forecast_Gaussian_intrahour: Do intra-hourly Gaussian error distribution forecast

In kdayday/solarbenchmarks: Intra-Hour and Intra-Day Probabilistic Solar Forecast Benchmarks

Description

Generates an intra-hourly Gaussian error distribution forecast using a doubly truncated Gaussian distribution. The distribution's mean is a smart persistence forecast, based on the most recent clear-sky index before each hourly issue time. The distribution's standard deviation is calculated as the standard deviation of the smart persistence errors over the past few hours.

Usage

forecast_Gaussian_intrahour(GHI, percentiles, sun_up, clearsky_GHI,
  ts_per_hour, nhours)

Arguments

GHI	A vector of the telemetry
percentiles	A vector of the percentiles corresponding to the desired forecast quantiles
sun_up	A vector of logicals, indicating whether the sun is up
clearsky_GHI	a vector of clear-sky irradiance estimates
ts_per_hour	Time-steps per hour, e.g., 12 for a 5-minute resolution forecast
nhours	Number of preceeding hours to collect training errors from, e.g., 1 or 2

Details

Valid for intra-hourly forecasts. For hourly-resolution forecasts, see forecast_Gaussian_hourly. Distribution is truncated at 0 on the low end and clear-sky GHI on the upper end. While training data is being collected at the beginning of the day, forecast starts as deterministic clear-sky forecast and gathers forecast errors starting after the 1st hour. This function is for hindcasting only, and uses the same clear-sky GHI estimates for both the historical and forecast values.

Value

a matrix of quantile forecasts at each valid time in the input data

See Also

Other forecast functions: forecast_CH_PeEn, forecast_Gaussian_hourly, forecast_NWP, forecast_PeEn_hourly, forecast_PeEn_intrahour, forecast_climatology, forecast_mcm

kdayday/solarbenchmarks documentation built on May 22, 2020, 10:33 p.m.