replaceable model Capacity in zone model

[saroele]

include stochastic capacities for state estimation by replaceable models for the capacity.
This avoids duplication of the zone models.

The only thing to change when redeclaring the capacity is to provide additional inputs for the stochastic variables.

Best to be implemented in a separate branch

[rubenbaetens]

Maybe the use of the term 'stochastic' isn't appropriate here as it assumes a probability theory, whether - if i understand it well - it's (just) an unknown (to be stated) variable ?

[saroele]

Including the w and v in the models is actually going from a deterministic to a stochastic thermal zone model. This property is then used in state estimation, where the stochastic 'noise' is estimated.

So I think it's not wrong to call it stochastic...