Fast Probabilistic Wind Power Forecasting via One-Step Consistency Distillation and Memory-Efficient CRPS Optimization
- 1School of Information and Intelligent Science, Donghua University, Shanghai 201620, China
- 2High Speed Aerodynamics Institute, China Aerodynamics Research and Development Center, Mianyang 621000, China
Abstract
Grid operators scheduling wind power act on the width of the predicted distribution, and diffusion-based forecasters have reported strong accuracy on wind benchmarks. That accuracy can come from a sampling chain of a hundred or more evaluations per member, with reported implementations incurring multi-second latency per farm against a dispatch cycle of minutes, while a one-step student distilled from such a chain can concentrate its members into a spread narrower than its own error. This paper introduces Consist-Wind, a one-step probabilistic wind power forecaster. A frozen point backbone supplies the conditioning context; a teacher diffusion model on numerical weather prediction (NWP) residuals is distilled into a one-step consistency student, one function evaluation per member, and a stabilized shifted-logit warp maps generated members to the physical power range. A fourth phase calibrates the student by training its ensemble against the continuous ranked probability score (CRPS), which recovers the spread and raises the accuracy of the one-step forecast. Evaluated at every update, its pairwise form governs the training footprint, so the sorted-rank form of the CRPS is used, quasilinear in time and linear in memory. The cost of calibration is decoupled from the ensemble width by a scalable ensemble calibration principle, with the objective on order statistics and query-chunked attention (QCA) inside the kernel launch limit, so the whole pipeline trains and runs on a single device with limited compute. In our experiments on GEFCom2014-W, SDWPF, and GEFCom2012-W, Consist-Wind achieves the lowest CRPS, 3.1% below the strongest reproduced baseline, reduces the calibration error of the distilled student by 45%, issues a forecast in 0.03 ms for a speedup of about 66× relative to the reproduced multi-step reference, and reduces the memory required to evaluate the pairwise penalty term by 98%, admitting 500-member calibration ensembles on the evaluation device. Distributional forecasts of this accuracy, trained within a limited-compute budget, bring uncertainty-aware scheduling within reach of operators on commodity hardware.
Related
- Diffusion Distillation for Efficient Weather Ensembles
- SDECast: Probabilistic Weather Forecasting in Continuous Time with Neural SDEs
- Probabilistic Forecasting of Electric Vehicle Charging in a Time-varying Hierarchical Infrastructure
- Combining Weather Forecast Aggregation and State-Space Models for Adaptive Probabilistic Electricity Load Forecasting
- From morning departures to evening returns: a probabilistic model for forecasting bike-sharing demand
BibTeX
@article{he2026fast,
title = {{Fast Probabilistic Wind Power Forecasting via One-Step Consistency Distillation and Memory-Efficient CRPS Optimization}},
author = {Chao He and Fangqi Zhou and Yu Wang and Hao Huang and Zijian Wang and Jianbo Zheng},
journal = {Mathematics},
year = {2026},
month = oct,
doi = {10.3390/math14203687},
url = {https://www.mdpi.com/2227-7390/14/20/3687},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Fast Probabilistic Wind Power Forecasting via One-Step Consistency Distillation and Memory-Efficient CRPS Optimization
AU - Chao He
AU - Fangqi Zhou
AU - Yu Wang
AU - Hao Huang
AU - Zijian Wang
AU - Jianbo Zheng
PY - 2026
DA - 2026/10/11
JO - Mathematics
DO - 10.3390/math14203687
UR - https://www.mdpi.com/2227-7390/14/20/3687
AB - Grid operators scheduling wind power act on the width of the predicted distribution, and diffusion-based forecasters have reported strong accuracy on wind benchmarks. That accuracy can come from a sampling chain of a hundred or more evaluations per member, with reported implementations incurring multi-second latency per farm against a dispatch cycle of minutes, while a one-step student distilled from such a chain can concentrate its members into a spread narrower than its own error. This paper introduces Consist-Wind, a one-step probabilistic wind power forecaster. A frozen point backbone supplies the conditioning context; a teacher diffusion model on numerical weather prediction (NWP) residuals is distilled into a one-step consistency student, one function evaluation per member, and a stabilized shifted-logit warp maps generated members to the physical power range. A fourth phase calibrates the student by training its ensemble against the continuous ranked probability score (CRPS), which recovers the spread and raises the accuracy of the one-step forecast. Evaluated at every update, its pairwise form governs the training footprint, so the sorted-rank form of the CRPS is used, quasilinear in time and linear in memory. The cost of calibration is decoupled from the ensemble width by a scalable ensemble calibration principle, with the objective on order statistics and query-chunked attention (QCA) inside the kernel launch limit, so the whole pipeline trains and runs on a single device with limited compute. In our experiments on GEFCom2014-W, SDWPF, and GEFCom2012-W, Consist-Wind achieves the lowest CRPS, 3.1% below the strongest reproduced baseline, reduces the calibration error of the distilled student by 45%, issues a forecast in 0.03 ms for a speedup of about 66× relative to the reproduced multi-step reference, and reduces the memory required to evaluate the pairwise penalty term by 98%, admitting 500-member calibration ensembles on the evaluation device. Distributional forecasts of this accuracy, trained within a limited-compute budget, bring uncertainty-aware scheduling within reach of operators on commodity hardware.
SN - 2227-7390
ER -