commonpower.control.safety_layer.safety_layers.ActionReplacementWithOptSafetyLayer

class ActionReplacementWithOptSafetyLayer(penalty: BasePenalty, solver: OptSolver | None = None)

Bases: BaseSafetyLayer

Action replacement safety layer. Action violating the constraints is replaced by safe action determined through an optimization method.

Parameters:

• penalty (BasePenalty) – class defining the penalty behavior for unsafe actions

• solver (OptSolver, optional) – solver for optimization problem, defaults to direct gurobi https://pyomo.readthedocs.io/en/stable/library_reference/solvers/gurobi_direct.html

Returns:

ActionReplacementWithOptSafetyLayer

Methods

action_distance	Euclidean norm between action expressed by the model and the action in dict
compute_safe_action	Checks whether the actions proposed by the controller satisfy the constraints of the controlled entities and replaces by pyomo-generated output if necessary.
get_penalty	Get penalty depending on the penalty class used.
initialize	Initializes the safety layer :param nodes: list of controlled entities to be safeguarded :type nodes: List[ModelEntity] :param top_level_nodes: list of controlled entities in highest level of model tree :type top_level_nodes: List[ModelEntity] :param solver: solver for optimization problem which will be called by Pyomo :type solver: OptSolver
is_action_feasible	Check if action is feasible.
prepare_model	Clone model from sys to run local optimization over
set_action_from_model	Corrects an action according to a model.
set_action_in_model	Sets the model values to the action to initialize the optimization.
solve_model	Finds a feasible action by solving a local model.

Attributes

DISTANCE_EPS

action_distance(model: ConcreteModel, action: dict) → SumExpression

Euclidean norm between action expressed by the model and the action in dict

Parameters:

• model (ConcreteModel) – pyomo optimization model

• action (dict) – action description

Returns:

SumExpression –

distance between the action and the model values expressed

in pyomo class, convert to float by running value() over it

compute_safe_action(action: Dict | None = None) → Tuple[Dict, bool, float]

Checks whether the actions proposed by the controller satisfy the constraints of the controlled entities and replaces by pyomo-generated output if necessary.

Parameters:

action (dict) – action suggested by the controller

Returns:

safe_action (dict) – verified action action_corrected (bool): whether the action was corrected or not correction_penalty (float): penalty for action correction (0 if action was not corrected)

get_penalty(action_distance: float) → float

Get penalty depending on the penalty class used.

Parameters:

action_distance (float) – distance between the safe and unsafe action, only used if penalty is distance based

Returns:

float – computed penalty for the action

is_action_feasible(action: dict) → dict

Check if action is feasible.

Parameters:

action (dict) – action to check

Returns:

bool – True if action is feasible

prepare_model() → ConcreteModel

Clone model from sys to run local optimization over

Returns:

ConcreteModel –

pyomo optimization model, representing

the part of the system under supervision of the safety

set_action_from_model(model: ConcreteModel, action: dict) → dict

Corrects an action according to a model.

Parameters:

• model (ConcreteModel) – pyomo model to ge the values from

• action (dict) – action to correct

Returns:

dict – safe action according to the model

set_action_in_model(model: ConcreteModel, action: dict, fix_values: bool = False) → ConcreteModel

Sets the model values to the action to initialize the optimization. If fix values is True, the model values will be fixed, which is useful to check the feasibility of the given action.

Parameters:

• model (ConcreteModel) – pyomo optimization model

• action (dict) – action to set into the model

• fix_values (bool) – optionally fixes the values for the optimization problem

Returns:

ConcreteModel – model with values set

solve_model(model: ConcreteModel) → float

Finds a feasible action by solving a local model. Returns the distance of the original unsafe action to the solved model.

Parameters:

model (ConcreteModel) – model to solve

Returns:

float – distance from action or None