Aug 2026· Sustainability· Vol 18, pp. 8479· 0 citations· 57 references
Abstract
Feature engineering has become a common step in AI-based HVAC forecasting, often involving variables calculated from raw building management system (BMS) measurements, such as temperature differences, setpoint tracking deviations, airflow balance indicators, rolling statistics, and temporal or seasonal descriptors. Accurate short-term forecasting can provide a baseline of expected operation for anomaly and fault detection and can support control optimization and operator decision making. However, real-world deployment is complicated due to differences in BMS sensor availability and data quality, as well as the preprocessing and maintenance burden associated with complex feature sets. The actual contribution of these features to the performance of AI forecasting remains underexplored, particularly for short-term prediction of air handling unit (AHU) operation. This study evaluates the impact of features on short-term AHU forecasting using three deep learning (DL) architectures: Temporal Convolutional Networks (TCNs), Long Short-Term Memory (LSTM) networks, and a hybrid CNN–LSTM model. An actual operational AHU dataset from a BMS was used to predict key operational variables, including supply and extract air temperatures, supply and extract fan operating signals, and supply air temperature setpoint-tracking error. Fan signal balance was additionally evaluated as a derived indicator calculated from the two predicted fan signals. Four input configurations were evaluated: (i) full (74 inputs), containing raw BMS measurements, short-cycle temporal variables, engineered and dynamic features, and annual-calendar information; (ii) no annual calendar (68 inputs), identical to full but excluding annual-calendar variables; (iii) raw + short-cycle temporal (20 inputs); and (iv) raw-only (12 inputs). The models used a 60-min input history to forecast the following 30-min at one-minute resolution. Persistence and Ridge models were included as reference baselines. All models were trained and tested on identical data splits and forecasting horizons to ensure a fair comparison. Each DL experiment was repeated across five independent runs, and performance was evaluated using MAE, RMSE, and R2. The TCN showed the strongest overall DL performance. Raw-only achieved the highest mean R2 in 11 of 15 architecture–target comparisons using just 12 inputs. The best mean DL R2 ranged from 0.916 for the fan signals to 0.993 for extract air temperature. Annual-calendar features improved the TCN results but provided no consistent benefit for the LSTM or CNN–LSTM. Ridge slightly outperformed the best DL configurations for temperature-related targets, reflecting the strong short-term continuity of these signals. These findings show that recent raw BMS measurements contain most of the information needed for accurate 30-min AHU forecasting, while explicit seasonal and engineered features provide limited additional value. The resulting simpler models may in the future be used as forecasting components in predictive control and fault detection systems. However, their control and energy-saving benefits must be tested separately.
Reliable power supply units are essential for Automatic Weather Stations (AWS) to maintain continuous data collection. However, traditional maintenance schedules often fail to prevent sudden equipment downtime. While machine learning can enable predictive maintenance, standard standalone models typically struggle to capture both immediate short-term anomalies and slow, long-term degradation. To address this gap, this study aims to evaluate and propose a hybrid Temporal Convolutional Network (TCN) and Long Short-Term Memory (LSTM) architecture specifically designed for AWS power supply forecasting. Using empirical time-series data, we monitored five operational parameters at 10-minute intervals from September 2023 to November 2024. Correlation analysis established battery temperature as a primary health indicator due to its strong inverse relationship with voltage (r = –0.87). Comparative evaluations demonstrated that while individual TCN and LSTM models exhibited architectural trade-offs, the proposed hybrid TCN-LSTM model achieved the highest predictive accuracy (R² = 0.9497; MAPE = 0.05%). The findings confirm that integrating these networks effectively balances rapid anomaly detection with stable long-term trend forecasting. Practically, this hybrid model can be integrated into AWS telemetry systems as a robust diagnostic tool, providing automated early warnings to prevent critical power failures.
Marzuki Sinambela, Rifqi Daffa Ul haq, Dibyo Susanto et al.· Indonesian Physical Review· 0 citations
HVAC systems use up about half of the total energy in smart buildings and are a key focus of optimization. The demand of HVAC energy is very difficult to forecast with high accuracy due to the nonlinear nature of HVAC operations, high temporal variability, and interdependencies among environmental and operational variables. Traditional forecasting methods like regression based models and ARIMA often do not reflect such multivariate dependencies resulting in incompetent energy management. This paper presents a multivariate Long Short-Term Memory (LSTM) model that will be developed to learn the long-term temporal dynamics of various variables related to HVAC. The model is trained and tested on a real-world benchmark dataset, which includes 11 sensor-derived features, and uses one fully connected LSTM layer with 50 hidden units trained using the Adam algorithm. Root Mean Square Error (RMSE) and the coefficient of determination (R2) are reported per variable as measures of forecast performance. The experimental findings indicate that the model is accurate, over 90% on most variables, a fact that justifies the fact that the model is effective in overcoming the weaknesses of the traditional methods and giving accurate predictions that can be incorporated into smart building energy management systems. Further research will focus on hybrid deep learning networks and TinyML networks to run on edge devices that are IoT-enabled.
Ali Abdullah.A.A Alsqaff, N. Alduais, Abdul-Malik H. Y. Saad et al.· 2026 6th International Confe...· 0 citations
This study compares ARIMA, LSTM, and temporal fusion transformer (TFT) models across three applications and shows that TFT consistently achieved superior forecasting performance and demonstrated greater robustness to increasing missingness, while k-NN generally provided the most effective imputation performance across datasets.
M. Hosseini, Mohamad Forouzanfar· Computer Science and Informa...· 0 citations
Flowchart summarizing the research workflow including data preprocessing, feature engineering, model development, validation strategy, and evaluation and analysis.
Short-term water demand forecasting plays a pivotal role in the efficient management and optimization of urban water systems. This study investigates the performance of six forecasting architectures (XGBoost, Random Forest, LightGBM, support vector regression (SVR), long short-term memory (LSTM), and Prophet) to assess whether complex deep learning methods yield superior results compared with advanced ensemble approaches driven by rigorous, domain-informed feature engineering. Leveraging approximately 35,000 hourly water demand observations coupled with ERA5-Land meteorological data, the forecasting task was formulated as a supervised learning problem. A rigorous feature engineering framework was proposed, extracting 27 informative predictors, including cyclical temporal encodings, 168-h autoregressive lags, dynamic anchor features, and weather-related regressors. Model evaluation revealed that tree-based gradient boosting algorithms fundamentally outperformed deep learning (LSTM) and structural (Prophet) models. XGBoost delivered the highest aggregate accuracy (R2 = 0.966, RMSE = 0.394), forming a statistically significant high-performance cluster with LightGBM, Random Forest, and SVR. Furthermore, a hierarchical Shapley additive explanations interpretability analysis demonstrated that the 1-week autoregressive lag and dynamic anchor features were the primary drivers of prediction, while meteorological variables acted as secondary microtuning parameters. Finally, introducing a novel residual-engineered hybrid architecture (e.g., Hybrid_XGB_XGB) yielded further statistically significant improvements, establishing a highly accurate and computationally efficient framework for operational forecasting.
Sadegh Yaghoobzadeh, M. Naseri· Water Practice & Technol...· 0 citations
The accelerating impact of climate change on meteorological dynamics and air quality across India poses a pressing challenge for urban sustainability and public health resilience. In this study, we present a data-driven AI framework based on Bidirectional Long Short-Term Memory (BiLSTM) networks to forecast PM₂.₅ concentrations using multivariate environmental data, including temperature, humidity, wind speed, UV index, and particulate matter levels. Comparative analysis with both conventional deep learning models (LSTM, GRU) and statistical baselines (ARIMAX, MLR) demonstrates the BiLSTM’s superior capacity in learning long-range temporal dependencies. Among the evaluated models, the proposed BiLSTM framework achieved the highest predictive performance with an
R
2
of 0.8113, MAE of 0.2988, and RMSE of 0.4359. Consequently, the model supports not only enhanced predictive accuracy but also operational daily, localized decision-making for air quality management, offering actionable insights for climate-aware urban planning and smart city governance. By integrating advanced AI techniques with high-dimensional environmental datasets, this work underscores the transformative role of computational intelligence in shaping adaptive, evidence-based sustainability strategies for future-ready cities.
Vinayak Gupta, Yajnaseni Dash, Ankush Goyal et al.· Frontiers in Climate· 0 citations
A framework for a comparative evaluation of four representative forecasting methods: the Seasonal Autoregressive Integrated Moving Average (SARIMA) model, Extreme Gradient Boosting, the Long Short-Term Memory (LSTM) neural network, and a hybrid Variational Mode Decomposition–LSTM (VMD-LSTM) model is proposed.
Pratiman Patel, Prajwal Pal· International Journal for Re...· 0 citations