LightGTS-Cov: Covariate-Enhanced Time Series Forecasting

2026-02-11

AI Summary

This work addresses the limitation of existing time series foundation models, which often neglect exogenous covariates or incorporate them through simplistic concatenation, thereby constraining predictive performance in covariate-rich scenarios such as electricity price and renewable energy forecasting. Building upon the lightweight, seasonality-aware LightGTS backbone, we propose a residual MLP plugin that explicitly integrates both historical and future-known covariates to refine the decoder outputs. This mechanism efficiently fuses temporally aligned exogenous information with minimal overhead—adding only approximately 0.1 million parameters. Experimental results demonstrate that the proposed model significantly outperforms LightGTS and other covariate-aware baselines on industrial tasks including electricity price and photovoltaic power forecasting, achieving both high accuracy and robust stability.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Planning, Routing, and Scheduling: Planning with Language Models
• Computer Vision: Multi-modal Vision

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Time series foundation models are typically pre-trained on large, multi-source datasets; however, they often ignore exogenous covariates or incorporate them via simple concatenation with the target series, which limits their effectiveness in covariate-rich applications such as electricity price forecasting and renewable energy forecasting. We introduce LightGTS-Cov, a covariate-enhanced extension of LightGTS that preserves its lightweight, period-aware backbone while explicitly incorporating both past and future-known covariates. Built on a $\sim$1M-parameter LightGTS backbone, LightGTS-Cov adds only a $\sim$0.1M-parameter MLP plug-in that integrates time-aligned covariates into the target forecasts by residually refining the outputs of the decoding process. Across covariate-aware benchmarks on electricity price and energy generation datasets, LightGTS-Cov consistently outperforms LightGTS and achieves superior performance over other covariate-aware baselines under both settings, regardless of whether future-known covariates are provided. We further demonstrate its practical value in two real-world energy case applications: long-term photovoltaic power forecasting with future weather forecasts and day-ahead electricity price forecasting with weather and dispatch-plan covariates. Across both applications, LightGTS-Cov achieves strong forecasting accuracy and stable operational performance after deployment, validating its effectiveness in real-world industrial settings.

Problem

• time series forecasting
• exogenous covariates
• electricity price forecasting
• renewable energy forecasting
• covariate-rich applications

Innovation

• covariate-enhanced forecasting
• lightweight time series model
• residual covariate integration
• future-known covariates
• energy forecasting