An Explainable, Attention-Enhanced, Bidirectional Long Short-Term Memory Neural Network for Joint 48-Hour Forecasting of Temperature, Irradiance, and Relative Humidity

2025/08/28 by Georgios Vamvouras, Vamvouras, Georgios, Konstantinos Braimakis +3
Computer Science · Engineering · Environmental Science · #Artificial Intelligence (cs.AI) #Building Energy and Comfort Optimization #FOS: Computer and information sciences #Hydrological Forecasting Using AI #Machine Learning (cs.LG) #Solar Radiation and Photovoltaics

paper · pdf · doi:10.48550/arxiv.2508.21109

openalex publication_date 2025/08/28 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

This paper presents a Deep Learning (DL) framework for 48-hour forecasting of temperature, solar irradiance, and relative humidity to support Model Predictive Control (MPC) in smart HVAC systems. The approach employs a stacked Bidirectional Long Short-Term Memory (BiLSTM) network with attention, capturing temporal and cross-feature dependencies by jointly predicting all three variables. Historical meteorological data (2019-2022) with encoded cyclical time features were used for training, while 2023 data evaluated generalization. The model achieved Mean Absolute Errors of 1.3 degrees Celsius (temperature), 31 W/m2 (irradiance), and 6.7 percentage points (humidity), outperforming state-of-the-art numerical weather prediction and machine learning benchmarks. Integrated Gradients quantified feature contributions, and attention weights revealed temporal patterns, enhancing interpretability. By combining multivariate forecasting, attention-based DL, and explainability, this work advances data-driven weather prediction. The demonstrated accuracy and transparency highlight the framework's potential for energy-efficient building control through reliable short-term meteorological forecasting.

Citations

• Temporal Pattern Attention for Multivariate Time Series Forecasting
• Axiomatic Attribution for Deep Networks
• CMA-ES for Hyperparameter Optimization of Deep Neural Networks

Related

• Deep Learning for Weather Forecasting: A CNN-LSTM Hybrid Model for Predicting Historical Temperature Data
• Using an Ancillary Neural Network to Capture Weekends and Holidays in an Adjoint Neural Network Architecture for Intelligent Building Management
• Dynamical prediction of two meteorological factors using the deep neural network and the long short term memory (1)
• Predicting the performance of hybrid ventilation in buildings using a multivariate attention-based biLSTM Encoder-Decoder neural network
• Skillful Twelve Hour Precipitation Forecasts using Large Context Neural Networks