00 - Setup check¶

Learning objectives

confirm that the toolchain of the tutorials is installed: trano (YAML to Modelica), rumoca (Modelica to CasADi, the commit uv sync builds), trano-optimization (collocation), casadi with IPOPT, mpctrano;
generate a first CasADi model from a Trano YAML with the FastModelMPC library and plan its heating.

Run this notebook first; every other tutorial assumes it passes. In continuous integration the notebooks run with MPCTRANO_QUICK=1 (shorter periods), see tutorials/README.md.

In [1]:

Copied!





import sys

if "google.colab" in sys.modules:  # Colab: fetch the tutorials, install mpctrano and rumoca (a no-op elsewhere)
    import subprocess

    subprocess.run(  # a fixed command; the clone is skipped when it exists
        ["git", "clone", "-q", "--depth", "1", "https://github.com/andoludo/mpctrano", "/content/mpctrano"], check=False
    )
    sys.path.insert(0, "/content/mpctrano")
    from tutorials import _colab

    _colab.setup()
import sys

if "google.colab" in sys.modules:  # Colab: fetch the tutorials, install mpctrano and rumoca (a no-op elsewhere)
    import subprocess

    subprocess.run(  # a fixed command; the clone is skipped when it exists
        ["git", "clone", "-q", "--depth", "1", "https://github.com/andoludo/mpctrano", "/content/mpctrano"], check=False
    )
    sys.path.insert(0, "/content/mpctrano")
    from tutorials import _colab

    _colab.setup()

In [2]:

Copied!





from importlib.metadata import version

import casadi
import pandas as pd
import rumoca

import mpctrano
from tutorials._common import MODEL, QUICK, horizon

print(f"trano {version('trano')}")
print(f"trano-optimization {version('trano-optimization')}")
print(f"rumoca {rumoca.version()}")
print(f"casadi {casadi.__version__}")
print(f"mpctrano {mpctrano.__version__}")
print(f"quick mode: {QUICK}")
assert mpctrano.model.rumoca_exports_fast_models(), "rumoca cannot export the FastModel models: run `uv sync`"
from importlib.metadata import version

import casadi
import pandas as pd
import rumoca

import mpctrano
from tutorials._common import MODEL, QUICK, horizon

print(f"trano {version('trano')}")
print(f"trano-optimization {version('trano-optimization')}")
print(f"rumoca {rumoca.version()}")
print(f"casadi {casadi.__version__}")
print(f"mpctrano {mpctrano.__version__}")
print(f"quick mode: {QUICK}")
assert mpctrano.model.rumoca_exports_fast_models(), "rumoca cannot export the FastModel models: run `uv sync`"

trano 0.58.0
trano-optimization 0.2.0
rumoca 0.10.3
casadi 3.8.1
mpctrano 0.2.0
quick mode: False

From a Trano YAML to a CasADi model¶

FastModel.from_yaml runs the whole chain: Trano renders the FastModelMPC Modelica model of the building (building_mpc: the walls, the windows, the heat pump, its buffer, the mixing valve and the radiator, the controls of the systems as inputs), rumoca translates it into the CasADi ODE xdot = f(t, x, u, p), and the roles of the variables are read from the Modelica model.

In [3]:

Copied!





model = mpctrano.FastModel.from_yaml(MODEL)
print("states      ", model.n_states)
print("controls    ", model.control_names)
print("disturbances", model.disturbance_names)
print("comfort     ", model.comfort_states)
print("meters      ", model.meters)
print("parameters  ", model.n_parameters)
model = mpctrano.FastModel.from_yaml(MODEL)
print("states      ", model.n_states)
print("controls    ", model.control_names)
print("disturbances", model.disturbance_names)
print("comfort     ", model.comfort_states)
print("meters      ", model.meters)
print("parameters  ", model.n_parameters)

states       27
controls     ['boiler_001_y.y', 'three_way_valve_001_y.y', 'valve_001_y.y']
disturbances ['weather_inputs.TDryBul', 'weather_inputs.TBlaSky', 'weather_inputs.winSpe', 'weather_inputs.winDir', 'weather_inputs.HDirNor', 'weather_inputs.HDifHor', 'weather_inputs.HGloHor', 'occupancy_1_gai.QCon', 'occupancy_1_gai.QRad']
comfort      ['space_001.TAir']
meters       {'boiler_001.EEle': 'electricity'}
parameters   581

IPOPT is available¶

A plan of the MPC on the model confirms that CasADi can call IPOPT (with SPRAL when CasADi ships it).

In [4]:

Copied!





from mpctrano import ComfortSpec, FlatTariff, ModelPredictiveController, MPCSettings, WeatherForecaster

steps = horizon(12)
controller = ModelPredictiveController(model, MPCSettings(horizon=steps), ComfortSpec(), FlatTariff(price=0.30))
forecaster = WeatherForecaster.from_model(model)
solution = controller.solve(model.initial_state, forecaster.forecast(pd.Timestamp("2023-01-16"), steps, 3600.0))
assert solution.success, solution.status
print(f"IPOPT: {solution.status} in {solution.iterations} iterations, {solution.solve_time:.1f} s")
print(solution.controls.round(2))
from mpctrano import ComfortSpec, FlatTariff, ModelPredictiveController, MPCSettings, WeatherForecaster

steps = horizon(12)
controller = ModelPredictiveController(model, MPCSettings(horizon=steps), ComfortSpec(), FlatTariff(price=0.30))
forecaster = WeatherForecaster.from_model(model)
solution = controller.solve(model.initial_state, forecaster.forecast(pd.Timestamp("2023-01-16"), steps, 3600.0))
assert solution.success, solution.status
print(f"IPOPT: {solution.status} in {solution.iterations} iterations, {solution.solve_time:.1f} s")
print(solution.controls.round(2))

IPOPT: Solve_Succeeded in 34 iterations, 4.7 s
                     boiler_001_y.y  three_way_valve_001_y.y  valve_001_y.y
2023-01-16 00:00:00             0.0                     0.03           0.03
2023-01-16 01:00:00             0.0                     0.04           0.04
2023-01-16 02:00:00             0.0                     0.06           0.06
2023-01-16 03:00:00             0.0                     0.10           0.10
2023-01-16 04:00:00             0.0                     0.17           0.17
2023-01-16 05:00:00             0.0                     0.26           0.26
2023-01-16 06:00:00             0.0                     0.39           0.39
2023-01-16 07:00:00             0.0                     0.32           0.32
2023-01-16 08:00:00             0.0                     0.29           0.29
2023-01-16 09:00:00             0.0                     0.29           0.29
2023-01-16 10:00:00             0.0                     0.32           0.32
2023-01-16 11:00:00             0.0                     0.37           0.37

Key takeaways: the toolchain works end to end; the next notebooks build on the same FastModel object. Further reading: tutorials/README.md (installation and troubleshooting), docs/index.md (the workflow of the package).