Market-Information-Aware Gated-LoRA of Foundation Models for Transferable Day-Ahead Electricity Price Forecasting

2026-08-11

AI Summary

This study addresses the challenge of day-ahead electricity price forecasting in data-scarce or newly established power markets, where insufficient historical data hinders effective model training. To overcome this limitation, the authors propose a cross-market transfer approach based on the Chronos-2 time-series foundation model. By constructing a multi-source market information interface and introducing a Gated Low-Rank Adapter (Gated-LoRA) that integrates structured market inputs with operational state dependencies, the method enables parameter-efficient fine-tuning without requiring labeled data from the target market. Evaluated under a leave-one-market-out protocol across four Chinese provincial spot markets, the proposed approach reduces average MAE and RMSE by 6.24%/7.99% compared to an information-aware zero-shot Chronos-2 baseline and by 3.05%/3.52% relative to standard Source-LoRA, demonstrating significantly improved forecasting performance in data-limited scenarios.

Abstract

Electricity price forecasting is crucial for market participants but remains difficult because prices are volatile, market-specific, and closely tied to anticipated system conditions. Existing supervised methods depend largely on market-specific historical data, limiting their use in newly established or data-scarce markets. This paper proposes a market-information-aware adaptation framework that transfers the Chronos-2 time-series foundation model to day-ahead electricity price forecasting. It first constructs a multi-source market information (MSMI) interface aligning 7-day price context with pre-clearing supply--demand, reserve, maintenance, generator-capacity, and intertie variables, and then trains a source-domain gated low-rank adapter (LoRA), updating about $1\%$ of model parameters without target-market labels. The gate scales the frozen source adapter according to reserve-tightness and operating-state signals. A leave-one-market-out protocol is adopted for evaluating cross-market transferability. Experiments on four Chinese provincial day-ahead spot markets show that the proposed framework reduces the average MAE/RMSE by $6.24\%/7.99\%$ relative to market-information-aware zero-shot Chronos-2 and by $3.05\%/3.52\%$ relative to vanilla Source-LoRA. Experiments show that the gain is not reproduced by a learned global scalar or by random gate initialization, while the additional improvement over Source-LoRA is limited. These results suggest that market-structured inputs and state-dependent gated LoRA can provide a practical transfer path for data-scarce electricity markets.

Problem

• electricity price forecasting
• data-scarce markets
• transfer learning
• market-specific volatility
• cross-market generalization

Innovation

• Gated-LoRA
• market-information-aware adaptation
• foundation model transfer
• day-ahead electricity price forecasting
• multi-source market information