Load forecasting improves by mixing shared and local demand patterns
From Shared Demand Patterns to Local Uncertainty: Probabilistic Load Forecasting by Mixing Compact Adaptations
Machine LearningArtificial Intelligence
Summary
Predicting how much electricity will be used at small scales like homes or transformers is tricky because people’s habits and weather vary a lot. The authors created a method that combines a general model of common usage with small, flexible adjustments for each customer. This keeps the prediction system efficient while handling many unique behaviors. Their tests showed better accuracy and forecasting quality than usual methods, with less computational cost.
What this means in practice
- •For power system operators: Improve local electricity demand predictions at customer and transformer level with scalable models using shared and adapted parameters.
- •For smart grid technology developers: Develop forecasting tools that efficiently balance accuracy and resource use by mixing common demand trends with customer-specific adjustments.
Authors
Haoran Li, Zhe Cheng, Yang Weng
Abstract
Probabilistic load forecasting has been widely studied for power-system operation and planning, but customer- and transformer-level forecasting introduces a distinct scalability challenge. At these levels, load uncertainty is strongly affected by customer behavior, weather, and mixed load composition, making it difficult for a single shared model to capture heterogeneous patterns. Using separate probabilistic models can improve local accuracy, but becomes costly to train, store, update, and validate at scale. To address this challenge, we develop a scalable customer-aware forecasting framework that learns common demand behavior through a shared model while adapting only a compact subset of parameters. Rather than using an independent model for each load or assigning each load to a specialized model, the proposed design learns a small bank of low-dimensional adaptation components and allows each load to combine them according to its forecasting characteristics. This preserves shared knowledge across customers while providing sufficient flexibility for heterogeneous and mixed load compositions. Experiments on 590 load profiles from the SMART-DS dataset show consistent improvements in deterministic accuracy and probabilistic quality over statistical, neural-network, Transformer-based, and pretrained time-series baselines, while retaining low storage and inference costs.