repository·main·Indexed 18 days ago
https://github.com/sienna-platform/powersystems.jlA rigorous data modeling package for power systems analysis providing core data structures for the Sienna Platform ecosystem, including PowerSimulations.jl and PowerSimulationsDynamics.jl. It supports a wide range of device data (generators, transmission, HVDC, storage, loads), parsing for MATPOWER, PSS/e, and RTS-GMLC formats, and a two-layer architecture separating static components from dynamic components. Requires Julia v1.6+.
MarketBidCost is an OperationalCost data structure designed to model production costs similar to US electricity market auctions. It allows for joint bidding of energy and ancillary services. You can use it to define incremental or decremental offer curves, either as static curves or as time-varying data using TimeSeriesData.A System is the central data container in PowerSystems.jl. It acts as a registry that holds all Component objects (representing physical or logical elements like generators or buses) and references to their associated time series data.
Data Storage Model:
PowerSystems.jl uses a hybrid storage approach to manage memory efficiently:
This design ensures that only the relevant portions of data are loaded during a query, preventing high memory overhead when working with large datasets.
A Component is any element within a power system model, such as generators, loads, buses, transmission lines, or services. Components are organized into an abstract type hierarchy based on their role in the system.
Ownership Constraint:
A component instance can belong to at most one System at a time. If you attempt to add a component to a second System without first removing it from its original System, an error will be raised. This prevents silent aliasing and ensures unambiguous data ownership.
To support serialization and de-serialization of custom components, your struct must meet these criteria:
PowerSystems.Component (which is a subtype of InfrastructureSystemsType). This ensures InfrastructureSystems handles the serialize and deserialize methods automatically.serialize/deserialize methods.deserialize method must have a keyword-only constructor, as the deserialization process splats dictionary key/value pairs into the constructor.PowerSystems.jl components, you must serialize those components as UUIDs rather than actual values. This allows the deserializer to restore references to existing objects rather than creating duplicate copies.deserialize method.Forecasts are used for simulation with receding horizons. They are defined by a resolution (time between steps in the horizon), an interval (time between forecast updates), and a horizon (number of forecasted values).
PowerSystems.jl provides three specific types of forecast structs:
Deterministic: A point forecast containing only predicted values without uncertainty.Probabilistic: Stores discretized cumulative distribution functions (CDFs) or probability distribution functions (PDFs) for each time step.Scenarios: Stores a set of probable trajectories, where each trajectory is assumed to have equal probability.The PowerSystems.jl library uses a structured type hierarchy to categorize different components of a power system. You can explore the complete tree of types by using the TypeTree utility within a Julia environment. This hierarchy is organized around a root type (such as InfrastructureSystemsType) and branches out into specific system components.
using PowerSystems
# To visualize the hierarchy, you can use the TypeTree utility
# (Note: This requires the TypeTree package and specific doc utilities)
import TypeTree: tt
# Example of printing the hierarchy under InfrastructureSystemsType
print(join(tt(PowerSystems.IS.InfrastructureSystemsType), ""))RenewableGenerationCost type is used to model the cost functions associated with renewable generation assets. It is part of the PowerSystems module and is typically used to define how much it costs to operate or dispatch renewable resources within a power system model.PowerSystems.jl supports a wide range of device data types for modeling power systems, including:
In PowerSystems.jl, hydro reservoir topology is used to define the source and destination of water for hydropower units. This is achieved by linking HydroReservoir components to HydroTurbine or HydroPumpTurbine units through explicit component relationships.
Key distinction:
HydroPowerPlant supplemental attribute) describes which units share the same intake pipe for plant-level constraints.Elevations defined on the reservoir and turbine structs are used to support head calculations for downstream simulation packages.
ACTransmission has been introduced to better distinguish between AC transmission objects connected between ACBus objects. Additionally, a new TwoTerminalHVDC abstract type was added to capture HVDC links connected between ACBus objects.PowerSystems.jl provides a rigorous data model using Julia structures for electric energy systems modeling. It is designed to be agnostic to specific mathematical models and serves as the foundation for the following Sienna applications:
Key capabilities include:
.raw, .dyr), MATPOWER, and configurable tabular data (e.g., CSV).