Set/alter command working issues in customized model

[PandaRP01]

HI Team

I am developing a customized model(i.e. TAPCON) to change the tap ratio of the Line model.
To change the tap ration of the Line model, I defined as follows
method-1

for ii in range(self.n):
            self.system.Line.set(src='itap', idx=self.rfidx.v[ii], attr='v', value=self.irtapp.v[ii])

or

Method-2

for ii in range(self.n):
            self.system.Line.alter(src='itap',idx=self.rfidx.v[ii], value=self.irtapp.v[ii])

In both the way, it is unable to change the itap in the Line model.

Also, when I run andes prepare -i, it is generating the code properly. Also, during tds simulation, the output of the TAPCON model, i..e irtapp is changing as expected.

Could you please highlight my mistake in this?

Regards
Rakesh Panda

[cuihantao]

Hi Rakesh, where did you place these code in the model? The models are mostly descriptive, and those in __init__() will not be executed in run time. You might need to use other hook functions like g_numeric()

[PandaRP01]

Hello Dr. Cui

Thank you for your response.

I have created a new folder inside models and placed my new customized model at thata folder. There I created init.py and added the model their.

Also, I added my model in init. py (inside models) and also group.py .

Could you please explain .. how to use g_numeric() ??

I will try to explore this function from my side too.

Regards
Rakesh panda

[cuihantao]

The folder and file organization do not matter for how the program is executed.

See here for an example. In g_numeric, you are updating the residuals by your custom function. The Jacobian will need to be updated accordingly.

andes/andes/core/block.py

Line 736 in 3adca81

class PIControllerNumeric(Block):

If you are only looking to adjust the tap as the program runs based on some logic, you may use the perturbation file. See https://docs.andes.app/en/latest/getting_started/formats/disturbance.html

[PandaRP01]

Hello Dr. Cui

Thank you for this help. It helps me to write my own g_numeric() and j_numeric().

However, when my J_numeric() function running, it is showing errors as follows:

TypeError                                 Traceback (most recent call last)
Cell In[10], line 7
      4 #ssPV.TDS.config.qrt = 1
      5 #ssPV.TDS.config.max_iter = 100
      6 ssPV.TDS.config.tf =50
----> 7 ssPV.TDS.run()

File ~\Andes_dev\andes\routines\tds.py:352, in TDS.run(self, no_summary, **kwargs)
    350 # only initializing at t=0 allows to continue when `run` is called again.
    351 if system.dae.t == 0:
--> 352     self.init()
    353 else:  # resume simulation
    354     self.init_resume()

File ~\Andes_dev\andes\routines\tds.py:209, in TDS.init(self)
    206 system.set_output_subidx(models=system.exist.pflow_tds)
    208 system.dae.clear_ts()
--> 209 system.store_sparse_pattern(models=system.exist.pflow_tds)
    210 system.store_adder_setter(models=system.exist.pflow_tds)
    211 system.store_no_check_init(models=system.exist.pflow_tds)

File ~\Andes_dev\andes\system.py:1161, in System.store_sparse_pattern(self, models)
   1159 for row, col, val in mdl.triplets.zip_ijv(jname):
   1160     print("in:",val,"row:", row,"col: ",col)
-> 1161     ii.extend(row)
   1162     jj.extend(col)
   1163     vv.extend(np.zeros_like(row))

TypeError: 'int' object is not iterable

How to overcome this error?
I have another question, could you please help me on "how existing ANDES model is updating J-numeric and g_numeric() function automatically without writing J-numeric() inside the model file."? How the linkage has been established to do this operation?