Neural Architecture Search for global multi-step Forecasting of Energy Production Time Series

2025-10-27

AI Summary

Balancing prediction accuracy, computational efficiency, and cross-scenario generalization remains challenging for short-term forecasting in dynamic energy systems. Method: This paper proposes an automated neural architecture search (NAS) framework tailored to energy time series. It introduces a lightweight, time-series-aware search space and a novel multi-objective optimization function jointly penalizing multi-step prediction error, inference latency, and generalization robustness. Crucially, it integrates NAS with ensemble learning to automatically discover and combine optimal lightweight architectures. Results: Extensive experiments demonstrate that the proposed method significantly outperforms state-of-the-art Transformer-based and pre-trained models on global multi-step short-term forecasting tasks. It achieves simultaneous improvements in prediction accuracy, inference speed, and cross-scenario generalization—thereby offering strong practical deployability for real-world energy system applications.

Technology Category

• Search and Optimization: Learning to Search
• Machine Learning: Deep Neural Architectures and Foundation Models
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

The dynamic energy sector requires both predictive accuracy and runtime efficiency for short-term forecasting of energy generation under operational constraints, where timely and precise predictions are crucial. The manual configuration of complex methods, which can generate accurate global multi-step predictions without suffering from a computational bottleneck, represents a procedure with significant time requirements and high risk for human-made errors. A further intricacy arises from the temporal dynamics present in energy-related data. Additionally, the generalization to unseen data is imperative for continuously deploying forecasting techniques over time. To overcome these challenges, in this research, we design a neural architecture search (NAS)-based framework for the automated discovery of time series models that strike a balance between computational efficiency, predictive performance, and generalization power for the global, multi-step short-term forecasting of energy production time series. In particular, we introduce a search space consisting only of efficient components, which can capture distinctive patterns of energy time series. Furthermore, we formulate a novel objective function that accounts for performance generalization in temporal context and the maximal exploration of different regions of our high-dimensional search space. The results obtained on energy production time series show that an ensemble of lightweight architectures discovered with NAS outperforms state-of-the-art techniques, such as Transformers, as well as pre-trained forecasting models, in terms of both efficiency and accuracy.

Problem

• Automating neural architecture design for energy production time series forecasting
• Balancing computational efficiency with predictive accuracy in forecasting
• Ensuring model generalization across temporal dynamics in energy data

Innovation

• Automated neural architecture search for energy forecasting
• Efficient search space capturing energy time series patterns
• Novel objective function ensuring generalization and exploration