Handler classes related to bounds¶

BoundHandler¶

class simu.core.model.bound.BoundHandler¶

The boundary handler keeps process model properties that must be strictly positive in order to remain within the model’s mathematical domain.

Thermodynamic (material) properties, such as temperature, pressure and molar quantities do not need to be added, as their bounds are to be defined in the simu.ThermoContribution object(s) that constraint(s) their domain.

Note

These bounds do not represent inequality constraints in an optimization problem, but true domain boundaries that need to be respected by any solver in order to keep a feasible vector of independent variables.

It is also generally not a good idea to impose bounds that are based on the expected range for a solved model, as a solver might need to step through values that are not physically reasonable but yet within the model domain.

Typical examples of bounds defined here are:

Temperature difference over a heat exchanger when using the logarithmic mean temperature. A temperature cross-over would cause negative arguments to the log function.

Process model parameters that might be subject to change, such as geometric dimensions of equipment.

A bad example is using these bounds to keep a temperature value between a lower and an upper limit.

__init__()¶

add(name: str, bound: Quantity)¶

Add a quantity to the bound handler to signal the solver that its value must remain strictly positive.

If the unit includes an offset, such as degC or barg, this offset is eliminated:

>>> from simu import SymbolQuantity
>>> # in a real case, this better be a dependent property
>>> b = SymbolQuantity("T", "degC")
>>> handler = BoundHandler()
>>> handler.add("T", b)
>>> print(f"{handler["T"]}")
T delta_degree_Celsius

If the intention is really to define a bound at a given absolute value, this must explicitly be done as follows:

>>> from simu import Quantity
>>> handler.add("T2", b - Quantity(31.4159, "degC"))
>>> print(f"{handler["T2"]}")
(T-31.4159) delta_degree_Celsius

create_proxy() → Mapping[str, Quantity]¶: Create a proxy object

BoundProxy¶

simu.core.model.bound.BoundProxy¶: A map of quantities, interpreted as variables to stay positive