Time Series Modifications

These functions are used to modify and simplify the use of time series data.

Functions are found in pyflow_acdc.grid_modifications.

Renewable Source Zone

Renewable Source Zone is an object designed to unify multiple renewable sources sharing the same time series data. For instance, if multiple wind turbines generate identical power output, they can be grouped into a single wind power plant zone.

Add Renewable Source Zone

add_RenSource_zone(grid, name)

Adds a renewable source zone to the grid.

Parameter	Type	Description
grid	Grid	Grid to modify
name	str	Zone name
Returns	Ren_source_zone	Created renewable zone

Example

zone = pyf.add_RenSource_zone(grid, "WindZone1")

Assign Renewable to Zone

assign_RenToZone(grid, ren_source_name, new_zone_name)

Assigns a renewable source to a zone.

Parameter	Type	Description
grid	Grid	Grid containing source
ren_source_name	str	Name of renewable source
new_zone_name	str	Name of target zone

Example

This example shows how to assign two renewable sources to the same zone. Each renewable source has a base power of 22 MW. You can either create the renewbale sources and then the zones to assign them to or create the zone first and then add the renewable sources to it.

grid = pyf.Grid(100)

node = pyf.add_AC_node(grid,66, name="node1")
node2 = pyf.add_AC_node(grid,66, name="node2")



pyf.add_RenSource(grid, 'node1', 22, ren_source_name='wind1')
pyf.add_RenSource(grid, 'node2', 22, ren_source_name='wind2')

zone = pyf.add_RenSource_zone(grid, "WindZone1")

pyf.assign_RenToZone(grid, "wind1", "WindZone1")
pyf.assign_RenToZone(grid, "wind2", "WindZone1")

Or you can create the zone first and then add the renewable sources to it. The assigment of zones will be called when adding the renewable sources.

grid = pyf.Grid(100)

node = pyf.add_AC_node(grid,66, name="node1")
node2 = pyf.add_AC_node(grid,66, name="node2")

pyf.add_RenSource_zone(grid, "WindZone1")

pyf.add_RenSource(grid, 'node1', 22, ren_source_name='wind1', zone="WindZone1")
pyf.add_RenSource(grid, 'node2', 22, ren_source_name='wind2', zone="WindZone1")

Price Zone

Price zones function also like renewable sources zone when dealing with time series data. As they allow to group multiple nodes and generators with the same price data. Check the Add Price Zone and Add DCDC Converter for more information.

Time Series data

Add Time Series

Time series data can be added to the grid by using the add_TimeSeries() function. This function allows to add time series data to a specific component. Time series data is imported from csv files.

add_TimeSeries(grid, Time_Series_data, associated=None, TS_type=None)

Adds time series data to grid components.

Parameter	Type	Description
grid	Grid	Grid to modify
Time_Series_data	DataFrame	Time series data
associated	str	Object name
TS_type	str	Time series type

Accepted types

The time series update will depend on which type it can be associated to, this is why it is important to have different names for each object even if they are in different classes. The following types are accepted:

Type	Description	Object Associated	Affected object variable
‘Load’	Load time series	Price_Zone, Node_AC, Node_DC	obj.PLi_factor
‘price’	Price time series	Price_Zone, Node_AC, Node_DC	obj.price
‘WPP’, ‘OWPP’, ‘SF’, ‘REN’	Per unit power available of base power	Ren_source_zone, Ren_Source	obj.PRGi_available

The following parameters are only available for price zones [1].

Type	Description	Object Associated	Affected object variable
‘a_CG’	Quadratic factor of cost of generation	Price_Zone	obj.a_CG
‘b_CG’	Linear factor of cost of generation	Price_Zone	obj.b_CG
‘c_CG’	Constant factor of cost of generation	Price_Zone	obj.c_CG
‘PGL_min’	Minimum value of P_N for the price zone	Price_Zone	obj.PGL_min
‘PGL_max’	Maximum value of P_N for the price zone	Price_Zone	obj.PGL_max

Data format

Data format depends on the selection of associated and TS_type. The value in associated is the name of the object which it refers to, this can be a node, a renewable source, a price zone or a renewable source zone.

For a dataset of length n, the CSV will follow this format: Position 0 is treated as the column name when imported into pandas. The first line should contain only headers and no data. Then there are cases where:

associated and TS_type assigned by user

position value 0 time series name 1 start of the data n end of the data

0 Load_n1 Load_n2 price_n1 1 0.95 0.55 20 2 0.75 0.95 30 3 0.84 0.72 30

n1_load_data = pd.DataFrame({"Load_n1": [0.95, 0.75, 0.84]})
n2_load_data = pd.DataFrame({"Load_n2": [0.55, 0.95, 0.72]})
price_data   = pd.DataFrame({"price_n1": [20, 30, 40]})

pyf.add_TimeSeries(grid, n1_load_data,associated="node1", TS_type="Load")
pyf.add_TimeSeries(grid, n2_load_data,associated="node2", TS_type="Load")
pyf.add_TimeSeries(grid, price_data,associated="node1", TS_type="price")

only associated assigned by user

position value 0 time series name 1 TS_type 2 start of the data n+1 end of the data

0 Load_n1 Load_n2 price_n1 1 Load Load price 2 0.95 0.55 20 3 0.75 0.95 30 4 0.84 0.72 40

n1_data = pd.DataFrame({"Load_n1": ['Load', 0.95, 0.75, 0.84],
                        "price_n1": ['price', 20, 30, 40]})
n2_data = pd.DataFrame({"Load_n2": ['Load', 0.55, 0.95, 0.72]})
pyf.add_TimeSeries(grid, n1_data,associated="node1")
pyf.add_TimeSeries(grid, n2_data,associated="node2")

only TS_type assigned by user

position value 0 time series name 1 Object name 2 start of the data n+1 end of the data

0 Load_n1 Load_n2 price_n1 1 node1 node2 node1 2 0.95 0.55 20 3 0.75 0.95 30 4 0.84 0.72 40

load_data  = pd.DataFrame({"Load_n1": ['node1', 0.95, 0.75, 0.84],
                           "Load_n2": ['node2', 0.55, 0.95, 0.72]})
price_data = pd.DataFrame({"price_n1": ['node1', 20, 30, 40]})
pyf.add_TimeSeries(grid, load_data, TS_type="Load")
pyf.add_TimeSeries(grid, price_data, TS_type="price")

associated = None and TS_type = None

position value 0 time series name 1 Object name 2 TS type 3 start of the data n+2 end of the data

0 Load_n1 Load_n2 price_n1 1 node1 node2 node1 2 Load Load price 3 0.95 0.75 20 4 0.75 0.84 30 5 0.84 0.72 40

load_data = pd.DataFrame({"Load_n1": ['node1', 'Load', 0.95, 0.75, 0.84],
                          "Load_n2": ['node2', 'Load', 0.55, 0.95, 0.72],
                          "price_n1":['node1', 'price', 20, 30, 40]})
pyf.add_TimeSeries(grid, load_data)

Time-Series Clustering

These functions are found in pyflow_acdc.Time_series_clustering.

identify_correlations(grid, time_series=[], correlation_threshold=0, cv_threshold=0, central_market=[], print_details=False, correlation_decisions=[])

Detects highly correlated time-series and builds reduction decisions.

cluster_TS(grid, n_clusters, time_series=[], central_market=[], algorithm='Kmeans', cv_threshold=0, correlation_threshold=0.8, print_details=False, correlation_decisions=[], critical_idx=[], base_critical_ratio=0.5, scaler_type='robust', **kwargs)

Clusters time-series profiles into representative operating states.

run_clustering_analysis_and_plot(grid, algorithms=['kmeans', 'kmedoids', 'ward', 'pam_hierarchical'], n_clusters_list=DEFAULT_CLUSTER_NUMBERS, path='clustering_results', time_series=[], print_details=False, ts_options=[None, 0, 0.8], correlation_decisions=[True, '2', True], plotting_options=[None, 'svg'], identifier=None)

Runs clustering sweeps and exports comparison plots/artifacts.

cluster_analysis(grid, clustering_options)

Applies clustering configuration and returns representative scenarios for TS/TEP workflows.

References

[1]

B. C. Valerio, V. A. Lacerda, M. Cheah-Mañe, P. Gebraad, and O. Gomis-Bellmunt, “Optimizing offshore wind integration through multi-terminal DC grids: a market-based OPF framework for the North Sea interconnectors,” IET Conference Proceedings, vol. 2025, no. 6, pp. 150–155, 2025. doi: 10.1049/icp.2025.1198