ugr-sail / sinergym

Gym environment for building simulation and control using reinforcement learning
https://ugr-sail.github.io/sinergym/
MIT License
131 stars 35 forks source link

(v3.2.3) - Sinergym IncrementalWrapper #391

Closed AlejandroCN7 closed 8 months ago

AlejandroCN7 commented 8 months ago

Description

A wrapper to transform some of the continuous environment variables into actions indicating an increase/decrease in their current value, rather than directly setting the value. To compute the possible increments/decrements for each variable, dictionary is specified as an argument, indicating the name of each variable to be transformed as the key, and the value being a tuple of values called delta and step. This achieves a set of possible increments for each desired variable.

The following figure illustrates its operation, basically the values are rounded with nearest increment value and summed with current real values of simulation:

incremental_wrapper

Types of changes

Checklist:

Changelog: