Timeseries power model without forecast - ACOPF

Specific Domains Optimization (Mathematical)
, , ,
1

Hello,
I have a model in PowerSystems.jl that looks like this:

image
image1460Ă—1234 46.1 KB

I just want to run an acopf and dcopf with the data I have imported - I do not want to do any forecasting. However, I cannot get past the DecisionModel() as it tells me I do not have any forecasted data. Is there a way I can run an ACOPF without forecasting in PowerSimulations.jl?

2

This is a question for @jdlara-berkeley.

Do you want to solve a single-period OPF problem? You might want to use PowerModels.jl instead.

3

@odow @jdlara-berkeley I went with PowerSimulations.jl since I wanted to run this as a timeseries model (744 hours/timesteps) and it did not seem as though that was the purpose of PowerModels.jl.

1 Like
4

I"m not much help here, I’m afraid.

There’s a slack for Sienna that might be faster than waiting for Jose to look at discourse questions:

5

Thanks @odow - I believe I have the proper set up with network_model_ed = NetworkModel(ACPPowerModel; use_slacks=true, power_flow_evaluation=PowerFlows.ACPowerFlow()). Curious if you have ran into issues with the build!() - I keep getting │ KeyError: key “DENVER 52 1” not found |
yet,

  • for load in get_components(PowerLoad, sys)

  • ts_keys = InfrastructureSystems.get_time_series_keys(load)

  • println("Load: ", load.name, " | Bus: ", load.bus.name, " | Time Series Keys: ", ts_keys)

  • end

prints: Load: DENVER 52 1 | Bus: DENVER 52 1 | Time Series Keys: TimeSeriesKey[ForecastKey(DeterministicSingleTimeSeries, "DENVER 52 1", DateTime("2021-01-01T00:00:00"), Millisecond(3600000), Hour(8760), Hour(1), 1, Dict{String, Any}()), StaticTimeSeriesKey(SingleTimeSeries, "DENVER 52 1", DateTime("2021-01-01T00:00:00"), Millisecond(3600000), 8760, Dict{String, Any}())]
and

  • has_component(PowerLoad, sys, "DENVER 52 1")  # Should return `true`

prints: `true’

6

@joaquimg might be able to answer this too.