Cyclical Temporal Encoding and Hybrid Deep Ensembles for Multistep Energy Forecasting

2025-12-03

AI Summary

To address insufficient accuracy in multi-step power load forecasting, this paper proposes a unified framework integrating periodic time encoding with a hybrid deep learning ensemble. Methodologically, it employs sine-cosine encoding to explicitly model temporal periodicity, jointly leveraging LSTM for long-term temporal dependency capture, CNN for local pattern extraction, and an MLP-based meta-learner for adaptive multi-step prediction fusion; a step-specific meta-learning mechanism is further designed to tailor predictions across varying horizons. The key contribution lies in the first systematic integration—within short-term energy forecasting—of periodic time encoding, calendar-aware features, and a hybrid ensemble architecture, rigorously validated via correlation analysis and ablation studies. Experiments on national-scale annual electricity consumption data demonstrate that the proposed model consistently outperforms state-of-the-art methods across seven forecasting horizons, achieving average reductions of 12.6% in RMSE and 10.3% in MAE, thereby significantly enhancing consistency and robustness in multi-step forecasting.

Technology Category

• Machine Learning: Ensemble Methods
• Search and Optimization: Metareasoning and Metaheuristics
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments

Abstract

Accurate electricity consumption forecasting is essential for demand management and smart grid operations. This paper introduces a unified deep learning framework that integrates cyclical temporal encoding with hybrid LSTM-CNN architectures to enhance multistep energy forecasting. We systematically transform calendar-based attributes using sine cosine encodings to preserve periodic structure and evaluate their predictive relevance through correlation analysis. To exploit both long-term seasonal effects and short-term local patterns, we employ an ensemble model composed of an LSTM, a CNN, and a meta-learner of MLP regressors specialized for each forecast horizon. Using a one year national consumption dataset, we conduct an extensive experimental study including ablation analyses with and without cyclical encodings and calendar features and comparisons with established baselines from the literature. Results demonstrate consistent improvements across all seven forecast horizons, with our hybrid model achieving lower RMSE and MAE than individual architectures and prior methods. These findings confirm the benefit of combining cyclical temporal representations with complementary deep learning structures. To our knowledge, this is the first work to jointly evaluate temporal encodings, calendar-based features, and hybrid ensemble architectures within a unified short-term energy forecasting framework.

Problem

• Enhancing multistep electricity consumption forecasting accuracy
• Integrating cyclical temporal encoding with hybrid LSTM-CNN architectures
• Evaluating temporal encodings and ensemble models for energy prediction

Innovation

• Cyclical temporal encoding with sine cosine functions
• Hybrid LSTM-CNN ensemble with meta-learner MLP
• Unified framework integrating encodings, features, and architectures