Energy Management Tools Research Articles

Predicting Electricity Consumption using an Innovative Deep Energy Predictor Model for Enhanced Accuracy and Efficiency

Accurate forecasting of electricity consumption is crucial for efficient energy management and planning. Traditional models often struggle to capture the complex, nonlinear relationships inherent in energy consumption data, leading to less reliable predictions. In the quest to improve the accuracy and efficiency of electricity consumption predictions, we introduce the Deep Energy Predictor Model (DEPM). This innovative hybrid model combines the strengths of XGBoost for classification and Deep Neural Networks (DNN) for deep learning (DL) capabilities. The model leverages advanced feature extraction techniques using Cascaded ResNet, ensuring the capture of intricate patterns within the data. Implemented in Python, the DEPM highlights an impressive accuracy of 98%, with a precision of 0.97, recall of 0.99, and an F1-score of 0.98, setting a new standard for predictive models in the energy sector. This study elaborates on the model architecture, implementation details, and the evaluation metrics that underscore the model's superior performance. Our results demonstrate the potential of DEPM to significantly enhance the reliability of electricity consumption forecasts, providing a robust tool for energy management and planning.

Weather-aware energy management for unmannedaerial vehicles: a machine learning application with global data integration

This study introduces a machine learning (ML) framework to predict unmanned aerial vehicle (UAV) energy requirements under diverse environmental conditions. The framework correlates UAV flight patterns with publicly accessible weather data, to yield an energy management tool applicable to a wide range of UAV configurations. The model employs the Cross-industry standard process for data mining and advanced feature engineering, offering an in-depth analysis of meteorological factors and UAV energy demands. The study assesses several multi-regression linear and ML models, whereby ensemble models gradient boosting (GB) and eXtreme gradient boosting demonstrate superior performance and accuracy. Specifically, the GB model achieved a test mean absolute error (MAE) of 0.0395 V (V) for voltage, 0.808 A (A) for current, and 9.758 mA-hours (mAh) for discharge, with prediction accuracy of over 99.9% for voltage and discharge, and 97% for current, derived from the coefficient of determination (R2). A novel integration of real-world UAV logs and weather data underpins the development of a weather-aware ML prediction model for UAV energy consumption. Our framework is capable of concurrently predicting three components of energy and power with almost uniform accuracy, a feature not found in contemporary models. Empirical test flights show a discrepancy of only 0.005 W-hour (Wh) between total predicted and actual energy consumption. This work enhances both efficiency and safety in UAV operations. The resulting energy-predictive flight planning tool sets a new benchmark for artificial intelligence (AI) applications in intelligent automation for UAVs.

Energy management based on coalitionnal game subdivision applied to energy communities

The energy transition requires rethinking how we produce and consume energy. Energy communities (EC) provide a recent legal framework for sharing energy, aiming to reduce energy bills and the environmental footprint of their participants. One of the challenges is adapting economic models to this technological upheaval. In this context, cooperative games, based on game theory, are valuable tools for modeling energy management through cooperation. However, despite their promising characteristics, cooperative games are limited by their computational complexity. The required computation time to solve cooperative games increases exponentially with the number of participants, restricting their application in energy management. This paper aims to propose a solution to apply cooperative game theory tools to larger communities using a multidisciplinary approach. For this purpose, a game subdivision approach based on the specific properties of energy communities is proposed. This methodology will be shown to be efficient in terms of computation time. While the game theory concepts are depreciated by limiting computing time, the sub-games method can become an interesting tool in energy management. Advantages and drawbacks in terms of energy management and game theory are discussed in this paper.

Optimal energy management and scheduling of a microgrid considering hydrogen storage and PEMFC with uncertainties

Renewable energy sources have been widely installed and operated in power systems, particularly in microgrids in the form of distributed generation units. This issue requires efficient energy management tools which take into account the inherent uncertainties of such energy resources. Thus, this paper presents a stochastic framework aimed at scheduling the renewable energy-based and thermal units in a coordinated way. The generation units comprise fuel cell units with proton exchange membrane known as PEMFC-CHP producing heat and power, concurrently. Moreover, the uncertainties arising from wind and solar power as well as market prices are characterized by deploying scenario-based optimization. The mentioned framework considers storing hydrogen and the model is presented within a stochastic mixed-integer nonlinear programming (MINLP) framework. The resulting problem is simulated on a modified 33-bus distribution network and tackled using the modified marine predators algorithm (MMPA)algorithm. The obtained results indicate that the revenue increases by more than 5% compared to other optimization algorithms. Furthermore, taking into account CHP will increase the total profit of the system by more than 15%.

Revolutionizing power grid loss prediction with advanced hybrid time series deep learning model

Accurate prediction of grid loss in power distribution networks is pivotal for efficient energy management and pricing strategies. Traditional forecasting approaches often struggle to capture the complex temporal dynamics and external influences inherent in grid loss data. In response, this research presents a novel hybrid time-series deep learning model: Gated Recurrent Units with Temporal Convolutional Networks (GRU-TCN), designed to enhance grid loss prediction accuracy. The proposed model integrates the temporal sensitivity of GRU with the local context awareness of TCN, exploiting their complementary strengths. A learnable attention mechanism fuses the outputs of both architectures, enabling the model to discern significant features for accurate prediction. The model is evaluated using well-established metrics across distinct temporal phases: training, testing, and future projection. Results showcase Resulting in encouraging Figures for mean absolute error, root mean squared error, and mean absolute percentage error, the model’s capacity to capture both long-term trends and transitory patterns. The GRU-TCN hybrid model represents a pioneering approach to power grid loss prediction, offering a flexible and precise tool for energy management. This research not only advances predictive accuracy but also lays the foundation for a smarter and more sustainable energy ecosystem, poised to transform the landscape of energy forecasting.

Advanced proactive anomaly detection in multi-pattern home appliances for energy optimization

As far as home appliances are concerned, a major issue arises as malfunctions frequently go unnoticed by homeowners, consuming a significant amount of energy. Identifying this problem highlights the need for an error detection technique, which is vital to minimize energy waste while enhancing household appliance efficiency. This paper introduces an innovative method for advanced proactive anomaly detection in multi-pattern home appliances using variational autoencoders (VAE). Specifically, a CNN-LSTM based VAE integrating a novel dynamic threshold method is exploited for identifying abnormal consumption usage. The overall suggested framework is a complete integrated approach comprising a training and testing phase. Both phases are initiated from robust feature engineering to enhance the multipattern operation program classification. Additionally, a smoothing and decomposition process is applied to optimize the performance of the CNN-LSTM VAE. The efficacy of the method is evaluated on data from 26 different front-load washing machines, demonstrating its effectiveness in identifying anomaly points across various washing programs. Furthermore, compared to the same approach with a static threshold, the proposed method showed improvements of up to 11. 4% on the F1 score. Finally, simulated use case scenarios indicate a reduction of nearly 30% of energy consumption, due to error prevention. As a result, the suggested approach is a robust and applicable tool for energy and demand side management.

Infiltration Models in EnergyPlus: Empirical Assessment for a Case Study in a Seven-Story Building

The current decarbonization transition to be achieved by 2050 according to the European Council has given great prominence to the use of Digital Twins as tools for energy management. For their correct operation, it is essential to control the uncertainties of the energy models, which lead to differences between the measured and predicted data. One of the key parameters that is most difficult to assess numerically is air leakage. The existent infiltration models available in EnergyPlus were developed to be applied in low-rise residential buildings with fewer than three stories. Therefore, it is common to rely on air leakage equations employing predefined coefficients. This research presents an empirical assessment of the performance of two EnergyPlus air leakage models, the “Effective Leakage Area” and the “Flow Coefficient”, in predicting dynamic infiltration within the attic of a seven-story building. Blower door tests, along with the application of CO2 tracer gas, were conducted to establish coefficients for the models. Then, they were evaluated in three independent periods according to the criteria established in the American Society for Testing Material D5157 Standard. Those models that only used in situ coefficients consistently met the standard across all three periods, demonstrating for both equations their accurate performance and reliability. For the best model derived from tracer gas data, the R2 and NMSE values are 0.94 and 0.019, respectively. In contrast, the model developed using blower door test data and EnergyPlus default values presented a 64% reduction in accuracy compared to the best one. This discrepancy could potentially lead to misleading energy estimates. Although other software options exist for estimating infiltration, this study specifically targets EnergyPlus users. Therefore, these findings offer valuable insights to make more informed decisions when implementing the infiltration models into energy simulations for high-rise buildings using EnergyPlus.

Comparative study of univariate and multivariate strategy for short-term forecasting of heat demand density: Exploring single and hybrid deep learning models

Accurate short-term forecasting of heating energy demand is needed for achieving optimal building energy management, cost savings, environmental sustainability, and responsible energy consumption. Furthermore, short-term heating energy prediction contributes to zero-energy building performance in cold climates. Given the critical importance of short-term forecasting in heating energy management, this study evaluated six prevalent deep-learning algorithms to predict energy load, including single and hybrid models. The overall best-performing predictors were hybrid models using Convolutional Neural Networks, regardless of whether they were multivariate or univariate. Nevertheless, while the multivariate models performed better in the first hour, the univariate models often were more accurate in the final 24 h. Thus, the best-performing predictor of the first timestep was a multivariate hybrid Convolutional Neural Network–Recurrent Neural Network model with a coefficient of determination (R²) of 0.98 and the lowest mean absolute error. Yet, the best-performing predictor of the final timestep was the univariate hybrid model Convolutional Neural Network–Long Short-Term Memory with an R² of 0.80. Also, the prediction accuracy of the best-performing multivariate hybrid models reduced faster per hour compared to the univariate models. These findings suggest that multivariate models may be better suited for early timestep predictions, while univariate models may be better suited for later time steps. Hence, combining the models can enhance accuracy at various timesteps for achieving high fidelity in forecasting and offering a comprehensive tool for energy management.

Nonintrusive Load Monitoring (NILM) Using a Deep Learning Model with a Transformer-Based Attention Mechanism and Temporal Pooling

Nonintrusive load monitoring (NILM) is an important technique for energy management and conservation. In this paper, a deep learning model based on an attention mechanism, temporal pooling, residual connections, and transformers is proposed. This article presents a novel approach for NILM to accurately discern energy consumption patterns of individual household appliances. The proposed method entails a sequence of layers, including encoders, transformers, attention, temporal pooling, and residual connections, offering a comprehensive solution for NILM while effectively capturing appliance-specific energy usage in a household. The proposed model was evaluated using UK-DALE, REDD, and REFIT datasets in both seen and unseen cases. It shows that the proposed model in this paper performs better than other methods stated in other papers in terms of F1-score and total error of the results (in terms of SAE). This model achieved an F1-score equal to 92.96 as well as a total SAE equal to −0.036, which shows its effectiveness in accurately diagnosing and estimating the energy consumption of individual home appliances. The findings of this research show that the proposed model can be a tool for energy management in residential and commercial buildings.

From White to Black-Box Models: A Review of Simulation Tools for Building Energy Management and Their Application in Consulting Practices

Buildings consume significant energy worldwide and account for a substantial proportion of greenhouse gas emissions. Therefore, building energy management has become critical with the increasing demand for sustainable buildings and energy-efficient systems. Simulation tools have become crucial in assessing the effectiveness of buildings and their energy systems, and they are widely used in building energy management. These simulation tools can be categorized into white-box and black-box models based on the level of detail and transparency of the model’s inputs and outputs. This review publication comprehensively analyzes the white-box, black-box, and web tool models for building energy simulation tools. We also examine the different simulation scales, ranging from single-family homes to districts and cities, and the various modelling approaches, such as steady-state, quasi-steady-state, and dynamic. This review aims to pinpoint the advantages and drawbacks of various simulation tools, offering guidance for upcoming research in the field of building energy management. We aim to help researchers, building designers, and engineers better understand the available simulation tools and make informed decisions when selecting and using them.

Predictive Modeling of Energy Consumption in the Steel Industry Using CatBoost Regression: A Data-Driven Approach for Sustainable Energy Management

This article presents a machine learning model for predicting energy consumption in the steel industry, which aids in energy management, cost reduction, environmental regulation compliance, informed decision-making for future energy investments, and contributes to sustainability. The dataset used for the prediction model comprises 11 attributes and 35,040 instances. The CatBoost prediction algorithm was employed for energy consumption prediction, and hyperparameter optimization was performed using GridSearchCV with 5-fold cross-validation. The developed model has undergone a comparative analysis based on both Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE) metrics, demonstrating its promise for accurate energy consumption prediction on both the training and test sets. The proposed model accurately predicts energy consumption for different load types, achieving impressive results on both the training set (RMSE=0.382, R2=0.999, MAPE=1.139) and the test set (RMSE=1.073, R2=0.998, MAPE=1.142). These findings highlight the potential of CatBoost as a valuable tool for energy management and conservation, enabling organizations to make informed decisions, optimize resource allocation, and promote sustainability.

Advancements in Home Energy Management Systems: A Review of Energy Optimization Strategies and Algorithmic Approaches

Most of the energy generated globally is used by residential and commercial structures. Due to the expansion of the world economy and population, there is a growing need for energy. Due to in Effective energy management, this has resulted in increased unit electricity prices, frequent strain on the primary electrical system, and carbon emissions. The Home Energy Management System optimizes electricity usage, reduces costs, and limits CO2 emissions by using advanced algorithms to improve the peak-to-average ratio (PAR) and dynamically adjust energy consumption. By using the Services Category (Load Shifting, Valley Filling..., households significantly improve their energy efficiency, resulting in substantial savings. Integrating smoothly with smart grids, HEMS enables real-time monitoring, empowering owners to do informed energy consumption choices. The incorporation of renewable energy sources further enhances their impact, reducing dependence on traditional grids and contributing to a collective reduction in CO2 emissions. HEMS are crucial tools for sustainable and efficient residential energy management, especially in a future of smart grids and environmentally friendly lifestyles. In this article, we are carried out a comparative study of the different algorithms used in the field of Home Energy Management (HEMS) to promote sustainable residential energy consumption. We highlight the goals, results, and limitations of each algorithm.

Short-term multi-step forecasting of rooftop solar power generation using a combined data decomposition and deep learning model of EEMD-GRU

In this study, an integrated forecasting model was developed by combining the ensemble empirical mode decomposition (EEMD) model and gated recurrent unit (GRU) neural network to accurately predict the rooftop solar power output at a specific power unit located in Tay Ninh province, Vietnam. The EEMD method was employed to decompose the solar power signals into multiple frequencies, allowing for a more comprehensive analysis. Subsequently, the GRU network, known for its ability to capture long-term dependencies, was utilized to forecast future values for each decomposition series. By merging the forecasted values obtained from the decomposition series, the final prediction for the solar power output was generated. To evaluate the efficacy of our proposed approach, a comparative analysis was undertaken against other forecasting models, including a single artificial neural network, long short-term memory network, and GRU, all of which solely considered the solar power series as input features. The experimental results provided compelling evidence of the superior performance of the EEMD-GRU model, especially when incorporating weather variables into the forecasting process, achieving the best results in all three forecasting scenarios (1-step, 2-step, and 3-step). For both forecasting targets, Inverter 155 and 156, the n-RMSE indices were 1.35%, 3.5%, and 4.8%, respectively, significantly lower than the compared single models. This integration of weather variables enhances the model's accuracy and reliability in predicting rooftop solar power output, establishing it as a valuable tool for efficient energy management in the region.

Using machine learning for the optimisation of operations and management in electric systems and networks

This research employs the Random Forest Machine Learning model to predict electricity consumption and detect anomalies in electrical networks. Addressing the energy sector’s challenges, such as supply reliability and renewable energy integration, this model processes historical electricity consumption data, weather conditions, and network events to efficiently forecast demand and identify anomalies. Data cleansing and normalisation preceded the training phase, where the model was fine-tuned using historical data to balance forecast accuracy and overfitting avoidance. The dataset was divided into training (80%) and testing (20%) sets for performance evaluation. Through cross-validation, optimal model hyperparameters were determined. The findings highlight the model’s efficacy in accurately predicting daily electricity consumption in a small, homogenous town. The model achieved a Mean Absolute Error (MAE) of 198.73 MWh and a coefficient of determination (R²) of 0.9387. Temperature, humidity, and wind speed were identified as key influencing factors on consumption levels. Conclusively, the Random Forest model presents a valuable tool for energy management, offering precise consumption forecasting and anomaly detection capabilities. Future work will address computational demands and enhance model integration with other Machine Learning methods for improved performance. This contribution is significant for efficient energy system planning and operation.

Research on the Analysis of Influencing Factors of Energy Consumption and Energy Saving Countermeasures of Green Buildings Based on DEMATEL-ISM

Abstract In the quest for sustainability, managing and optimizing energy consumption within green buildings has emerged as a critical area of focus. The complexity of factors affecting energy use demands more sophisticated analysis than what traditional methods offer. This study leverages the DEMATEL method to isolate crucial influencing factors. It employs the ISM method to map their logical interconnections, forming a comprehensive model of energy consumption determinants in green buildings. We identify building design, energy efficiency, and green technology as critical elements, with building design exerting the most influence (weight: 0.35). Through an in-depth fuzzy DEMATEL-ISM analysis, we develop specific energy-saving measures—such as design improvement and technology enhancement—that promise up to a 20% reduction in energy consumption. Our research introduces a groundbreaking green building energy management tool and provides a solid foundation for policy development and practical applications.

Industrial cluster energy systems integration and management tool

Industrial cluster energy systems integration and management tool

Coordinated pricing mechanism for parking clusters considering interval-guided uncertainty-aware strategies

With the increasing penetration of electric vehicles into the existing power networks, the development of proper energy management tools become essential to mitigate the damaging effects of emerging high-power charging modes. To facilitate the management of vehicles, they can be gathered into clusters in order to aggregate their charging profiles. This paradigm is called cluster of vehicles and can be forward developed by gathering a group of parking, thus forming a parking cluster where the characteristics of different parking lots are aggregated. In this paper, a coordinated charging price mechanism for clusters of parking lots is proposed. The new proposal is casted as a game-based framework which seeks for the equilibrium between the cluster coordinator and parking lots, instead of conventional tools where the pricing strategy is decided by a central entity disregarding the interests of users. The developed model is established as a bi-level optimization problem, which is further linearized to be tractable and solvable by off-the-shelf solvers. The implications of vehicle-to-grid are also studied by fully modelling bi-directional power flows in chargers. In addition, two uncertainty-aware strategies are proposed to cope with uncertain energy margins in parking lots. A case study is conducted and various results are analysed and discussed. It is shown that enabling vehicle-to-grid characteristics discloses significant economic benefits for users and the cluster coordinator. In addition, vehicle-to-grid impacts notably on the risk-averse character of the uncertainty-aware strategies proposed. The developed pricing mechanism is also compared with the conventional pricing policies based on key of repartitions, showing that the new proposal is able to reduce the cost for users, avoiding to directly translate the energy cost to charging points.

Are the next-generation households ready for the energy transition? A survey on their positioning and practice with energy management tools

In the last decades, significant effort has been put towards technological advancement in housing for energy transition. Massive retrofitting actions have been called for, and innovative technologies for smart energy management at home have been deployed. However, undesired energy trends in housing suggest that relevant factors have been neglected. Among these, increasing importance is now given to occupants’ behaviour, and their capacity to interact with energy management devices available in dwellings. This study investigates what is the position of next-generation users on energy transition at home. Two years ago, the authors launched a survey to explore people’s awareness of energy use practices, interaction with metering devices, and user motivation to change when informed. As a pilot survey, over 300 people from the academy were involved to see what was the position of a sample which was supposed to be informed more than the average, in Italy. The test yielded early outcomes on how people become more interested to change as they gain knowledge and are offered suggestions. Despite the expectations, the sample’s level of awareness was low. This suggested that a more user-centred approach is needed for wide-scale progress. Especially results from the youngest were below prospects. The questionnaire was relaunched to examine if the pandemic, energy crisis and latest news on climate change have affected positions of the youngsters. A testing session involving university students was performed, and results have been compared with the previous. As a result, reflections on the energy use patterns of the next-generation households are provided.

Life cycle energy consumption prediction based on an extended system boundary with the Bi-LSTM model: An empirical study of China

Global warming and natural disasters have wreaked havoc on Earth these days. Energy conservation has been a priority under social development pressure, especially in the building sector. Energy consumption prediction serves as a vital tool for effective building energy management, guiding energy policy making and services distribution. However, despite the application of advanced technologies, achieving accurate energy consumption prediction still faces challenges due to potential influencing factors. This study aims to explore previously unidentified factors to enhance building energy consumption prediction accuracy. An extended life cycle energy boundary of buildings including embodied, operational, and mobile energies is proposed. To optimize the prediction processes, a hybrid Life Cycle Assessment (LCA) approach and a non-linear Bidirectional Long Short-Term Memory (Bi-LSTM) model are combined. An empirical study was conducted to predict the life cycle energy consumption of urban residential buildings in China by 2035. Results show: (1) life cycle energy consumption increased dramatically from 195.837 (±11.77) Mtce in 2000 to 1151.69 (±80.38) Mtce in 2022, but then achieved a slow decline to 796.998 ((±51.46)) Mtce by 2035; (2) there were four significant development stages in the residential field: rapid growth (2000–2010), sharp growth (2010–2018), slow growth (2018–2022), and stable fall (2022–2035); (3) mobile energy related to building location accounted for up to 28% of total energy consumption, and daily commuting is its largest emitter. However, calculation errors caused by diverse data sources and the fundamental model itself need to be addressed later. This study provides an original pathway of combining LCA with a non-linear model to improve energy consumption prediction and the completed primary historical data to help local governments’ energy decision-making.

A fully robust home energy management model considering real time price and on-board vehicle batteries

On-board batteries from electric vehicles can be used in domestic installations as an auxiliary source of energy through their vehicle-to-home features. To properly manage this kind of assets, Home Energy Management systems are expected to be essential tools in future smart grids, enabling a more efficient use of energy in residential applications. However, these tools normally perform under severe uncertain environments caused by intermittent renewable generation, uncertain prices, uncertain demand as well as uncertain behaviour of electric vehicles. In this regard, further research should be conducted in order to develop advanced robust energy management tools that account for the effect of uncertainties. So far, most of the existing literature focused on simple stochastic approaches, that may pose various issues, while other more sophisticated uncertainty models have not been sufficiently studied. In this regard, this paper proposes a fully robust Home Energy Management model, which accounts for all the inherent uncertainties that may arise in domestic installations. In contrast to previous researches, robust optimization is applied to all the uncertainties, resulting in a tractable, but robust yet paradigm model which allows to adapt the degree of robustness. A case study is conducted on a benchmark prosumer paradigm to validate the new proposal as well as highlight the benefits of flexible appliances. Results reveal the usefulness of the new tool, which allows to handle very unfavourable uncertain profiles, especially in case of the electric vehicles, for which a very low initial state-of-charge and early departure time can be assumed, leading to more conservative results that suppose incrementing the electricity bill by 38 %.