Prosumer flexibility scheduling

The codes and data provided here are used for the experiment simulations described in:

@article{mashlakov2021decentral,
  title={Decentralized cooperative scheduling of prosumer flexibility under forecast uncertainty},
  author={Mashlakov, A., Pournaras, E., Nardelli, P.H.J., and Honkapuro, S.},
  journal={Applied Energy},
  volume={290},
  pages={116706},
  year={2021},
  publisher={Elsevier}
}

Repository structure

• Data sets for forecasting and scheduling
• Prosumer flexibility modeling in R
• Flexibility coordination package in java
• Battery control simulation in python

Installation for Linux machine

$ mkdir flexibility-scheduling
$ cd flexibility-scheduling
$ git clone https://github.com/aleksei-mashlakov/flexibility-scheduling.git .

Data

Carbon intensity data is fetched from National Grid ESO

To get the carbon intensity data:

$ wget https://data.nationalgrideso.com/backend/dataset/88313ae5-94e4-4ddc-a790-593554d8c6b9/resource/f93d1835-75bc-43e5-84ad-12472b180a98/download/df_fuel_ckan.csv -P ./datasets

The preprocessing was done using:

$ python ./data_preprocessing/preprocess_carbon_intensity.py

Household net load data is fetched from Network Revolution project

To get the net load data:

$ wget http://www.networkrevolution.co.uk/go.php?id=409&link=TC5.zip -P ./datasets
$ wget http://www.networkrevolution.co.uk/go.php?id=409&link=TC2Auto.zip -P ./datasets
$ unzip TC5.zip TC2Auto.zip -d ./datasets

The preprocessing was done using:

$ python preprocess_net_load_data.py

Grid Frequency data is fetched from National Grid ESO

To get the frequency data for year 2019:

$ bash ./datasets/get_frequency.sh

Prosumer flexibility modeling in R

Prerequisites

1. Installed R (version>=3.6)

Instructions for linux are here or here

Ubuntu 18 in short:

$ sudo add-apt-repository 'deb https://cloud.r-project.org/bin/linux/ubuntu bionic-cran35/'
$ sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-keys E298A3A825C0D65DFD57CBB651716619E084DAB9
$ sudo apt-get update
$ sudo apt-get install r-base

2. (Optional) Installed Gurobi optimizer

Gurobi speeds up the simulation but for the simple example we use the other solver.

Get and activate academic license. Check instructions from here.

$ tar xvfz gurobi9.0.2_linux64.tar.gz
$ sudo mv gurobi.lic /opt/gurobi902/linux64

Consider this to run Gurobi in R. Users of the bash shell should add the following lines to their .bashrc files in /etc/bash:

$ export GUROBI_HOME="/opt/gurobi902/linux64"
$ export PATH="${PATH}:${GUROBI_HOME}/bin"
$ export LD_LIBRARY_PATH="${LD_LIBRARY_PATH}:${GUROBI_HOME}/lib"

3. Installed nohub

$ sudo apt-get install nohup

Flexibility forecasting

Forecasts quantiles for carbon intensity, household net load, and battery grid frequency response.

$ gunzip ./datasets/whole home power import 30 min 12 months averaged.csv.gz
$ nohup Rscript ./R/plan_forecasting.R &

Flexibility modeling

$ nohup Rscript ./R/plan_modeling.R &

Flexibility coordination package in java

Prerequisites

1. Installed java

$ sudo apt install openjdk-11-jre-default

2. Installed I-EPOS (it is already in the repository but the full instructions below)

To run 1 day example of flexibility coordination. For the rest days you need to run Flexibility modeling and place the results in ./I-EPOS/datasets.

$ cd ./I-EPOS
$ ./prepare.sh 1 1

Full I-EPOS instructions below:

$ wget https://github.com/epournaras/EPOS/releases/download/0.0.2/Release-0.0.2.zip -P ./I-EPOS/
$ sudo apt install unzip
$ unzip ./I-EPOS/Release-0.0.2.zip -d ./I-EPOS/
$ rm -rf ./I-EPOS/Release-0.0.2.zip
$ unzip ./I-EPOS/Release-0.0.2.zip -d ./I-EPOS/ && mv ./I-EPOS/Release-0.0.2/* ./I-EPOS/ && rm -rf ./I-EPOS/Release-0.0.2*
$ mv ./datasets/1/ ./I-EPOS/datasets/1
$ change number of agent to 150 and  ./I-EPOS/conf  
$ change ./I-EPOS/conf/epos.properties file
...............................
### Dataset ###
#The folder name in the datasets path. Make sure it has no spaces, tabs or newlines (alphanum and underscore preferred)
dataset=1


### Basic epos properties ###
# any integer > 0
numSimulations=50

# any integer > 0
numIterations=30

# any integer > 0
numAgents=150

# any integer > 0
numPlans=19

# any integer > 0
numChildren=2

# exact dimensionality from the dataset
planDim=48
..................................

Battery control simulation

The battery control simulations were conducted with the following file:

$ python ./Simulation/battery_control.py