esa 0.5.3

A python package that makes PowerWorld Simauto easier yet more powerful to use

Easy SimAuto (ESA)

This Python package provides an easy to use and light-weight wrapper for interfacing with PowerWorld's Simulator Automation Server (SimAuto).

This file contains some motivation for the ESA package, a quick start, installation directions, common issues, and some examples/snippets.

Why Use ESA?

Directly interacting with PowerWorld via the Windows COM object can be quite cumbersome. Data type inputs and outputs can be odd, returns come back unlabeled, and you have to use pywin32 to interface with SimAuto.

ESA makes all these tasks quick and easy, is well documented, automagically translates data to the appropriate types, and uses Pandas DataFrames and Series where possible. For some motivating examples, please the the "Quick Start" section of this document.

Citation

If you use ESA in any of your work, please use the following citation:

@misc{ESA,
  author = {Zeyu Mao, Brandon Thayer and Yijing Liu},
  title = {Easy SimAuto (ESA)},
  year = {2019},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/mzy2240/ESA}},
  commit = {<copy + paste the specific commit you used here>}
}

Documentation

ESA is documented here.

Quick Start

The following quick start example uses the IEEE 14 bus case, which can be found in the repository or from Texas A&M.

You can find API documentation here.

# Import the SimAuto Wrapper (SAW)
>>> from esa import SAW

# Initialize SAW instance using 14 bus test case. Adapt path as needed
# for your file system.
>>> saw = SAW(FileName=r'C:\Users\myuser\git\ESA\tests\cases\ieee_14\IEEE 14 bus.pwb')

# Solve the power flow.
>>> saw.SolvePowerFlow()

# Retrieve power flow results for buses. This will return a Pandas 
# DataFrame to make your life easier.
>>> bus_data = saw.get_power_flow_results('bus')
>>> print(bus_data)
    BusNum BusName  BusPUVolt   BusAngle    BusNetMW  BusNetMVR
0        1   Bus 1   1.060000   0.000000  232.391691 -16.549389
1        2   Bus 2   1.045000  -4.982553   18.300001  30.855957
2        3   Bus 3   1.010000 -12.725027  -94.199997   6.074852
3        4   Bus 4   1.017672 -10.312829  -47.799999   3.900000
4        5   Bus 5   1.019515  -8.773799   -7.600000  -1.600000
5        6   Bus 6   1.070000 -14.220869  -11.200000   5.229700
6        7   Bus 7   1.061520 -13.359558    0.000000   0.000000
7        8   Bus 8   1.090000 -13.359571    0.000000  17.623067
8        9   Bus 9   1.055933 -14.938458  -29.499999   4.584888
9       10  Bus 10   1.050986 -15.097221   -9.000000  -5.800000
10      11  Bus 11   1.056907 -14.790552   -3.500000  -1.800000
11      12  Bus 12   1.055189 -15.075512   -6.100000  -1.600000
12      13  Bus 13   1.050383 -15.156196  -13.500001  -5.800000
13      14  Bus 14   1.035531 -16.033565  -14.900000  -5.000000

# Retrieve power flow results for generators.
>>> gen_data = saw.get_power_flow_results('gen')
>>> print(gen_data)
   BusNum GenID       GenMW     GenMVR
0       1     1  232.391691 -16.549389
1       2     1   40.000001  43.555957
2       3     1    0.000000  25.074852
3       6     1    0.000000  12.729700
4       8     1    0.000000  17.623067

# Let's change generator injections! But first, we need to know which 
# fields PowerWorld needs in order to identify generators. These fields
# are known as key fields.
>>> gen_key_fields = saw.get_key_fields_for_object_type('gen')
>>> print(gen_key_fields['internal_field_name'])
key_field_index
0    BusNum
1     GenID
Name: internal_field_name, dtype: object
>>> key_fields = gen_key_fields['internal_field_name'].tolist()
>>> print(key_fields)
['BusNum', 'GenID']

# Change generator active power injection at buses 3 and 8 via SimAuto
# function.
>>> params = key_fields + ['GenMW']
>>> values = [[3, '1', 30], [8, '1', 50]]
>>> saw.ChangeParametersMultipleElement(ObjectType='gen', ParamList=params, ValueList=values)

# Did it work? Spoiler: it does!
>>> new_gen_data = saw.GetParametersMultipleElement(ObjectType='gen', ParamList=params)
>>> print(new_gen_data)
   BusNum GenID       GenMW
0       1     1  232.391691
1       2     1   40.000001
2       3     1   30.000001
3       6     1    0.000000
4       8     1   50.000000

# It would seem the generator active power injections have changed. Let's 
# re-run the power flow and see if bus voltages and angles change. Spoiler:
# they do.
>>> saw.SolvePowerFlow()
>>> new_bus_data = saw.get_power_flow_results('bus')
>>> cols = ['BusPUVolt', 'BusAngle']
>>> print(bus_data[cols] - new_bus_data[cols])
       BusPUVolt   BusAngle
0   0.000000e+00   0.000000
1  -1.100000e-07  -2.015596
2  -5.700000e-07  -4.813164
3  -8.650700e-03  -3.920185
4  -7.207540e-03  -3.238592
5  -5.900000e-07  -4.586528
6  -4.628790e-03  -7.309167
7  -3.190000e-06 -11.655362
8  -7.189370e-03  -6.284631
9  -6.256150e-03  -5.987861
10 -3.514030e-03  -5.297895
11 -2.400800e-04  -4.709888
12 -1.351040e-03  -4.827348
13 -4.736110e-03  -5.662158

# Wouldn't it be easier if we could change parameters with a DataFrame?
# Wouldn't it be nice if we didn't have to manually check if our updates
# were respected? You're in luck!
#
# Create a copy of the gen_data DataFrame so that we can modify its 
# values and use it to update parameters in PowerWorld.
>>> gen_copy = gen_data.copy()
# Change generation at buses 2, 3 and 6.
>>> gen_copy.loc[gen_copy['BusNum'].isin([2, 3, 6]), 'GenMW'] = [0.0, 100.0, 100.0]
>>> print(gen_copy)
   BusNum GenID       GenMW     GenMVR
0       1     1  232.391691 -16.549389
1       2     1    0.000000  43.555957
2       3     1  100.000000  25.074852
3       6     1  100.000000  12.729700
4       8     1    0.000000  17.623067

# Use helper function to both command the generators and to confirm that
# PowerWorld respected the command. This is incredibly useful because
# if you directly use ChangeParametersMultipleElements, PowerWorld may
# unexpectedly not update the parameter you tried to change! If the 
# following does not raise an exception, we're in good shape (it doesn't)!
>>> saw.change_and_confirm_params_multiple_element(ObjectType='gen', command_df=gen_copy.drop('GenMVR', axis=1))

# Run the power flow and observe the change in generation at the slack
# bus (bus 1).
>>> saw.SolvePowerFlow()
>>> print(saw.get_power_flow_results('gen'))
   BusNum GenID       GenMW     GenMVR
0       1     1   62.128144  14.986289
1       2     1    0.000000  10.385347
2       3     1  100.000000   0.000000
3       6     1  100.000000  -3.893420
4       8     1    0.000000  17.399502

# What if we try to change generator voltage set points? Start by getting
# a DataFrame with the current settings. Remember to always access the
# key fields so that when we want to update parameters later PowerWorld
# knows how to find the generators.
>>> gen_v = saw.GetParametersMultipleElement('gen', key_fields + ['GenRegPUVolt'])
>>> print(gen_v)
   BusNum GenID  GenRegPUVolt
0       1     1      1.060000
1       2     1      1.045000
2       3     1      1.010001
3       6     1      1.070001
4       8     1      1.090003
>>> gen_v['GenRegPUVolt'] = 1.0
>>> print(gen_v)
   BusNum GenID  GenRegPUVolt
0       1     1           1.0
1       2     1           1.0
2       3     1           1.0
3       6     1           1.0
4       8     1           1.0
>>> saw.change_and_confirm_params_multiple_element('gen', gen_v)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "C:\Users\blthayer\git\gym-powerworld\venv\lib\site-packages\esa\saw.py", line 201, in change_and_confirm_params_multiple_element
    raise CommandNotRespectedError(m)
esa.saw.CommandNotRespectedError: After calling ChangeParametersMultipleElement, not all parameters were actually changed within PowerWorld. Try again with a different parameter (e.g. use GenVoltSet instead of GenRegPUVolt).

# So, PowerWorld didn't respect that command, but we've been saved from
# future confusion by the helper function.

# Let's call the LoadState SimAuto function.
>>> saw.LoadState()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "C:\Users\blthayer\git\gym-powerworld\venv\lib\site-packages\esa\saw.py", line 635, in LoadState
    raise NotImplementedError(NIE_MSG)
NotImplementedError: This method is either not complete or untested. We appreciate contributions, so if you would like to complete and test this method, please read contributing.md. If there is commented out code, you can uncomment it and re-install esa from source at your own risk.

# This behavior is expected - if we have not implemented/tested a SimAuto
# function, it will raise a NotImplementedError.

# Finally, make sure to clean up after yourself so you don't have COM
# objects hanging around.
saw.exit()

Pre-requisites

• Microsoft Windows 10 Operating System (PowerWorld is Windows only).
• PowerWorld Simulator with SimAuto add-on installed.
• Git Large File Storage (LFS) (OPTIONAL: required to download case files and run tests). After installing Git LFS, simply change directories to this repository, and run git lfs install. You will likely need to run a git pull or git lfs pull after installing and setting up Git LFS. After initial setup, you shouldn't need to do anything else with Git LFS.
• Python >=3.5.

Notes on Python versions

The authors of ESA have tested with Python 3.5, 3.6, 3.7, and 3.8. Many users may find it easiest to use Anaconda, but this is not recommended for users familiar with using Pip and/or virtual environments directly (or via PyCharm), as Anaconda provides an unnecessarily bloated installation.

Important Notes for PyCharm + Python 3.8

If you use PyCharm to automatically create virtual environments for you, there's a little extra work to do to get everything working for Python 3.8. Start up a terminal inside PyCharm (click on the Terminal button which defaults to the lower left area). In the terminal, run: python -m pip install -U --force-reinstall pip. Note you may need to run this command twice - mine failed the first time.

Installation

This section covers installation via Pip, installation from source, and some important post installation steps that must be taken.

Installation with Pip (easiest)

Use the Python package manager pip to install ESA:

pip install esa

Installation from Source

If you want to make modifications to ESA for your own purposes, you'll likely want to clone the Git repository and install from source. Fortunately, this is quite easy. The following directions will assume that your project is at C:\Users\myuser\git\myproject and that your virtual environment is at C:\Users\myuser\git\myproject\venv. If you're new to virtual environments, Python provides a nice tutorial.

Start by cloning ESA into C:\Users\myuser\git\ESA. This can be accomplished in a variety of ways, but perhaps the simplest is by using Git Bash:

cd ~/git
git clone https://github.com/mzy2240/ESA.git

After the cloning has completed, close Git Bash and open up a command prompt. Run the following:

cd C:\Users\myuser\git\myproject
venv\Scripts\activate.bat

Your prompt should change to be prefixed by (venv) indicating that your virtual environment has been activated. Now, perform the following:

cd ../ESA
python setup.py install

Post-Installation (optional)

You need to run a post-installation script related to a pre-requisite Python package, pywin32. As per pywin32's directions, you'll need to run the following with an elevated (administrator) command prompt after navigating to your virtual environment's directory:

Scripts\activate.bat
python Scripts/pywin32_postinstall.py -install

(this Scripts directory can be found within your virtual environment where your Python packages are installed. If you followed along in the "Installation from Source" example, this Scripts directory would be found at C:\Users\myuser\git\myproject\venv.)

License

MIT

Contributing

We welcome contributions! Please read out contributing.md.

Common Issues

This section will describes some (maybe) common issues and their solutions.

Installation/Virtual Environment Issues

If you have issues installing ESA and/or its dependencies, you may need to do some manual work installing pre-requisites in your virtual environment. Hopefully following these simple directions will help fix most issues.

1. Start fresh! Completely remove your virtual environment and recreate it. PyCharm makes this pretty easy, or you can do so manually using Python's guide. The remaining directions will assume you're typing commands into your activated virtual envrionment.
2. Reinstall pip: python -m pip install --upgrade --force-reinstall pip. We're intentionally using python -m pip instead of just pip to avoid possible path issues. Note that you might need to run this command twice (the first may fail for some unknown reason).
3. Check out ESA's setup.py file and look for install_requires. It'll look something like ['pandas', 'numpy', 'pywin32', 'pypiwin32'].
4. Using what we found under install_requires, install ESA's dependencies manually. To avoid dependencies on a compiler, we'll get binary distributions only: python -m pip install --upgrade --only-binary :all: pandas numpy pywin32 pypiwin32
   • If this command fails, you may need to pick and choose which dependencies you grab binary distributions for, and which you get other types of distributions for. Here's the Python documentation. If we only want to get binary distributions for pandas and numpy, we'd modify the previous command to instead read like so: python -m pip install --upgrade --only-binary pandas,numpy pandas numpy pywin32 pypiwin32
5. After you've installed ESA's dependencies, it's time to install ESA: python -m pip install esa

Errors/Issues Initializing a SAW Instance

This section will cover some common issues when attempting to initialize a SAW instance. The first thing to check is that your arguments are correct - you can find the documentation here.

esa.saw.PowerWorldError: OpenCase: Errors have occurred

You may see an error stack trace that looks something like the following:

Traceback (most recent call last):
  File "<input>", line 1, in <module>
  File "C:\Users\brand\git\ESA\esa\saw.py", line 111, in __init__
    self.OpenCase(FileName=FileName)
  File "C:\Users\brand\git\ESA\esa\saw.py", line 680, in OpenCase
    return self._call_simauto('OpenCase', self.pwb_file_path)
  File "C:\Users\brand\git\ESA\esa\saw.py", line 1101, in _call_simauto
    raise PowerWorldError(output[0])
esa.saw.PowerWorldError: OpenCase: Errors have occurred

Often, this is due to a bad path specification. Ensure you're providing a full file path, including the file extension (.pwb), and that the file exists at the exact path you specified.

TypeError: This COM object can not automate the makepy process - please run makepy manually for this object

If you see an error like the above, try initializing your SAW object again but set early_bind=False. While we're unsure of the root cause of this issue, it seems to be related to the fact that early_bind=True preemptively creates some Python files related to the SimAuto COM API.

Examples and Snippets

Add Lines to Model

This example shows how to add transmission lines to a model. You can find the case and .csv file referenced in the tests directory. This example will assume you execute this code within the repository at the top level (this level).

from esa import SAW
import pandas as pd
import os

# File with lines to add.
line_df = pd.read_csv(os.path.join('tests', 'data', 'CandidateLines.csv'))

# Fire up a SAW object. Ensure CreateIfNotFound is True so that we can
# use ChangeParametersMultipleElement to create new objects.
this_dir = os.path.dirname(os.path.abspath(__file__))
saw = SAW(FileName=os.path.join(this_dir, 'tests', 'cases', 'tx2000',
                                'tx2000_base.PWB'),
          CreateIfNotFound=True, early_bind=True)

# Rename columns to match PowerWorld variables.
line_df.rename(
    # TODO: Will need to update this renaming once
    #   https://github.com/mzy2240/ESA/issues/1#issue-525219427
    #   is addressed.
    columns={
        'From Number': 'BusNum',
        'To Number': 'BusNum:1',
        'Ckt': 'LineCircuit',
        'R': 'LineR',
        'X': 'LineX',
        'B': 'LineC',
        'Lim MVA A': 'LineAMVA'
    },
    inplace=True)

# We're required to set other limits too.
line_df['LineAMVA:1'] = 0.0
line_df['LineAMVA:2'] = 0.0

# Move into edit mode so we can add lines.
saw.RunScriptCommand("EnterMode(EDIT);")

# Create the lines.
saw.change_and_confirm_params_multiple_element(
    ObjectType='branch', command_df=line_df)

# Close the object so we don't get COM objects hanging around.
saw.exit()