How can I see the injected active and reactive power at a specific node?

[cqss11]

How can I see the injected active and reactive power at a specific node? I couldn't find it in the ANDES user manual and I only see the bus.a or.v.
Thanks for your reply.

[jinningwang]

We don't have variables to store injected power of each node. The reason is to reduce unnecessary memory consumption.
You can get the power injection with power flow equations with bus voltage and angle. Feel free to check the Example working with data.

[cqss11]

Thanks for your reply. And can I see the initial power of a certain node at a specific location in the code?

[jinningwang]

After calculating the injected power, you can get the entire time series data, where you can find the initial value.

[cuihantao]

Jinning, Do we have a documentation mentioning that line injections are the Line.a1.e, Line.a2.e, Line.v1.e, and Line.v2.e fields? For the original question, one can use a perturbation file to access these fields at the end of each step. Maybe you can help develop an example script. Get Outlook for iOS<https://aka.ms/o0ukef>

…

________________________________ From: Jinning Wang ***@***.***> Sent: Tuesday, August 1, 2023 3:02:42 PM To: CURENT/andes ***@***.***> Cc: Subscribed ***@***.***> Subject: Re: [CURENT/andes] How can I see the injected active and reactive power at a specific node? (Discussion #471) After calculating the injected power, you can get the entire time series data, where you can find the initial value. — Reply to this email directly, view it on GitHub<#471 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ABSNZAYG6PO5KNLR2R3URODXTFVHFANCNFSM6AAAAAA2YDEQQ4>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[jinningwang]

I don't see such discussion in the existing documentation and examples. I'll do the example script and make a PR recently.

[jinningwang]

An example is attached here, Power Injection.pdf

@cuihantao, would you kindly verify if I have correctly understood your idea?

[cuihantao]

Jinning,

This is clearly written. But why all the total P injections do not sum up to zero?

[jinningwang]

Updated one is attached, now the summed power is zero.

The reason is find_idx only returns first occurrence when multiple matches are there. For example, both Line_1 and Line_2 have "bus1" as "1", find_idx will only return Line_1.

[cuihantao]

Thanks a lot, Jinning.

I guess uses might also want to obtain these numbers at each step. They will then be able to obtain, for example, line injections following a disturbance. Can you also provide an example? You can use a pert file or come up with other methods.

[jinningwang]

Here, a further clarified example is attached, Power Injection.pdf

The pert.py code is excerpted below:

def pert(t, system):
    """
    Demonstration of perturbation function.

    This file should be applied to built-in IEEE 14-bus
    case "ieee14/ieee14_full.xlsx".

    In the given case, `Line` "Line_1" and "Line_2" are
    connected to `Bus` "1" through parameter `bus1`,
    and line injected active power is inspected through
    `Line.a1.e`.
    """
    Pinj_line = system.Line.get(src='a1', attr='e',
                                idx=['Line_1', 'Line_2'])
    print(f"t={t:.5f}, pinj_line={Pinj_line.sum():.5f}")

[MuaYoi]

Thanks a lot, Wang. I code a pert.py to calculate what I want like this:
def pert(t,system): Pinj_line = system.StaticLoad.get(src='bus',attr='e',idx=['PQ_1']) pass
But how can I use the variable Pinj_line outside the function pert?
And another question is how I can get timing data for a specific device in system.dae.ts.x or system.dae.ts.y. It seems that I am ablet to get only one kind of data like system.GENROU.Pe, but actually I only want the Pe of GENROU_1.

[jinningwang]

1. Good question. To the best of my knowledge, pert function is not designed to have any return value and the function call callpert does not provide any return interface, as shown below. Then I guess there is not a direct way to feedback the pinj_line value outside the pert function.
   

   andes/andes/routines/tds.py

   Line 386 in 6203731

   self.callpert(dae.t, system)

2. Actually example 2 in the documentation can answer this question, feel free to check it. https://docs.andes.app/en/latest/_examples/ex2.html#working-with-data

[MuaYoi]

Hi, Wang. I have read the section extracting variable data, in that example, the code ss.GENROU.omega returns an array [4,5,6,7]. I know that the array represents all GENROU's omega in system ss, but it seems that I can't work out which generator each element in the array corresponds to.

[jinningwang]

I believe in this section, https://docs.andes.app/en/latest/_examples/ex2.html#extracting-variable-data, desired variables number can be found, then the desired section of variables can be extracted.

[MuaYoi]

Hi Wang!
I am happy that I have found a way. If I want to find omega data of GENROU_1, I will use ps.GENROU.get(attr='a',src='omega',idx='GENROU_1') to get a index in ps.dae.ts.x, then I do what I want.
Thanks for your help and wish you a happy life!

[cuihantao]

@MuaYoi

But how can I use the variable Pinj_line outside the function pert?

It is unclear how "outside the function" means. Do you want to use it after the simulation is done, or do use the values at each step? You can provide at least some pseudo code for how you want to use it.

And another question is how I can get timing data for a specific device in system.dae.ts.x or system.dae.ts.y. It seems that I am ablet to get only one kind of data like system.GENROU.Pe, but actually I only want the Pe of GENROU_1.

The timestamps are the same for all the variables.

Like Jinning mentioned, https://docs.andes.app/en/latest/_examples/ex2.html#extracting-variable-data should give you some good ideas. In addition, the columns of ss.dae.ts.y[:, ss.GENROU.v.Pe] correspond to generators with indices ss.GENTOU.idx.v.

[MuaYoi]

Thanks for response Dr. Cui.

1. I want to use it after the simulation is done, just store Pinj_line of every step in an array, and I want to use the array to do something.
2. I have found a way. If I want to find omega data of GENROU_1, I will use ps.GENROU.get(attr='a',src='omega',idx='GENROU_1') to get a index in ps.dae.ts.x, then I do what I want.
   Thanks for your help and wish you a happy life!

[jinningwang]

For 1, a quick way would be store the desired data in a CSV file, as the pert function leaves nothing after calling. The drawback is I/O with hard-disk at every time step can significantly slow down the simulation.
A more hacker way can be revise the pert function and corresponding calling function in tds.py to leave a interface parameter that can be stored and returned after simulation.

[MuaYoi]

Thanks a lot, maybe I will try it.

[cuihantao]

Another easier way is to use Jupyter so that you have an interactive session. Next, in the pert function, duck tape the ss object by saving whatever you need as a field of ss. For example, ss.temp = data. When done, you can access ss.temp retrieve the data. This is a commonly used hacking technique in Python. Get Outlook for iOS<https://aka.ms/o0ukef>

…

________________________________ From: Jinning Wang ***@***.***> Sent: Thursday, November 30, 2023 7:57:31 PM To: CURENT/andes ***@***.***> Cc: Hantao Cui ***@***.***>; Mention ***@***.***> Subject: Re: [CURENT/andes] How can I see the injected active and reactive power at a specific node? (Discussion #471) For 1, a quick way would be store the desired data in a CSV file, as the pert function leaves nothing after calling. The drawback is I/O with hard-disk at every time step can significantly slow down the simulation. A more hacker way can be revise the pert function and corresponding calling function in tds.py to leave a interface parameter that can be stored and returned after simulation. — Reply to this email directly, view it on GitHub<#471 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ABSNZA6FPYDU5TO6CWRAGU3YHE2QXAVCNFSM6AAAAAA2YDEQQ6VHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM3TOMRUGA4DO>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[MuaYoi]

Thanks a lot. I will try it.