Energy Consumption Prediction Research Articles

Transfer learning and source domain restructuring-based BiLSTM approach for building energy consumption prediction

ABSTRACT Currently, building energy consumption prediction typically relies on vast amounts of historical data. However, for newly constructed buildings, the scarcity of data leads to reduced prediction accuracy. To address this challenge, this paper proposes a novel approach that integrates transfer learning with a source domain reconstruction-based BiLSTM model for building energy consumption prediction. In the first stage, both source and target domains are clustered into profile types using k-means. For each profile type in the target domain, the most similar profiles in the source domain are identified using Maximum Mean Discrepancy and Dynamic Time Warping. The source domain is then reconstructed by combining these identified profiles based on their proportions in the target domain. Subsequently, a feature extraction method based on EMD-CWT-Conv is introduced. Empirical Mode Decomposition is first applied to decompose and filter the source domain data. Continuous Wavelet Transform is then employed to extract distinctive frequency-domain and time-domain features from both the reconstructed source domain and the target domain. Final predictions are made using transfer learning and fine-tuning. Experiments on a grocery shop and a school show that the proposed approach reduces Mean Absolute Percentage Error by at least 13.19% and 17.67%, respectively.

A Comparative Study of Building Energy Consumption Prediction Methods

Continued usage of fossil fuel-based electricity generation is one of the main factors for the greenhouse effect. As the greenhouse effect increase, the ambient temperature of the earth will also increase. As the temperature increase, the usage of air conditioning in buildings is also expected to increase. Since that more than 50 % of building energy consumption goes to air conditioning, it is utterly important to look into the methods that can be used to predict the impact of variations in temperature towards building energy consumption. This paper hence tries to perform a conceptual review on the prevalent methods that have been used for prediction of energy consumption, grouped in the category known as the white box, grey box and black box. The comparison of the different methods used in forecasting of energy consumption is discussed in terms of application, input, specific approach, advantage and disadvantage. It is hoped that the work carried out in this paper will be the reference to jump start the research work of others in this field of study.

Application of a decision tree approach to predict energy consumption in lightweight buildings under subtropical climate

PurposeLightweight building systems have emerged as alternatives to reduce the high environmental impact of conventional masonry. However, in subtropical climates, the low thermal inertia of lightweight envelopes negatively affects energy performance. The purpose of this paper is to investigate the thermophysical parameters that influence heating and cooling energy consumption in lightweight residential buildings under subtropical climates and develop a model to predict these parameters using statistical and machine learning tools.Design/methodology/approachA database was created with computer simulation data on the energy performance of 2048 building conditions generated by factorial combination of 10 parameters. Sensitivity analysis was performed to identify which parameters contribute most to energy performance indicators. Subsequently, decision trees were created using a classification and regression tree (CART) algorithm to visualize parameters and improve energy performance indicators, particularly cooling energy consumption.FindingsLow thermal transmittance and ground contact are interesting strategies for low thermal capacity buildings. Furthermore, the findings showed that relying only on the most influential properties does not ensure good energy performance; rather, it is the adequate combination of envelope properties that leads to good energy efficiency. The tree developed by CART can be used as a guide to assist designers and researchers in the initial selection of building envelopes, demonstrating the impact of each choice on electrical energy consumption for indoor climate control.Originality/valueBy adopting a global approach to assess the thermal performance of lightweight buildings, this study makes a significant contribution to synthesizing the results of a complex and time-consuming methodology into a guide for optimizing envelope design decisions and directing efforts and resources toward efficient strategies.

A novel graph convolutional network-based interpretable method for chiller energy consumption prediction considering the spatiotemporal coupling between variables

The prediction of chiller energy consumption is critical for lowering the building's energy consumption. Deep neural networks have been widely applied in this field. However, traditional deep learning methods only consider operational data and do not incorporate empirical knowledge and working condition information that are beneficial for chiller energy consumption prediction. The graph convolutional network (GCN) can handle this problem well, as it is able to incorporate empirical knowledge and working condition information into the network. However, it is important to avoid the excessive influence of empirical knowledge and working condition information on the model's accuracy. To adjust the weight of the influence of the association graph and training set on model accuracy in the GCN network, a coefficient θ was introduced. Herein, a novel GCN (θ-GCN)-based method for the energy consumption prediction of chillers is proposed. This proposed method first combines empirical knowledge and working condition information to obtain the adjacency matrix among operational data and then constructs an association graph based on this adjacency matrix. Secondly, due to the temporal correlation of the collected dataset, a gated recurrent unit (GRU) is used for feature extraction. Lastly, the features extracted through GRU are input into the constructed association graph, and the θ-GCN network is used for training and energy consumption prediction. The proposed method was experimentally verified in an actual chiller system and compared with existing methods; the outcomes indicated that the established method could obtain better prediction accuracy.

An Improved Neural Network Algorithm for Energy Consumption Forecasting

Accurate and efficient forecasting of energy consumption is a crucial prerequisite for effective energy planning and policymaking. The BP neural network has been widely used in forecasting, machine learning, and various other fields due to its nonlinear fitting ability. In order to improve the prediction accuracy of the BP neural network, this paper introduces the concept of forecast lead time and establishes a mathematical model accordingly. Prior to training the neural network, the input layer data are preprocessed based on the forecast lead time model. The training and forecasting results of the BP neural network when and when not considering forecast lead time are compared and verified. The findings demonstrate that the forecast lead time model can significantly improve the prediction speed and accuracy, proving to be highly applicable for short-term energy consumption forecasting.

Modeling of hybrid renewable resources and comprehensive feasibility analysis of grid interactive energy system for commercial building: a case study

ABSTRACT The growing energy demand in commercial buildings and businesses poses significant environmental challenges due to the dependence on fossil fuel-based energy. To address this issue, increasing the use of renewable energy sources and decreasing reliance on fossil fuels is crucial. Commercial buildings, in particular, have many opportunities to harness energy from renewable resources, such as solar, wind, and biomass. This study analyzed historical energy consumption data from a commercial building using MATLAB to forecast future energy demands. A range of machine learning algorithms were employed to accurately predict energy consumption, including Ensemble Boosting, Ensemble Bagging, Decision Trees, Support Vector Machines (SVM), and Neural Networks (NN). A Hybrid Renewable Energy (HRE) system was designed to meet the building’s energy needs, integrating renewable sources with grid-interactive inverters. Performance metrics for the forecasting models were calculated and documented, followed by a detailed techno-economic and environmental feasibility analysis of the HRE system. Different levels of renewable energy penetration were explored to optimize HRE designs using load-following dispatch algorithms tailored to the commercial building’s specific needs. The study revealed that achieving a renewable energy proportion of 12.6% resulted in the lowest cost of electricity (COE) at $0.135 and a Net Present Cost (NPC) of $8.84 million. Additionally, a reliability study using load flow techniques confirmed the stability of the optimized Hybrid Renewable Energy system. The energy costs for varying levels of renewable penetration, ranging from 10% to 40% is also explored in this research. In summary, transitioning to renewable energy sources in commercial buildings can reduce environmental impact while ensuring a sustainable energy supply.

Will it get there? A deep learning model for predicting next-trip state of charge in Urban Green Freight Delivery with electric vehicles

To enhance urban freight efficiency and green development, China has implemented the Urban Green Freight Delivery (UGFD) project, which includes optimizing vehicle traffic control policies and increasing the number of new energy vehicles (NEV). However, range anxiety is a significant challenge for freight drivers performing delivery tasks with electric vehicles (a major component of NEV). We constructed a prediction model for the state of charge (SOC), or battery remaining energy percentage when UGFD vehicles reach the next trip point, aiming to alleviate this issue. The model consists of three modules: (1) a vehicle SOC context prediction module, (2) a vehicle energy consumption prediction module, and (3) a multi-perspective SOC prediction value fusion module. Specifically, in the SOC context prediction module, historical SOC sequences, vehicle status (loading/unloading, charging), and time intervals between SOC points are used to accurately describe context change trends, and directly predict the vehicle SOC at the next trip point. The energy consumption prediction module combines community-level and grid-level geographical location information for the vehicle stops using weather, vehicle parameters, etc., to model the spatial dynamic correlation of energy consumption. The vehicle SOC at the next trip point is the difference between the current vehicle SOC and the predicted energy consumption. The multi-perspective SOC prediction value fusion module is a combination of the predicted values from the context and energy consumption perspectives, resulting in the final vehicle SOC prediction value. Taking Suzhou, China as an example, the results show that the mean absolute error, root mean square error, and symmetric mean absolute percentage error for the constructed model are 23.67%, 10.39%, and 20.03% less, respectively, than for the baseline models focusing on SOC short-term time series prediction. The research results can provide scientific evidence for formulating refined energy management, charging station layout, and freight delivery optimization approaches.

Deformation Evolution and Perceptual Prediction for Additive Manufacturing of Lightweight Composite Driven by Hybrid Digital Twins

This paper proposes a deformation evolution and perceptual prediction methodology for additive manufacturing of lightweight composite driven by hybrid digital twins (HDT). In order to improve manufacturing quality of irregular lightweight composite through boosting conceptual design in aeronautic and aerospace engineering, the HDT meaning hybridization of physical and digital domains, including deformation and energy efficiency can be built, where the essential parameters can be perceptually predicted in advance, by virtue of the fusion of physical sensors and digital information. The long short term memory (LSTM) can be employed to void vanishing gradient problem and improve predicting precision via Recurrent Neural Networks, thereby laying a foundation for the HDT. The diverse manufacturing requirements of different regions are integrated into the parameters designing phase by attaching region weights confirmed via empiricism and in-service simulation. The effects of slicing strategy and external support structures on manufacturing quality are considered from the perspective of improving dimensional accuracy. The manufacturing efficiency and comprehensive costs are accounted as consideration factors, which are perceptually predicted via LSTM. The designed manufacturing parameters through HDT were virtually examined by evaluating the deformation and equivalent stress distributions of fabricated lightweight component with composite material through AM process simulation. The physical experiments were conducted to verify the HDT-based pre-designing and optimization method of manufacturing parameters via fused deposition modeling (FDM). The energy consumption of actual manufacturing process was measured via digital power meter and applied to evaluate accuracy of perceptual prediction outcomes. The dimensional accuracy and distortion distribution of the manufactured lightweight prototype made with composite material were measured through the coordinate measuring machine (CMM) and 3D optical scanner. The proposed method demonstrates effectiveness in improving manufacturing quality and accurately predicting energy consumption, which have been verified with a three-way solenoid valve element, in which the maximum deformation was reduced by 39.78% and the mean absolute percentage error for perceptual prediction was 3.76%.

Life cycle assessment of metal powder production: a Bayesian stochastic Kriging model-based autonomous estimation

Metal powder contributes to the environmental burdens of additive manufacturing (AM) substantially. Current life cycle assessments (LCAs) of metal powders present considerable variations of lifecycle environmental inventory due to process divergence, spatial heterogeneity, or temporal fluctuation. Most importantly, the amounts of LCA studies on metal powder are limited and primarily confined to partial material types. To this end, based on the data surveyed from a metal powder supplier, this study conducted an LCA of titanium and nickel alloy produced by electrode-inducted and vacuum-inducted melting gas atomization, respectively. Given that energy consumption dominates the environmental burden of powder production and is influenced by metal materials’ physical properties, we proposed a Bayesian stochastic Kriging model to estimate the energy consumption during the gas atomization process. This model considered the inherent uncertainties of training data and adaptively updated the parameters of interest when new environmental data on gas atomization were available. With the predicted energy use information of specific powder, the corresponding lifecycle environmental impacts can be further autonomously estimated in conjunction with the other surveyed powder production stages. Results indicated the environmental impact of titanium alloy powder is slightly higher than that of nickel alloy powder and their lifecycle carbon emissions are around 20 kg CO2 equivalency. The proposed Bayesian stochastic Kriging model showed more accurate predictions of energy consumption compared with conventional Kriging and stochastic Kriging models. This study enables data imputation of energy consumption during gas atomization given the physical properties and producing technique of powder materials.

Electric vehicle energy consumption prediction for unknown route types using deep neural networks by combining static and dynamic data

Accurate energy consumption prediction of electric vehicles (EVs) is crucial for drivers considering long trips. All the data should be provided beforehand to determine the energy consumption at the beginning of the trip. Although dynamic vehicle data (vehicle speed, state-of-charge, acceleration, etc.) cannot be known before the trip, factors related to the specified route (route type, elevation, average speed, weather, driving time, etc.) can be used to predict the consumed energy. These factors can be categorized as static and dynamic features, and thus, the question of how to effectively use static and dynamic data arises. This paper investigates the problem of predicting the energy consumption of an EV for a predetermined trip using a deep neural network (DNN) model that effectively uses static features along with dynamic segment features. Furthermore, we address the problem where the route types are unknown in advance. To include more information in the prediction model, we clustered the speed profiles using shape-based clustering with dynamic time warping (DTW) to predict the route type and used the cluster labels as static inputs. Real driving data collected from various drivers of a specific EV were used to train the DNN. The proposed DNN model was compared with the average energy consumption (AEC) model and five machine learning models. The results show that labels obtained from shape-based clustering improved the prediction more than feature-based cluster labels. The prediction errors were minimized with the proposed DNN model, where static features are introduced to the first and second layers twice.

Characterizing energy-related occupant behavior towards urban residential buildings in China with statistical database establishment

Energy-related occupant behavior has significant impacts on the accuracy of energy consumption predictions. However, existing research on urban residential energy-related occupant behavior in China lacks a comprehensive understanding of its characteristics across different climates. Thus, focusing on energy-related occupant behavior of heating, cooling, cooking, domestic hot water use, lighting, operating other electric appliances, window opening, and shading, this paper proposes an index system and data collection method to reveal its inherent spatial and temporal characteristics. Then, the “Chinese Urban Residential Energy-Related Occupant Behavior Database” was established through large-scale investigations involving 8511 valid questionnaires conducted in the Severe Cold Zone, Cold Zone, Hot Summer and Cold Winter Zone, Hot Summer and Warm Winter Zone, and Mild Zone. Furthermore, typical operational patterns of each behavior were refined using K-Means clustering. Results revealed significant seasonal and regional variability in the characteristics of the investigated behaviors in each zone. The typical operational patterns of heating, cooling, window opening, and shading varied by zone and exhibited “part-time, part-space” characteristics. Cooking and electric appliance operational patterns differ depending on local customs. The index system and data collection method for energy-related occupant behavior could potentially serve as a template for establishing behavioral databases in other countries.

Analysis and prediction of energy consumption in office buildings with variable refrigerant flow systems: A case study

Accurately predicting building air-conditioning energy consumption is particularly important for energy management. However, for small datasets, there is a lack of simple and effective prediction methods applicable to multiple buildings. In order to fully explore the energy consumption mode of Variable Refrigerant Flow (VRF) air conditioning system and improve the accuracy of the model. This study uses the weather data of an outdoor meteorological data collection system and the air conditioning system and energy consumption data of an office building in Jinan. Initial analysis revealed that occupant behavior and daily variations in outdoor air temperatures significantly impact the energy consumption of air conditioning systems. Then, by applying the random forest algorithm and Pearson correlation analysis, the study identified daily average outdoor air temperature and VRF indoor unit running time as critical features for prediction. Subsequently, three machine learning models—Multiple Linear Regression (MLR), Back Propagation Neural Network (BPNN), and Long Short-Term Memory (LSTM) neural network—were developed and assessed using a 4:1 training-to-testing set ratio. Comparative analysis revealed that all models performed robustly, with the MLR model exhibited superior predictive accuracy (R2 (Coefficient of Determination) = 0.98, MAE (Mean Absolute Error) = 0.005 kWh/(m2·day), and RMSE (Root Mean Square Error) = 0.007 kWh/(m2·day) for the testing set) and simplicity. Furthermore, the effectiveness of the MLR model in handling small sample sizes was confirmed through cross-validation of data from two additional office buildings, highlighting its suitability for predicting energy consumption of VRF air conditioning systems in office buildings.

Real-time optimal hierarchy control of HVAC system with maximized ancillary service support in smart buildings: An experimental approach

This article proposes a real-time optimum control algorithm for Heating, Ventilation, and Air Conditioning (HVAC) units of a commercial building aimed to maximize the ancillary power available for frequency regulation services. First, the prediction of energy consumption of air conditioning zones using long short-term memory (LSTM) is achieved, which helps forecast the aggregated regulation capacity bid of the building in the energy markets. Furthermore, to minimize energy costs and thermal discomfort of building dwellers, an optimization algorithm is designed considering the day-ahead electricity tariffs. Moreover, using installed sensors in a lab-scaled hardware prototype, the proposed algorithm is also shown to be capable of interacting with a Building Energy Management System (BEMS) and investigating the electrical and thermal properties of the HVAC unit. The BEMS controller uses an ethernet protocol to interface with the sensors installed at various operational points throughout the HVAC system. The optimal operation of the HVAC system has been executed with the real-time thermal feedback system, which counterbalances the uncertain thermal load or renewable power availability in the building confronted in real-time operation. The simulation results show the performance of the Artificial Neural Network (ANN) based compressor model in deep learning integrated with the other components of the HVAC unit and thermal model of the air-conditioning zone. Finally, the experimental results display the real-time implementation of multiple thermal and electrical conditions within the proposed control scheme of the HVAC unit. This shows the potential of ancillary services capacity through commercial buildings can be maximized with flexible loads and thereby help in power grid support.

Optimizing energy efficiency through building orientation and building information modelling (BIM) in diverse terrains: A case study in Pakistan

Building orientation plays a key role in enhancing the built environment's energy efficiency through the application of Building Information Modelling (BIM) and energy simulation. Due to their significant global energy consumption, buildings are the focus of energy optimization efforts. This study explores how building orientation, when combined with modern methods like BIM and energy modelling, affects energy usage and savings throughout Pakistan's numerous geographic terrains. The methodology is quantitative, beginning with an extensive review of literature on energy analysis, BIM applications, and urban building energy modelling. Autodesk Insight 360 is employed for energy simulations, which enables the precise prediction of energy consumption by considering various factors such as building orientation, window-to-wall ratios, shading, wall and roof construction, infiltration rates, lighting efficiency, occupancy controls, plug load efficiency, and HVAC systems. Energy simulation of the data indicates that optimizing building orientation alone can result in an average energy savings of 18 %, while combining orientation optimization with improvements in window arrangements and construction materials can achieve savings of up to 30 % over 30 years. The energy simulation results were validated through case study representing typical terrains and building types (commercial and residential), with simulated energy expenditures showing a high correlation (R2 = 0.92) with actual utility bills, adjusted for local energy rates. The findings of this study highlight substantial financial benefits, with potential annual savings ranging from $2500 to $4000 for residential buildings and $10,000 to $15,000 for commercial buildings, depending on building size and location. These results suggest broader applicability to diverse geographical areas and building types in future research. This research contributes significantly to energy-efficient building practices, paving the way for sustainable construction and reduced environmental impact across the varied landscapes of Pakistan.

Energy consumption forecasting: a case study on Bhashan Char island in Bangladesh

Energy consumption forecasting: a case study on Bhashan Char island in Bangladesh

Integrating autoencoder and decision tree models for enhanced energy consumption forecasting in microgrids: A meteorological data-driven approach in Djibouti

At this time, as the world and nations move to reduce the use of fossil fuels, research is oriented toward improving the energy consumption of people and buildings. Recent methods, mainly computing techniques such as deep learning, are proposed in the literature. This paper proposes a model that integrates the autoencoder with the decision tree model using six months of meteorological data, including solar radiation, wind speed, temperature, and other meteorological data. This study aims to advance the energy consumption prediction of loads connected to the microgrid. A case study of a microgrid park of a campus in Djibouti is presented to test the proposed model. Autoencoders were exploited to extract the main features of the meteorological data. Then the decision tree is proposed to predict the energy consumption using the resulting encoded features. The evaluation of the training of the autoencoder model gave a favorable result with a mean squared error of 0.002 and an R2 value of 0.99. Hyperparameter tuning scenarios facilitated the exploration of the decision tree model. Ensemble decision trees performed better than individual trees in this model, achieving a mean absolute error of 1.4 % and an R2 value of 0.997. It showed that hyperparameter tuning improved the results due to the best architecture fit for the decision tree. Moreover, the proposed model was validated by comparing it with the literature on KNN, SVR, and MLP models commonly used in microgrid energy management. The autoencoder decision tree outperformed other compared methods, achieving an explained variance of 0.997 and an MAE of 1.7 % in the standard decision tree regressor scenario. These results demonstrated the capabilities of machine learning and weather data. Combining an autoencoder and the decision tree model will open a new door to energy management improvement.

Physical, socio-psychological, and behavioural determinants of household energy consumption in the UK

Determining which attitudes and behaviours predict household energy consumption can help accelerate the low-carbon energy transition. Conventional approaches in this domain are limited, often relying on survey methods that produce data on individuals’ motivations and self-reported activities without pairing these with actual energy consumption records, which are particularly hard to collect for large, nationally representative samples. This challenge precludes the development of empirical evidence on which attitudes and behaviours influence patterns of energy consumption, thus limiting the extent to which these can inform energy interventions or conservation programs. This study demonstrates a novel methodology for estimating energy consumption in the absence of actual energy records by using a large, publicly available data set of energy consumption in the UK. We develop a predictive model using the Smart Energy Research Laboratory (SERL) data portal (with records from nearly 13,000 UK households) and then use this model to predict energy consumption (both electric and gas) for a sample of 1,000 UK householders for which we separately collect over 200 variables relating to climate change attitudes and practices. Our approach uses a set of over 50 independent variables that are shared between the data sets, allowing us to train a model on the SERL data and use it to analyse the relationship between energy consumption and the opinions, motivations, and daily practices of survey respondents. Results show that electricity consumption is influenced by a broader range of factors compared to gas. Household energy use is best explained by physical dwelling characteristics, socio-demographic variables, and certain behavioural and attitudinal measures. Notably, pro-environmental attitudes, frugality, and conscientiousness correlate with lower energy use, while income and consumerism are linked to higher consumption. We discuss how these findings can inform efforts to decarbonise home energy use in the UK.

Identification and control of hydrothermal carbonisation process with energy consumption assessment

The sustainable and cost-effective production of hydrochar requires energy efficient processing. In this study, data collected from laboratory-scale hydrothermal carbonisation experiments were used to identify dynamic models to predict reactor temperature based on estimated heating power. The performance was then analysed through process simulations with model predictive control. The optimal control model exhibited root mean squared error (RMSE) in the range of 3.19–11.47 °C in model validation with data sets of 31 experiments, whereas for process models it was 1.49–4.79 °C using their training data. In simulations with the optimal control model at the target temperatures of 180–200 °C for 4–6 h, the optimal mean values of mean absolute percentage error (MAPE), overshoot, and specific energy consumption were 0.56%, 2.50 °C, and 40.86 kWh/kg, respectively. Simulations indicated also that strict temperature control settings (average standard deviation s = 1.24 °C at target temperature) increased specific energy consumption slightly (x¯ = 4.4%) when compared with control settings that allowed higher variance around a setpoint (s = 5.11 °C). However, the energy consumption was generally more dependent on the target temperature and processing time, which was also validated based on the measurements of 23 additional experiments. The presented dynamic modelling approach enables accurate real-time process control and predictive energy consumption optimisations for hydrothermal carbonisation.

Anomaly detection and prediction of energy consumption for smart homes using machine learning

AbstractAs technology advances, smart homes are being increasingly adopted, thus generating massive data that pose new research challenges. We propose a machine learning framework for monitoring energy consumption in smart home devices. The proposed framework involves an anomaly detection module, followed by a predictive model to forecast energy consumption patterns in a typical smart home. We employ three outlier‐based techniques for anomaly detection: (1) local outlier factor, (2) connectivity‐based outlier factor, and (3) cluster‐based local outlier factor. Furthermore, we apply random forest, linear regression, decision tree, and the ensemble techniques of adaptive, gradient, and extreme gradient boosting to anomaly free data to develop baseline models that predict the energy consumption patterns of smart home devices. The framework is evaluated on three publicly available energy datasets collected from various smart homes. The experimental results reveal that the cluster‐based local outlier factor with extreme gradient boosting achieves promising results with high prediction accuracy.

Enhanced Data Processing and Machine Learning Techniques for Energy Consumption Forecasting

Energy consumption plays a significant role in global warming. In order to achieve carbon neutrality and enhance energy efficiency through a stable energy supply, it is necessary to pursue the development of innovative architectures designed to optimize and analyze time series data. Therefore, this study presents a new architecture that highlights the critical role of preprocessing in improving predictive performance and demonstrates its scalability across various energy domains. The architecture, which discerns patterns indicative of time series characteristics, is founded on three core components: data preparation, process optimization methods, and prediction. The core of this architecture is the identification of patterns within the time series and the determination of optimal data processing techniques, with a strong emphasis on preprocessing methods. The experimental results for heat energy demonstrate the potential for data optimization to achieve performance gains, thereby confirming the critical role of preprocessing. This study also confirms that the proposed architecture consistently enhances predictive outcomes, irrespective of the model employed, through the evaluation of five distinct prediction models. Moreover, experiments extending to electric energy validate the architecture’s scalability and efficacy in predicting various energy types using analogous input variables. Furthermore, this research employs explainable artificial intelligence to elucidate the determinants influencing energy prediction, thereby contributing to the management of low-carbon energy supply and demand.