Skip to content

Design Problems

adore.api.schema.adore_schema.DesignProblem 

DesignProblem

Represents an interpretation of the choices and Quantities Of Interest of the design space as a design (optimization) problem. A design problem is defined as a problem where the goal is to find the best possible objective values (minimal or maximal, depending on the definition of the objectives), while making sure that all constraints are satisfied (i.e. their value is above or below the reference value, depending on the definition of the constraint), by systematically varying the design variable values. In an optimization problem, the design variables are varied using an optimization algorithm, but in principle it is also possible to manually tune the design variables. QOIs can also be interpreted as metrics, which means that they are considered generic output of the analysis, but do not mean anything to the design problem.

constraints 
constraints: List[Constraint] = constraints
design_vars 
design_vars: List[Union[DiscreteDesignVar, ContinuousDesignVar]] = design_vars
encoder 
encoder: Optional[str] = encoder
fixed_stats 
fixed_stats: Optional[Statistics] = fixed_stats
id 
id: int = id
metrics 
metrics: List[Metric] = metrics
name 
name: str = name
notes 
notes: Optional[str] = notes
objectives 
objectives: List[Objective] = objectives
opt_config 
opt_config: Optional[OptConfig] = opt_config
ref 
ref: Optional[str] = ref

Optional reference, used for linking to external interfaces.

stats 
stats: Optional[Statistics] = stats

adore.api.schema.adore_schema.DiscreteDesignVar 

DiscreteDesignVar

Represents a discrete design variable, which can have any value from a predefined list of options. A discrete design variable is either derived from a QOI or from an architecture choice.

component_idx 
component_idx: Optional[int] = component_idx
decision_id 
decision_id: int = decision_id
fixed_option_index 
fixed_option_index: Optional[int] = fixed_option_index

Optionally a fixed option to choose. This effectively means the design variable is not active in the design problem.

id 
id: int = id
name 
name: str = name
notes 
notes: Optional[str] = notes
options 
options: List[str] = options

The list of options to choose from.

qoi_id 
qoi_id: Optional[int] = qoi_id
ref 
ref: Optional[str] = ref

Optional reference, used for linking to external interfaces.

system_idx 
system_idx: Optional[List[int]] = system_idx

adore.api.schema.adore_schema.ContinuousDesignVar 

ContinuousDesignVar

Represents a continuous design variable, which can have any value between the lower and upper bound (inclusive). A continuous design variable is always derived from a QOI.

bounds 
bounds: List[float] = bounds

The upper and lower bound (inclusive) of the design variable.

component_idx 
component_idx: Optional[int] = component_idx
decision_id 
decision_id: int = decision_id
fixed_value 
fixed_value: Optional[float] = fixed_value

Optionally a fixed value. This effectively means the design variable is not active in the design problem.

id 
id: int = id
name 
name: str = name
notes 
notes: Optional[str] = notes
qoi_id 
qoi_id: int = qoi_id
ref 
ref: Optional[str] = ref

Optional reference, used for linking to external interfaces.

system_idx 
system_idx: Optional[List[int]] = system_idx

adore.api.schema.adore_schema.Objective 

Objective

Represents an optimization objective, to be either minimized or maximized. An objective is always derived from a QOI.

always_exists 
always_exists: bool = always_exists

Whether this element exists in all architectures (allows it to be used as an objective).

component_idx 
component_idx: Optional[int] = component_idx
id 
id: int = id
maximize 
maximize: bool = maximize

Whether this is maximization (true) or minimization (false) target.

name 
name: str = name
notes 
notes: Optional[str] = notes
original_type 
original_type: QOIType = original_type

Original type of the associated QOI.

qoi_id 
qoi_id: int = qoi_id
ref 
ref: Optional[str] = ref

Optional reference, used for linking to external interfaces.

system_idx 
system_idx: Optional[List[int]] = system_idx

adore.api.schema.adore_schema.Constraint 

Constraint

Represents an optimization inequality constraint. The value should be either above, below or equal to the reference value. A constraint is always derived from a QOI.

always_exists 
always_exists: bool = always_exists

Whether this element exists in all architectures (allows it to be used as an objective).

component_idx 
component_idx: Optional[int] = component_idx
id 
id: int = id
name 
name: str = name
notes 
notes: Optional[str] = notes
original_type 
original_type: QOIType = original_type

Original type of the associated QOI.

pos_feasible 
pos_feasible: bool = pos_feasible

Whether positive (i.e. greater than or equal to the reference value) yields a satisfied (feasible) constraint or not.

qoi_id 
qoi_id: int = qoi_id
ref 
ref: Optional[str] = ref

Optional reference, used for linking to external interfaces.

ref_value 
ref_value: float = ref_value

The reference value for determining whether the constraint is satisfied or not.

system_idx 
system_idx: Optional[List[int]] = system_idx

adore.api.schema.adore_schema.Objective 

Objective

Represents an optimization objective, to be either minimized or maximized. An objective is always derived from a QOI.

always_exists 
always_exists: bool = always_exists

Whether this element exists in all architectures (allows it to be used as an objective).

component_idx 
component_idx: Optional[int] = component_idx
id 
id: int = id
maximize 
maximize: bool = maximize

Whether this is maximization (true) or minimization (false) target.

name 
name: str = name
notes 
notes: Optional[str] = notes
original_type 
original_type: QOIType = original_type

Original type of the associated QOI.

qoi_id 
qoi_id: int = qoi_id
ref 
ref: Optional[str] = ref

Optional reference, used for linking to external interfaces.

system_idx 
system_idx: Optional[List[int]] = system_idx

Types

adore.api.schema.adore_schema_defs.DesignVar 

DesignVar = Union[DiscreteDesignVar, ContinuousDesignVar]