summer19

Paper (draft): https://realgjl.com/paper/summer19.pdf

This summer intern project was in fact a fork of my undergraduate thesis project realgjl/sfcNordic.

Acknowledgments: Prof. Petros Aristidou, Dr Sultan Alghamdi, et al.

Suummary

1. Built a communication layer on top of an existing Smart Grid simulator.
2. Designed physical and mathematical models (centralised secondary frequency control controller) in Python. Designed analytical models for testing and training results, such as the impacts of time delay and emergencies’ impacts on the stability of control (insensitive data analysis) in Python and MATLAB.
3. Collaborated with three senior researchers, and performed a comparative analysis with decentralised networks and optimized algorithm performance to ensure that the network can be restored to normal within 300 seconds after a cyber attack or massive power outage.

Installing Python 3 applications and libraries with Conda

Download the Miniconda installer

At a command prompt, enter (Mac):

curl -o ./Miniconda3-latest-MacOSX-x86_64.sh -k https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-x86_64.sh

In Linux (Ubuntu, CentOS, ...):

wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh

Run the installer

Mac:

bash Miniconda3-latest-MacOSX-x86_64.sh

Linux:

bash Miniconda3-latest-Linux-x86_64.sh

Setting the install directory

Follow the prompts on the installer screens. If you are unsure about any setting, accept the defaults. You can change them later. To make the changes take effect, close and then re-open your terminal window. To test your installation, in your terminal window or Anaconda Prompt, run the command conda list. For a successful installation, a list of installed packages appears.

Check if the Python has been installed successfully (version 3.x): python -V

For Mac users (macOS 10.15 Catalina, macOS 11.0 Big Sur, and later)

The default shell of terminal is now "zsh" instead of "bash".

According to the offcial doc from conda (search "zsh" in the webpage), we need to find the path/direction of miniconda3's folder, for instance, in my case:

/Users/realgjl/miniconda3

Then in the Terminal.app:

source /Users/realgjl/miniconda3/bin/activate

conda init zsh

Check the python version again and/or check if "conda" command works.

Suggestions: add path to the system

Go to the file "/etc/paths", and add the path of python binary folder (like "/Users/realgjl/miniconda3/bin") to it. Make sure this new miniconda directory is the first one, meaning that it will have precedence.

p.s., to show the hidden files in Mac, press: "command" + "shift" + ".".

Configuring Conda

Before first use, the conda package management system needs some initial configuration.

Make sure all the components are updated to their latest versions by entering:

conda update conda

at the command prompt. If there are any updates, you will be prompted to agree their installation.

Install libraries

Necessaries: pytorch (Deep Learning package) scikit-learn (general Machine Learning package) jupyterlab notebook scipy pandas matplotlib

pip install scipy numpy pandas matplotlib
pip3 install torch torchvision torchaudio
pip install jupyterlab

It's preferred to install scipy instead of independently installing numpy, mkl and other independencies. Scipy will automatically install packages like intel-openmp, mkl, mkl-service, numpy and blas.

To make sure that you have the MKL libraries installed, you can test via ipython:

ipython

import numpy as np
np.__config__.show()

"blas_mkl_info" should NOT shown as "not available", like:

mkl_info:
    libraries = ['mkl_rt', 'pthread']
    library_dirs = ['/home/home01/el17jg/miniconda3/lib']
    define_macros = [('SCIPY_MKL_H', None), ('HAVE_CBLAS', None)]
    include_dirs = ['/home/home01/el17jg/miniconda3/include']

pyramses

conda install -c apetros pyramses

Test if pyramses is in your path via ipython:

import pyramses
ram = pyramses.sim()

Alternatively, you can install pyramses via pypi.

intel

If the error "libifport.so.5: : cannot open shared object file" occurs, you may have to install Intel's redistributables package. You can review this intel's forum as a reference.

gnuplot

You _may need to install gnuplot although it is not necessary if you remote to a supercomputer. You can choose to use gnuplot or not in ~/settings.dat:

$CALL_GP F;
$CALL_GP T;
# $CALL_GP F;

GB system frequency range (nominal: 50 Hz, or 1 pu)

	Frequency Range (Hz)	Frequency Range (pu)
Overshoot/Max/Min	±0.200 = ±200m	±0.40% = ±(1/250)
Settling Frequency	±0.010 = ±10m	±0.02% = ±(1/5000)

GB: Thermal generators in Central Area (G13-G29)

TOR tor_ENTSOE_simp or TOR CONSTANT

Machine Type	Machine Code	Generator	SNOM (MW)
SYNC_MACH	COALG15.02	G15	8027.401
SYNC_MACH	CCGTG16.01	G16	7631.982
SYNC_MACH	COALG16.02	G16	3911.223
SYNC_MACH	COALG17.02	G17	2240.000
SYNC_MACH	CCGTG18.01	G18	883.047
SYNC_MACH	COALG18.02	G18	1356.953
SYNC_MACH	CCGTG19.01	G19	2186.816
SYNC_MACH	NUCLG20.04	G20	1500.000
SYNC_MACH	CCGTG21.01	G21	835.000
SYNC_MACH	CCGTG22.01	G22	512.000
SYNC_MACH	CCGTG23.01	G23	5716.446
SYNC_MACH	COALG23.02	G23	1604.416
SYNC_MACH	CCGTG25.01	G25	2176.000
SYNC_MACH	CCGTG26.01	G26	3964.273
SYNC_MACH	COALG26.02	G26	1601.727
SYNC_MACH	CCGTG28.01	G28	1475.995
SYNC_MACH	COALG28.02	G28	194.005
SYNC_MACH	COALG29.01	G29	643.685
SYNC_MACH	COALG29.02	G29	530.811

Frequency and Power