Skip to content

ochre_gym

parse_dwelling_args

ochre_gym.parse_dwelling_args(dwelling_dict: Dict, building_path: Dict)

Parse the dwelling arguments.

Parameters:

Name Type Description Default
dwelling_dict Dict

Dictionary of dwelling arguments.

required
building_path Dict

Path to the building.

required

Returns:

Name Type Description
dwelling_dict Dict

Dictionary of dwelling arguments.

load

ochre_gym.load(env_name: str, **kwargs: str)

Load an ochre_gym Env by name. Override default arguments with kwargs. Default arguments can be found in the defaults.toml file in ochre_gym/buildings.

Parameters:

Name Type Description Default
env_name str

Name of the building environment.

required

Other Parameters:

Name Type Description
seed int

The random seed used to initialize the environment.

start_time str

Time to start controlling the building. Format: "YYYY-MM-DD HH:MM:SS".

time_res str

Time resolution for OCHRE. String. Format: "HH:MM".

end_time str

Time to stop controlling the building. Format: "YYYY-MM-DD HH:MM:SS".

episode_duration str

Amount of time to run each episode for. Number of episode steps is episode_duration / time_res. Ignored if end_time is specified. Format: "X days" or "X hours", where X > 0.

initialization_time str

Amount of time that OCHRE should run before starting to control the building, to initialize the physical building states. Format: "X days" or "X hours", where X > 0.

hpxml_file str

Path to the building properties file.

schedule_input_file str

Path to the building schedule file.

weather_file str

Path to the building weather file.

from_beopt bool

Whether using a BEopt building model.

verbosity int

OCHRE internal verbosity level. Default: 7.

log_to_file bool

Log to file. Default: False.

log_to_console bool

Log to console. Default: False.

log_output_filepath str

Path to a log file. Default: './ochre_gym.log'.

disable_uncontrollable_loads bool

Disable load due to uncontrolled appliances such as washer/dryer. Default: False.

vectorize_observations bool

Vectorize the observation space. If False, the observation space is a composite spaces.Dict. If True, it is a spaces.Box. Default: True.

use_all_ochre_observations bool

Whether to use all OCHRE observations or a reduced set of defaults. Default: True.

override_ochre_observations_with_keys List[str]

Only take these observations from OCHRE.

override_equipment_controls Dict

Override the default action space to use the provided equipment and control types. Dictionary with keys "OCHRE equipment type" and list of values "control type". For example, {"HVAC Heating": ["Setpoint"]} or {"HVAC Cooling": ["Duty Cycle"]}.

clip_actions bool

Clip the actions to the action space bounds. Default: False.

vectorize_actions bool

Vectorize the action space. If False, the action space is a composite spaces.Dict. If True, it is a spaces.Box. Default: True.

lookahead str

Length of price lookahead provided as part of observation, in "hour:minute" format. Default: 00:00.

thermal_comfort_unit_penalty float

Unit penalty for being outside the comfort band.

thermal_comfort_band_low float

Lower bound of comfort band for thermal comfort.

thermal_comfort_band_high float

Upper bound of comfort band for thermal comfort.

flat_energy_price float

Price of energy in $/kWh, used in PC program.

reward_scale float

Scale the reward by a constant.

dr_type str

The type of DR program considered, could be "TOU", "PC" or "RTP".

dr_subfolder str

The name of the subfolder in ochre_gym/energy_price containing the DR files. Default: env_name.

tou_price_file str

File name of the TOU price daily signal.

rtp_price_file str

File name of the RTP price signal.

pc_power_file str

File name of the PC power limit signal.

pc_unit_penalty float

Unit penalty for violating the power constraint.

Returns:

Name Type Description
env OchreEnv

The OCHRE Gym environment.