power_grid_model
- class power_grid_model.PowerGridModel(*_args, **_kwargs)
Bases:
object
Main class for Power Grid Model
- property batch_error: Optional[PowerGridBatchError]
Get the batch error object, if present
- Returns:
Batch error object, or None
- property all_component_count: Dict[str, int]
Get count of number of elements per component type. If the count for a component type is zero, it will not be in the returned dictionary.
- Returns:
A dictionary with
key: Component type name
value: Integer count of elements of this type
- copy() PowerGridModel
Copy the current model
- Returns:
A copy of PowerGridModel
- __init__(input_data: Dict[str, ndarray], system_frequency: float = 50.0)
Initialize the model from an input data set.
- Parameters:
input_data –
Input data dictionary
key: Component type name
value: 1D numpy structured array for this component input
system_frequency – Frequency of the power system, default 50 Hz
- update(*, update_data: Dict[str, ndarray])
Update the model with changes.
- Parameters:
update_data –
Update data dictionary
key: Component type name
value: 1D numpy structured array for this component update
- Returns:
None
- get_indexer(component_type: str, ids: ndarray)
Get array of indexers given array of ids for component type
- Parameters:
component_type – Type of component
ids – Array of ids
- Returns:
Array of indexers, same shape as input array ids
- calculate_power_flow(*, symmetric: bool = True, error_tolerance: float = 1e-08, max_iterations: int = 20, calculation_method: Union[CalculationMethod, str] = CalculationMethod.newton_raphson, update_data: Optional[Dict[str, Union[ndarray, Dict[str, ndarray]]]] = None, threading: int = -1, output_component_types: Optional[Union[Set[str], List[str]]] = None, continue_on_batch_error: bool = False) Dict[str, ndarray]
Calculate power flow once with the current model attributes. Or calculate in batch with the given update dataset in batch.
- Parameters:
symmetric (bool, optional) –
Whether to perform a three-phase symmetric calculation.
True: Three-phase symmetric calculation, even for asymmetric loads/generations (Default).
False: Three-phase asymmetric calculation.
error_tolerance (float, optional) – Error tolerance for voltage in p.u., applicable only when the calculation method is iterative.
max_iterations (int, optional) – Maximum number of iterations, applicable only when the calculation method is iterative.
calculation_method (an enumeration or string) –
The calculation method to use.
newton_raphson: Use Newton-Raphson iterative method (default).
linear: Use linear method.
update_data (dict, optional) –
None: Calculate power flow once with the current model attributes. Or a dictionary for batch calculation with batch update.
key: Component type name to be updated in batch.
value:
For homogeneous update batch (a 2D numpy structured array):
Dimension 0: Each batch.
Dimension 1: Each updated element per batch for this component type.
For inhomogeneous update batch (a dictionary containing two keys):
indptr: A 1D integer numpy array with length n_batch + 1. Given batch number k, the update array for this batch is data[indptr[k]:indptr[k + 1]]. This is the concept of compressed sparse structure. https://docs.scipy.org/doc/scipy/reference/generated/scipy.sparse.csr_matrix.html
data: 1D numpy structured array in flat.
threading (int, optional) –
Applicable only for batch calculation.
< 0: Sequential
= 0: Parallel, use number of hardware threads
> 0: Specify number of parallel threads
output_component_types ({set, list}, optional) – List or set of component types you want to be present in the output dict. By default, all component types will be in the output.
continue_on_batch_error (bool, optional) – If the program continues (instead of throwing error) if some scenarios fail.
- Returns:
Dictionary of results of all components.
key: Component type name to be updated in batch.
value:
For single calculation: 1D numpy structured array for the results of this component type.
For batch calculation: 2D numpy structured array for the results of this component type.
Dimension 0: Each batch.
Dimension 1: The result of each element for this component type.
- Raises:
Exception – In case an error in the core occurs, an exception will be thrown.
- calculate_state_estimation(*, symmetric: bool = True, error_tolerance: float = 1e-08, max_iterations: int = 20, calculation_method: Union[CalculationMethod, str] = CalculationMethod.iterative_linear, update_data: Optional[Dict[str, Union[ndarray, Dict[str, ndarray]]]] = None, threading: int = -1, output_component_types: Optional[Union[Set[str], List[str]]] = None, continue_on_batch_error: bool = False) Dict[str, ndarray]
Calculate state estimation once with the current model attributes. Or calculate in batch with the given update dataset in batch.
- Parameters:
symmetric (bool, optional) –
Whether to perform a three-phase symmetric calculation.
True: Three-phase symmetric calculation, even for asymmetric loads/generations (Default).
False: Three-phase asymmetric calculation.
error_tolerance (float, optional) – error tolerance for voltage in p.u., only applicable when the calculation method is iterative.
max_iterations (int, optional) – Maximum number of iterations, applicable only when the calculation method is iterative.
calculation_method (an enumeration) – Use iterative linear method.
update_data (dict, optional) –
None: Calculate state estimation once with the current model attributes. Or a dictionary for batch calculation with batch update.
key: Component type name to be updated in batch.
value:
For homogeneous update batch (a 2D numpy structured array):
Dimension 0: Each batch.
Dimension 1: Each updated element per batch for this component type.
For inhomogeneous update batch (a dictionary containing two keys):
indptr: A 1D integer numpy array with length n_batch + 1. Given batch number k, the update array for this batch is data[indptr[k]:indptr[k + 1]]. This is the concept of compressed sparse structure. https://docs.scipy.org/doc/scipy/reference/generated/scipy.sparse.csr_matrix.html
data: 1D numpy structured array in flat.
threading (int, optional) –
Applicable only for batch calculation.
< 0: Sequential
= 0: Parallel, use number of hardware threads
> 0: Specify number of parallel threads
output_component_types ({set, list}, optional) – List or set of component types you want to be present in the output dict. By default, all component types will be in the output.
continue_on_batch_error (bool, optional) – If the program continues (instead of throwing error) if some scenarios fail.
- Returns:
Dictionary of results of all components.
key: Component type name to be updated in batch.
value:
For single calculation: 1D numpy structured array for the results of this component type.
For batch calculation: 2D numpy structured array for the results of this component type.
Dimension 0: Each batch.
Dimension 1: The result of each element for this component type.
- Raises:
Exception – In case an error in the core occurs, an exception will be thrown.
- calculate_short_circuit(*, calculation_method: Union[CalculationMethod, str] = CalculationMethod.iec60909, update_data: Optional[Dict[str, Union[ndarray, Dict[str, ndarray]]]] = None, threading: int = -1, output_component_types: Optional[Union[Set[str], List[str]]] = None, continue_on_batch_error: bool = False, short_circuit_voltage_scaling: Union[ShortCircuitVoltageScaling, str] = ShortCircuitVoltageScaling.maximum) Dict[str, ndarray]
Calculate a short circuit once with the current model attributes. Or calculate in batch with the given update dataset in batch
- Parameters:
calculation_method (an enumeration) – Use the iec60909 standard.
update_data –
None: calculate a short circuit once with the current model attributes. Or a dictionary for batch calculation with batch update
key: Component type name to be updated in batch
value:
For homogeneous update batch (a 2D numpy structured array):
Dimension 0: each batch
Dimension 1: each updated element per batch for this component type
For inhomogeneous update batch (a dictionary containing two keys):
indptr: A 1D integer numpy array with length n_batch + 1. Given batch number k, the update array for this batch is data[indptr[k]:indptr[k + 1]]. This is the concept of compressed sparse structure. https://docs.scipy.org/doc/scipy/reference/generated/scipy.sparse.csr_matrix.html
data: 1D numpy structured array in flat.
threading (int, optional) –
Applicable only for batch calculation.
< 0: Sequential
= 0: Parallel, use number of hardware threads
> 0: Specify number of parallel threads
output_component_types ({set, list}, optional) – List or set of component types you want to be present in the output dict. By default, all component types will be in the output.
continue_on_batch_error (bool, optional) – If the program continues (instead of throwing error) if some scenarios fail.
- Returns:
Dictionary of results of all components.
key: Component type name to be updated in batch.
value:
For single calculation: 1D numpy structured array for the results of this component type.
For batch calculation: 2D numpy structured array for the results of this component type.
Dimension 0: Each batch.
Dimension 1: The result of each element for this component type.
- Raises:
Exception – In case an error in the core occurs, an exception will be thrown.
- power_grid_model.initialize_array(data_type: str, component_type: str, shape: Union[tuple, int], empty: bool = False)
Initializes an array for use in Power Grid Model calculations
- Parameters:
data_type – input, update, sym_output, or asym_output
component_type – one component type, e.g. node
shape – shape of initialization integer, it is a 1-dimensional array tuple, it is an N-dimensional (tuple.shape) array
empty – if leave the memory block un-initialized
- Returns:
np structured array with all entries as null value
enum
Common Enumerations
Note: these enumeration match the C++ arithmetic core, so don’t change the values unless you change them in C++ as well
- class power_grid_model.enum.LoadGenType(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)
Bases:
IntEnum
Load and Generator Types
- const_power = 0
- const_impedance = 1
- const_current = 2
- class power_grid_model.enum.WindingType(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)
Bases:
IntEnum
Transformer Winding Types
- wye = 0
- wye_n = 1
- delta = 2
- zigzag = 3
- zigzag_n = 4
- class power_grid_model.enum.BranchSide(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)
Bases:
IntEnum
Branch Sides
- from_side = 0
- to_side = 1
- class power_grid_model.enum.Branch3Side(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)
Bases:
IntEnum
Branch3 Sides
- side_1 = 0
- side_2 = 1
- side_3 = 2
- class power_grid_model.enum.CalculationType(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)
Bases:
IntEnum
Calculation Types
- power_flow = 0
- state_estimation = 1
- short_circuit = 2
- class power_grid_model.enum.CalculationMethod(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)
Bases:
IntEnum
Calculation Methods
- linear = 0
- newton_raphson = 1
- iterative_linear = 2
- iterative_current = 3
- linear_current = 4
- iec60909 = 5
- class power_grid_model.enum.MeasuredTerminalType(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)
Bases:
IntEnum
The type of asset measured by a (power) sensor
- branch_from = 0
Measuring the from-terminal between a branch (except link) and a node
- branch_to = 1
Measuring the to-terminal between a branch (except link) and a node
- source = 2
Measuring the terminal between a source and a node
- shunt = 3
Measuring the terminal between a shunt and a node
- load = 4
Measuring the terminal between a load and a node
- generator = 5
Measuring the terminal between a generator and a node
- branch3_1 = 6
Measuring the terminal-1 between a branch3 and a node
- branch3_2 = 7
Measuring the terminal-2 between a branch3 and a node
- branch3_3 = 8
Measuring the terminal-3 between a branch3 and a node
- node = 9
Measuring the total power injection into a node
- class power_grid_model.enum.FaultType(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)
Bases:
IntEnum
The type of fault represented by a fault component
- three_phase = 0
- single_phase_to_ground = 1
- two_phase = 2
- two_phase_to_ground = 3
- nan = -128
Unspecified fault type. Needs to be overloaded at the latest in the update_data
- class power_grid_model.enum.FaultPhase(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)
Bases:
IntEnum
The faulty phase(s) affected by the provided fault type
- abc = 0
All phases are faulty in a three-phase fault
- a = 1
The first phase is faulty in a single-phase-to-ground fault
- b = 2
The second phase is faulty in a single-phase-to-ground fault
- c = 3
The third phase is faulty in a single-phase-to-ground fault
- ab = 4
The first and second phase are faulty in a two-phase or two-phase-to-ground fault
- ac = 5
The first and third phase are faulty in a two-phase or two-phase-to-ground fault
- bc = 6
The second and third phase are faulty in a two-phase or two-phase-to-ground fault
- default_value = -1
Use the default fault phase. Depends on the fault_type.
- nan = -128
Unspecified fault phase. Needs to be overloaded at the latest in the update_data
validation
- power_grid_model.validation.validate_input_data(input_data: Dict[str, ndarray], calculation_type: Optional[CalculationType] = None, symmetric: bool = True) Optional[List[ValidationError]]
Validates the entire input dataset:
Is the data structure correct? (checking data types and numpy array shapes)
Are all required values provided? (checking NaNs)
Are all ID’s unique? (checking object identifiers across all components)
Are the supplied values valid? (checking limits and other logic as described in “Graph Data Model”)
- Parameters:
input_data – A power-grid-model input dataset
calculation_type – Supply a calculation method, to allow missing values for unused fields
symmetric – A boolean to state whether input data will be used for a symmetric or asymmetric calculation
- Returns:
None if the data is valid, or a list containing all validation errors.
- Raises:
Error – KeyError | TypeError | ValueError: if the data structure is invalid.
- power_grid_model.validation.validate_batch_data(input_data: Dict[str, ndarray], update_data: Dict[str, Union[ndarray, Dict[str, ndarray]]], calculation_type: Optional[CalculationType] = None, symmetric: bool = True) Optional[Dict[int, List[ValidationError]]]
The input dataset is validated:
Is the data structure correct? (checking data types and numpy array shapes)
Are all input data ID’s unique? (checking object identifiers across all components)
- For each batch the update data is validated:
Is the update data structure correct? (checking data types and numpy array shapes)
Are all update ID’s valid? (checking object identifiers across update and input data)
- Then (for each batch independently) the input dataset is updated with the batch’s update data and validated:
Are all required values provided? (checking NaNs)
Are the supplied values valid? (checking limits and other logic as described in “Graph Data Model”)
- Parameters:
input_data – A power-grid-model input dataset
update_data – A power-grid-model update dataset (one or more batches)
calculation_type – Supply a calculation method, to allow missing values for unused fields
symmetric – A boolean to state whether input data will be used for a symmetric or asymmetric calculation
- Returns:
None if the data is valid, or a dictionary containing all validation errors, where the key is the batch number (0-indexed).
- Raises:
Error – KeyError | TypeError | ValueError: if the data structure is invalid.
- power_grid_model.validation.assert_valid_input_data(input_data: Dict[str, ndarray], calculation_type: Optional[CalculationType] = None, symmetric: bool = True)
Validates the entire input dataset:
Is the data structure correct? (checking data types and numpy array shapes)
Are all required values provided? (checking NaNs)
Are all ID’s unique? (checking object identifiers across all components)
Are the supplied values valid? (checking limits and other logic as described in “Graph Data Model”)
- Parameters:
input_data – A power-grid-model input dataset
calculation_type – Supply a calculation method, to allow missing values for unused fields
symmetric – A boolean to state whether input data will be used for a symmetric or asymmetric calculation
- Raises:
KeyError | TypeError | ValueError – if the data structure is invalid.
ValidationException – if the contents are invalid.
- power_grid_model.validation.assert_valid_batch_data(input_data: Dict[str, ndarray], update_data: Dict[str, Union[ndarray, Dict[str, ndarray]]], calculation_type: Optional[CalculationType] = None, symmetric: bool = True)
The input dataset is validated:
Is the data structure correct? (checking data types and numpy array shapes)
Are all input data ID’s unique? (checking object identifiers across all components)
- For each batch the update data is validated:
Is the update data structure correct? (checking data types and numpy array shapes)
Are all update ID’s valid? (checking object identifiers across update and input data)
- Then (for each batch independently) the input dataset is updated with the batch’s update data and validated:
Are all required values provided? (checking NaNs)
Are the supplied values valid? (checking limits and other logic as described in “Graph Data Model”)
- Parameters:
input_data – a power-grid-model input dataset
update_data – a power-grid-model update dataset (one or more batches)
calculation_type – Supply a calculation method, to allow missing values for unused fields
symmetric – A boolean to state whether input data will be used for a symmetric or asymmetric calculation
- Raises:
KeyError | TypeError | ValueError – if the data structure is invalid.
ValidationException – if the contents are invalid.
- power_grid_model.validation.errors_to_string(errors: Optional[Union[List[ValidationError], Dict[int, List[ValidationError]]]], name: str = 'the data', details: bool = False, id_lookup: Optional[Union[List[str], Dict[int, str]]] = None) str
Convert a set of errors (list or dict) to a human readable string representation.
- Parameters:
errors – The error objects. List for input_data only, dict for batch data.
name – Human understandable name of the dataset, e.g. input_data, or update_data.
details – Display object ids and error specific information.
id_lookup – A list or dict (int->str) containing textual object ids
- Returns:
A human readable string representation of a set of errors.
errors
- class power_grid_model.validation.errors.ValidationError
The Validation Error is an abstract base class which should be extended by all validation errors. It supplies three public member variables: component, field and ids; storing information about the origin of the validation error. Error classes can extend the public members. For example:
- NotBetweenError(ValidationError):
component = ‘vehicle’ field = ‘direction’ id = [3, 14, 15, 92, 65, 35] ref_value = (-3.1416, 3.1416)
For convenience, a human readable representation of the error is supplied using the str() function. I.e. print(str(error)) will print a human readable error message like:
Field direction is not between -3.1416 and 3.1416 for 6 vehicles
- component: Optional[Union[str, List[str]]] = None
The component, or components, to which the error applies.
- field: Optional[Union[str, List[str], List[Tuple[str, str]]]] = None
The field, or fields, to which the error applies. A field can also be a tuple (component, field) when multiple components are being addressed.
- ids: Optional[Union[List[int], List[Tuple[str, int]]]] = None
The object identifiers to which the error applies. A field object identifier can also be a tuple (component, id) when multiple components are being addressed.
- property component_str: str
A string representation of the component to which this error applies
- property field_str: str
A string representation of the field to which this error applies
- get_context(id_lookup: Optional[Union[List[str], Dict[int, str]]] = None) Dict[str, Any]
Returns a dictionary that supplies (human readable) information about this error. Each member variable is included in the dictionary. If a function {field_name}_str() exists, the value is overwritten by that function.
- Parameters:
id_lookup – A list or dict (int->str) containing textual object ids
utils
- power_grid_model.utils.import_json_data(json_file: Path, data_type: str, ignore_extra: bool = False) Union[Dict[str, ndarray], Dict[str, Union[ndarray, Dict[str, ndarray]]]]
import json data
- Parameters:
json_file – path to the json file
data_type – type of data: input, update, sym_output, or asym_output
ignore_extra – Allow (and ignore) extra attributes in the json file
- Returns:
A single or batch dataset for power-grid-model
- power_grid_model.utils.export_json_data(json_file: Path, data: Union[Dict[str, ndarray], Dict[str, Union[ndarray, Dict[str, ndarray]]]], indent: Optional[int] = 2, compact: bool = False)
export json data
- Parameters:
json_file – path to json file
data – a single or batch dataset for power-grid-model
indent – indent of the file, default 2
compact – write components on a single line
- Returns:
Save to file