Architecture - Pyosch/powertac-server GitHub Wiki

Deployment configurations

The server can be operated in two different deployment configurations: a research configuration, and a tournament configuration.

Server structure

Core entities of the Power TAC environment with their dependency relationships and responsibilities.

Server Communication

Communication between the Server and Broker agents is by exchange of "command" instances, packaged as payloads in messages transported via ApacheMQ.

Logging and Exception handling

What gets logged, and how are logfiles managed? What exceptions are we using, which ones are recoverable, how does the server recover from exceptions, and where are they raised and handled?

Visualizer architecture

Game visualizer overview, technology overview and description.

Time management

How simulation time is measured and managed.

Tariffs and Contracts

Tariffs are published and revoked by the brokers. Customers can subscribe or negotiate a published tariff. A subscribed tariff is a contract and lasts one or more days. As the accounting service is responsible for storing tariffs and logging contract relationships it acts as an intermediary entity between brokers and customers.

Controllable capacity

Also known as "Demand-Side Management." Customers may have some portion of their consumption, production, or storage capacity that can be remotely controlled by the DU for balancing purposes, or by a Broker for cost reasons.

Server configuration

How the server is configured, how configuration information is represented and communicated to the server.

Customer model structure

Structure and operation of customer models. There's also information on specific customer models.

R-language Broker

Ideas for implementing a broker framework in the R language.

Scenarios

A collection of operational scenarios to help clarify and understand how the server works.

Grid integration

Linking of Power TAC and a grid model