From GridLAB-D Wiki
Jump to: navigation, search

Jcfuller 18:13, 23 April 2013 (UTC)

"First, frequency and angle can be estimated directly from the bus complex voltages, considering a measurement first order lag." - You don't really estimate directly, do you? Don't you estimate from an initial value and the change in voltage, or some combination of previous knowledge?

Melizondo 18:33, 25 April 2013 (UTC)

Changed to "First, frequency and angle can be estimated by numerical differentiation of the bus voltage angles, considering a measurement first order lag."


Jcfuller 18:13, 23 April 2013 (UTC)

How do you plan to switch between delta and steady-state modes, e.g., will you take the final delta frequency and apply it across the steady-state regime, then recalculate again once you go back into delta? Will this only be done in delta? Will you calculate this in steady-state?

Melizondo 18:33, 25 April 2013 (UTC)

Added a section on "Super-second implementation and transition between super-second and sub-second"


Jcfuller 18:13, 23 April 2013 (UTC)

It seems like there is a possibility of divergent answers across a wide-spread network, i.e., after a significant number of timestamps you could have divergent frequency calculations across the network. Is there a way (or a need) to create/maintain/calculate a global frequency?

Melizondo 18:33, 25 April 2013 (UTC)

There will only be divergence when the system is unstable. In this case the simulation should stop and teh problem should be solved. See also the new section "Super-second implementation and transition between super-second and sub-second."


Mrice99 17:42, 25 April 2013 (UTC)

Background and Simple frequency measurement discussion needs to be tempered with a time step less than 1/60 seconds (0.0167 s). So, that change in Voltage Phase is calculated per same frequency zero crossing. Does it make since to say phase angle change 10(degrees) in minute can be used to calculate change in frequency?

Melizondo 21:21, 29 April 2013 (UTC)

Added the sentence "Note that the frequency is estimated using information of each time step of the sub-second simulations." in the simple freqeuncy measurement.


Mrice99 17:42, 25 April 2013 (UTC)

I understand that for PLL we know theta at bus p in gridlab, bus in hardware PLL the theta,p is not a value easily calculated. Could you expand to explain how in the real world measurements could create a theta,p.

Melizondo 21:21, 29 April 2013 (UTC)

In the PLL section, added the paragraph:"The PLL device creates a signal that has the same phase and frequency of a reference signal, the bus voltage in this case. The phase can be used as a reference for inverters to control their active and reactive power exchange with the network. A PLL model for phasor dynamic simulation is described here."


Mrice99 17:42, 25 April 2013 (UTC)

Echoing Jason… how does the values from sub second calculations impact super second “grid” frequency. (never mind it appears to have been corrected).

Melizondo 21:21, 29 April 2013 (UTC)

Added a section on "Super-second implementation and transition between super-second and sub-second"


Jcfuller 16:22, 01 May 2013 (UTC)

I really only had one other question, and I figured you already knew how to handle that – validation? Will validation be against other software or engineer eyeballs or a textbook sample or ?

Melizondo 17:41, 02 May 2013 (UTC)

Added a section on Validation


Ftuffner 00:00, 3 May 2013 (UTC)

You may want to define the validation a little better. Is there a specific system you'd like to validate against? How long of a simulation will it be on? Is there a particular frequency this system operates at? It just needs a little more detail.

Jcfuller 00:19, 3 May 2013 (UTC)
Yes, I'd a little more detail. For example, how would you use PSLF (a single-phase representation) to compare answers to a three-phase representation? What if your validation models / tools don't jive?