necessary)
You can find this changes at: https://gist.github.com/luizhrbueno/8baa3bd647043f58a15c8984f0536e73
or in the branch: https://github.com/luizhrbueno/PhotoVoltaics/tree/Implement_AC_current_limitation_of_multi_phase_converter
— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/christiankral/PhotoVoltaics/pull/20#issuecomment-284695289, or mute the thread https://github.com/notifications/unsubscribe-auth/AD_Ymk-RZ3EjFDMpXZql5yaX-8auhXC1ks5rjT5WgaJpZM4MQ0on.
addition of output power limitation in the converter.