Accessed from Optimisation Analysis Settings dialog
Optimisation Constraints allow you to impose upper or lower limits on a building Key Performance Indicator (KPI) such as Cost or Discomfort. Examples of constraints that might be applied to an optimisation analysis are "Discomfort hours must be less than 200", "Daylight availability must be greater than 50%" or "Construction cost must be less than $6m".
Note: The constraint KPI can be the same as a KPI used for an Optimisation Objective.
When viewing optimisation results with constraints applied you will find that some of the results generated don't respect the constraints. This is normal, however you should also see that many more new points are generated that do meet constraints because the optimiser favours such design variants when creating new generations.
When using constraints you will sometimes find that the solution will appear not to converge to as clean a Pareto front as occurs without constraints applied. This is likely to be due to the way constraints are modelled as modified objectives in DesignBuilder. In a problem with 2 objectives and 1 constraint there are effectively 3 objective functions which will give rise to a 3-D Pareto front. This cannot be adequately visualised on a 2-D scatter graph, but it is likely that the Pareto Front would be cleaner were it being viewed in 3-D with the constraint as the 3rd dimension.
Your name for the constraint (e.g. "ASHRAE 55 discomfort hours < 200")
Select the Constraint Key Performance Indicator (KPI) from the list:
Note: The 4-Daylight availability constraint is not available in current versions.
Select whether the criterion refers to an upper or lower limit.
Enter the upper or lower limit value for the KPI.
Enter text to define the units of the value. This text does not affect the calculations.