Download presentation
Presentation is loading. Please wait.
1
by quantile regression
Wind forecasting by quantile regression Dr. Geoffrey Pritchard University of Auckland
3
Short term (within 2 hours)
The persistence forecast (“no change”) is hard to beat by much. Important to indicate situation-dependent uncertainty awareness of risks probabilistic forecast: scenarios, full pdf
4
MetService ePD - the kitchen-sink approach
5
Persistence forecast: 30min x 5
-2hr +30min TP-5 TP-4 TP-3 TP-2 TP-1 TP Half-hourly data Wind forecast is actual output in TP-5 Actual wind observed
6
72149 observations (~4 years)
7
72149 observations (~4 years)
10
Quantile regression Model for the t-quantile (0 < t < 1) of the conditional distribution of a response variable: Q(t) = S b i (t) xi coefficients explanatory variables Our xi will be (nonlinear) spline basis functions of current power.
12
Quantile regression fitting
To fit observations yi : minimize S rt(yi - b i (t) xi ) where rt is the function t-1 t Reduces to linear programming.
15
Additional predictor variables
Improve (?) quantile models by adding terms with wind direction barometric pressure time of day recent power variability etc. In single-scenario forecasting, get little improvement on persistence. In scenario generation, extra variables may help identify low- and high-uncertainty situations.
18
91 34 34
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.