olirsn / p2p_shiny_app

Shiny Peer-to-Peer electricity market Application

Antoine Rosin, Tiago Tousa and Pierre Pinson for DTU Elektro

The aim of the App is to raise the awareness of people about P2P electricity market. The user can visualize the exchange of energy between agents for various setups, and see the impact of individual preferences that reflect the willing to consume green or local.

Organization

The code is organized in four main folders:

• ui/ : containing all the ui*.R files that define the layout.
• server/ : containing the server*.R files that deal with the reactive variables.
• helper/ : containing the helper*.R files storing various functions needed to define the Leaflet map, solve the optmiization problem, analyze the results, ...
• starter/ : containing the starter_setup*.R files that defined the different setups used.

The app.R file gathered all the elements to lunch the app.

Setup file requirements

A setup file must define 3 variables:

• an agent_characteristic data frame containing for each agent:
  • LONG: the longitude of the agent.
  • LAT: the latitude of the agent.
  • NAME: the name of the agent.
  • GROUP: the group of the agent (village/community/group name).
  • TYPE: the type of the agent: 1 = consumer, 2 = conventional unit, 3 = wind turbine 4 = solar plant, 5 hydro.
• a prop_tot matrix containing for each agent:
  • first column: the minimum production capacity (if the agent is a consumer it will be negative).
  • second column: the maximum production capacity.
  • third column: the slope of the marginal cost/demand function.
  • fourth column: the origin of the marginal cost/demand function.
• a prop_emi vector containing for each agent the emission factor.

Important information

• Right now it is impossible to define more than two villages/communities/groups. The function 'setup_to_shiny' in the helper.R file has to be modified to allow it.
• If a new setup is added, one must also adapt the 'select_setup' selectInput button in the app.R file.
• Due to the RCI algorithm, solving the optimization problem can takes few seconds if most of the peferences are set to zero.

For any questions about the App contact Antoine Rosin at [email protected]