Smart Building Energy Management Research Articles

A Multi-Agent Deep Constrained Q-Learning Method for Smart Building Energy Management Under Uncertainties

A Multi-Agent Deep Constrained Q-Learning Method for Smart Building Energy Management Under Uncertainties

AI-powered deep learning for sustainable industry 4.0 and internet of things: Enhancing energy management in smart buildings

With the increasing demand for energy in urban areas, there is a pressing need to manage energy consumption more effectively. Buildings, especially commercial and industrial ones, are major consumers of energy. Implementing AI-powered solutions can help monitor, predict, and reduce energy usage, leading to substantial cost savings and more efficient energy use. Integrating Industry 4.0 and the Internet of Things (IoT) into smart buildings presents a significant opportunity for enhancing energy management through advanced technologies. The increasing demand for energy in urban areas necessitates the development of more efficient energy management strategies, particularly within smart buildings, which are significant energy consumers. This study proposes an AI-powered deep learning framework utilizing Convolutional Neural Networks (CNNs) to enhance energy management in smart buildings, leveraging the ASHRAE - Great Energy Predictor III dataset. The proposed framework leverages deep learning techniques to analyze historical energy data, weather conditions, and building characteristics to forecast future energy usage accurately. Additionally, the framework includes anomaly detection mechanisms to identify inefficiencies and faults in the energy management system. By optimizing energy consumption in real-time and implementing demand response strategies, the framework aims to reduce energy costs and enhance the overall efficiency of smart buildings. The proposed CNN-IoT-based energy management framework in smart buildings demonstrates significant advancements over existing methods, achieving an accuracy of 88 %. These performance metrics indicate a substantial improvement in prediction accuracy and efficiency compared to existing approaches such as SVM, ELM, and LSTM.

A smart building energy management incorporating clustering-based tariffs in the presence of domestic solar energy, battery, and electric vehicle

Smart buildings play a crucial role in optimizing energy management within the power network. As end-users of the power network, they have the ability to not only reduce economic costs for householders but also modify the technical indices of the power network. To promote efficient device management in smart homes (SH), demand response programs are recommended for consumers. This research investigates the application of clustering-based electricity pricing strategy aimed at effectively managing the energy devices of a residential smart home. The utilized method categorizes the electricity tariff into five rates according to the clustering of the real-time pricing program. Ward’s clustering method is utilized to cluster and determine new electricity tariffs. The primary goal of the energy management program is to minimize the building’s energy cost, which is accomplished through the utilization of the multi-verse optimizer. The smart home consists of essential and manageable appliances, a photovoltaic panel (PV), a sodium-sulfur (NaS) battery, and an electric vehicle (EV). The initial parameters of the PV and EV are modeled stochastically by their probability distribution functions and calculated using the Latin hypercube sampling algorithm. The smart building’s performance is assessed by taking into account various demand response programs. The numerical results present that the application of the clustering-based management method has resulted in a significant reduction of 23–43 % in the electricity cost of smart homes. Additionally, the smart home exhibits a more linear consumption pattern when considering the electricity tariffs based on the clustering approach.

Internet of things and deep learning-enhanced monitoring for energy efficiency in older buildings

Retrofitting older buildings for energy efficiency is paramount in today's sustainability and environmental awareness era. Older buildings contribute greatly to energy waste since they typically lack new energy-efficient technology. Reducing carbon emissions, lowering energy bills, and extending the life of these historic landmarks all depend on fixing the energy inefficiency that plagues these buildings. Despite advanced technologies' remarkable progress, the potential of the Internet of Things and deep learning in older buildings has not been unexplored. Major obstacles include expensive and out-of-date infrastructure and the difficulty of incorporating new technology into historically significant structures. Existing research has mostly ignored older infrastructures' unique requirements and limitations in favour of current or newly built services. In addition, comprehensive research on integrating deep learning with the Internet of Things in this specific environment has been lacking. Smart building energy management is made possible by the Internet of Things (IoT) and deep learning. Architectural limitations, outmoded infrastructure, and the necessity for non-invasive retrofitting solutions all contribute to the difficulty of energy monitoring and improvement in older buildings. This research proposes combining the IoT with Deep Learning-enhanced Predictive Energy Modeling (DL-PEM) to make an energy management system that can change and adapt to the needs of older buildings. Data from IoT sensors is collected on occupancy, temperature, lighting, and equipment usage and then analyzed using Deep Learning models to determine the most efficient energy consumption patterns. Beyond its energy-saving potential, this method has many potential uses. Spotting and fixing structural problems before they become major can improve occupant comfort, reduce maintenance costs, and pave the way for predictive maintenance. Integration with the Smart grid and demand response programs can be facilitated, too, improving the reliability of the power system as a whole. Our IoT and Deep Learning-based energy management solution optimizes energy usage, reduces expenses, and mitigates environmental impact in older buildings, as shown by extensive simulation studies. The system's performance is compared to more conventional methods, and its flexibility is evaluated in various building and usage contexts. The experimental outcomes show the suggested DL-PEM model increases the energy demand forecasting analysis, energy consumption analysis, thermal comfort optimization analysis, seasonal variation analysis, and occupancy data analysis.

Enhancing Energy Efficiency With Smart Building Energy Management System Using Machine Learning and IOT

Enhancing Energy Efficiency With Smart Building Energy Management System Using Machine Learning and IOT

Demand flexibility and cost-saving potentials via smart building energy management: Opportunities in residential space heating across the US

Leveraging demand-side flexibility resources (e.g., buildings) is a crucial and cost-effective strategy for addressing the operational and infrastructure-related challenges in power grids pursuing deep decarbonization with high renewable energy penetration. However, the demand flexibility opportunities and financial benefits for residential space heating, which are sizeable demand-side flexibility resources, through emerging building energy management solutions (i.e., smart control and phased change material (PCM) thermal storage) across the US are not fully understood. In this paper, we systematically assess the demand flexibility and cost-saving/revenue potentials in residential space heating through detailed building-level simulations for five consecutive years at a 5-min temporal resolution in 20 metro areas across the high-heating-demand regions of the US. The results show a high degree of synergy between PCM thermal storage and smart control, which enables substantial demand flexibility potential, reaching 98.5% of peak load shifting, and electricity cost-saving/revenue potential, reaching 338.3% of electricity cost reductions, for residential space heating in the US. By achieving such performance, adopting smart control and PCM thermal storage is financially viable in 50% of the tested metro areas. The results reveal that the demand flexibility and cost-saving/revenue potentials of residential space heating in the US are further enhanced by higher volatilities in electricity prices. Active PCM thermal storage has lower energy efficiency but much higher energy flexibility than passive PCM thermal storage. Based on the findings, recommendations for integrating PCM thermal storage and smart control systems within residential space heating are provided.

Optimizing smart building energy management systems through industry 4.0: A response surface methodology approach

Effective control of inner climatic conditions in a Heating ventilation and air conditioning (HVAC) system based on efficiency of the system is imperative to ensure occupant comfort, energy efficiency, and preservation of indoor air quality. The existing problems with HVAC system includes inefficient adjustment to internal room parameters and limited optimization based on varying loads. The current study investigates the influence of Industry 4.0 by assessing internal room parameters in an indoor ventilated room conditions for system output. Smart systems like AI and IoT, integrated with sensors, dynamically adjust indoor conditions based on external climate, enhancing HVAC system efficiency. Using Artificial Neural Networks (ANN) and Response Surface Methodology (RSM), integrated with AI and IoT sensors, optimally adjust indoor conditions in correlation with external climate, maximizing HVAC system efficiency. The study utilizes input parameters including Dry Bulb Temperature (DBT), Relative Humidity (RH), and Solar Radiation across diverse days to analyze system efficiency. ANN and RSM iteratively optimize system efficiency by training on RH, DBT, and air quality data. They converge on recommending 42% RH, 28 °C DBT, and 25 PM air quality, achieving 0.851 desirability. This approach enhances heat load (75.89%) and ventilation load (69.12%) efficiency significantly. Integration of Industry 4.0 sensors in the HVAC system resulted in a 16% efficiency increase and 15% cost effectiveness in the controlled room compared to prior measurements.

Machine Learning Applications for Smart Building Energy Utilization: A Survey

The United Nations launched sustainable development goals in 2015 that include goals for sustainable energy. From global energy consumption, households consume 20–30% of energy in Europe, North America and Asia; furthermore, the overall global energy consumption has steadily increased in the recent decades. Consequently, to meet the increased energy demand and to promote efficient energy consumption, there is a persistent need to develop applications enhancing utilization of energy in buildings. However, despite the potential significance of AI in this area, few surveys have systematically categorized these applications. Therefore, this paper presents a systematic review of the literature, and then creates a novel taxonomy for applications of smart building energy utilization. The contributions of this paper are (a) a systematic review of applications and machine learning methods for smart building energy utilization, (b) a novel taxonomy for the applications, (c) detailed analysis of these solutions and techniques used for the applications (electric grid, smart building energy management and control, maintenance and security, and personalization), and, finally, (d) a discussion on open issues and developments in the field.

Machine learning-based energy use prediction for the smart building energy management system

Smart building is a building development approach utilizing digital and communication technology to improve occupants' comfort inside the building and help increase energy usage efficiency in building operations. Despite its benefits, the smart building concept is still slowly adopted, particularly in developing countries. The advancement of computational techniques such as machine learning (ML) has helped building owners simulate and optimize various building performances in the building design process more accurately. Therefore, this study aims to assist energy efficiency design strategies in a building by identifying the features of the smart building characteristics that can potentially foster building energy efficiency. Furthermore, an ML model based on the features identified is then developed to predict the level of energy use. K-Nearest Neighbor (k-NN) algorithm is employed to develop the model with the openly accessible smart building energy usage datasets from Chulalongkorn University Building Energy Management System (CU-BEMS) as the training and testing datasets. The validation result shows that the predictive model has an average relative error value of 17.76%. The energy efficiency levels obtained from applying identified features range from 34.5% to 45.3%, depending on the reviewed floor. This paper also proposed the dashboard interface design for ML-based smart building energy management.

Deep Learning and Artificial Intelligence in Sustainability: A Review of SDGs, Renewable Energy, and Environmental Health

Artificial intelligence (AI) and deep learning (DL) have shown tremendous potential in driving sustainability across various sectors. This paper reviews recent advancements in AI and DL and explores their applications in achieving sustainable development goals (SDGs), renewable energy, environmental health, and smart building energy management. AI has the potential to contribute to 134 of the 169 targets across all SDGs, but the rapid development of these technologies necessitates comprehensive regulatory oversight to ensure transparency, safety, and ethical standards. In the renewable energy sector, AI and DL have been effectively utilized in optimizing energy management, fault detection, and power grid stability. They have also demonstrated promise in enhancing waste management and predictive analysis in photovoltaic power plants. In the field of environmental health, the integration of AI and DL has facilitated the analysis of complex spatial data, improving exposure modeling and disease prediction. However, challenges such as the explainability and transparency of AI and DL models, the scalability and high dimensionality of data, the integration with next-generation wireless networks, and ethics and privacy concerns need to be addressed. Future research should focus on enhancing the explainability and transparency of AI and DL models, developing scalable algorithms for processing large datasets, exploring the integration of AI with next-generation wireless networks, and addressing ethical and privacy considerations. Additionally, improving the energy efficiency of AI and DL models is crucial to ensure the sustainable use of these technologies. By addressing these challenges and fostering responsible and innovative use, AI and DL can significantly contribute to a more sustainable future.

A novel economic dispatch in the stand-alone system using improved butterfly optimization algorithm

Distributed renewable energy systems are now widely installed in many buildings, transforming the buildings into ‘electricity prosumers'. Additionally, managing shared energy usage and trade in smart buildings continues to be a significant difficulty. The main goal of solving such problems is to flatten the aggregate power consumption-generation curve and increase the local direct power trading among the participants as much as possible. This study provides a coordinated smart building energy-sharing concept for smart neighborhood buildings integrated with renewable energy sources and energy storage devices within the building itself. This neighborhood energy management model's primary objective is to reduce the total power cost of all customers of smart buildings in the neighborhood by increasing the use of locally produced renewable energy. In the first stage, a group of optimum consumption schedules for each HEMS is calculated by an Improved Butterfly Optimization Algorithm (IBOA). A neighborhood energy management system (NEMS) is established in the second stage based on a consensus algorithm. A group of four smart buildings is used as a test system to evaluate the effectiveness of the suggested neighborhood smart building energy management model. These buildings have varying load profiles and levels of integration of renewable energy. In this paper, the proposed framework is evaluated by comparing it with the Grey Wolf optimization (GWO) algorithm and W/O scheduling cases. With applying GWO, the total electricity cost, peak load, PAR, and waiting time are improved with 3873.723 cents, 21.6005 (kW), 7.162225 (kW), and 87 s respectively for ToU pricing and 11217.57 (cents), 18.0425(kW), 5.984825 (kW), and 98 s respectively for CPP tariff. However, using the IBOA Improves the total electricity cost, peak load, PAR, and waiting time by 3850.61 (cents), 20.1245 (kW), 6.7922 (kW), and 53 s respectively, for ToU and 10595.8 (cents), 17.6765(kW), 5.83255(kW), and 74 s for CPP tariff. Also, it is noted that the run time is improved using GWO and IBOA by 13% and 47%, respectively, for ToU and 2% and 26% for CPP. However, the number of iterations required to obtain the optimal solution is reduced using the GWO and IBOA by 60% and 81% for ToU and 55% and 80% for CPP tariffs. The results show significant improvements obtained by applying just intelligent programming and management.

Day-ahead forecasting of residential electric power consumption for energy management using Long Short-Term Memory encoder–decoder model

Energy management in smart buildings and energy communities needs short-term load demand forecasting for optimization-based scheduling, dispatch, and real-time operation. However, producing accurate forecasting for individual residential households is more challenging compared to the forecasting of load demand at the distribution level, which is smoother and benefits from statistical compensation of errors. This paper presents a day-ahead forecasting technique for individual residential load demand that is based on the Long Short-Term Memory encoder–decoder architecture, which is extended to consider possibly differing sets of past and future exogenous variables. A novel focus is posed on the validation of the proposed approach considering that it is tailored for use by energy management systems. A publicly available dataset was used for validation, and the approach was compared with three other methods, resulting in a reduction of the Mean Absolute Scaled Error by up to 8%.

Artificial Intelligence Assisted Energy Management in Smart Buildings: A Review

Günümüzde küresel iklim değişikliği sebebiyle enerji yönetim sistemlerine olan ihtiyaç giderek artmaktadır. Bu çalışmada akıllı binalarda yapay zekâ destekli enerji yönetimi ile ilgili literatürdeki çalışmalar incelenmiştir. İncelemeye alınan çalışmaların araştırılmasında YÖKTEZ, IEEE, Web of Science, Science Direct ve ProQuest veri tabanlarından yararlanılmıştır. Bu veri tabanlarında önceden belirlenen anahtar kelimeler ile arama yapılıp, konu ile yakın içerikteki çalışmaların incelenmiştir. İncelemede çalışmaların amaçları ve içerikleri ile ilgili bilgi verilmiştir. Ayrıca ilgili çalışmaların daha rahat analiz edilmesi için çalışmaların adını, türünü, gerçekleştirildiği ülkeyi, konusunu, veri toplama aracını, örneklem büyüklüğünü ve sınırlılıklarını içerecek şekilde literatürdeki 32 çalışmanın yer aldığı bir tablo oluşturulup, inceleme için bu tablodan yararlanılmıştır. İncelenen çalışmalar içeriklerine ve sonuçlarına göre yorumlanıp, ileride yapılabilecek çalışmalar hakkında öneriler sunulmuştur. Çalışma kapsamında yapılan inceleme sonucunda, akıllı binalarda enerji sistemlerinin büyük bir çoğunluğunun sadece elektrik enerjisinin yönetimi üzerinde durduğu görülmüştür. Buna ek olarak araştırma kapsamına dâhil edilen çalışmaların büyük bir çoğunluğunun Hindistan’da gerçekleştirildiği, yapılan inceleme sonucunda görülmüştür.

Large-scale comparison and demonstration of continual learning for adaptive data-driven building energy prediction

Data-driven models have been increasingly employed in smart building energy management. To avoid performance degradation over time, data-driven models need to be continually updated to adapt to the changes in building operations. However, several critical issues in the model update process raised wide concerns, especially the concept drift and catastrophic forgetting issues. The concept drift issue happens when the statistical properties of target variable change over time in unforeseen ways. The catastrophic forgetting issue refers to the process that the previously learnt knowledge or patterns may be diluted and eventually lost in model update. Although a few model update methods were proposed, there is a lack of comprehensive comparison of the methods for adaptive data-driven building energy prediction. This paper conducted a comprehensive investigation on the performance of three conventional model update methods and five emerging continual learning methods using 2-year data of 100 buildings extracted from an open-source dataset. The results show that continual learning methods are more effective in ensuring long-term accuracy while cutting down on computation time and data storage expenses. The CV-RMSE of Elastic weight consolidation and Gradient episodic memory decreased by around 14% and 8% on average compared with static model and accumulative learning. The comparison results are valuable to the development of adaptive data-driven building energy prediction models which are more reliable over time and robust against changing operation conditions, thus more practically applicable in smart building energy management.

Optimal energy management in smart buildings with electric vehicles based on economic and risk aspects using developed whale optimization algorithm

With the development of smart electricity consumption technology, consumers can benefit from an economically optimized home energy management system, considering their well-being and utilizing electrical appliances. The multi-objective modelling of energy management in smart buildings with the presence of electric vehicles is the main topic of this paper, which also takes risk and economic considerations into account. To make the problem more realistic, various practical and nonlinear constraints are taken into account, such as production limits, flexible loads, and cost-risk limitations and indicators. Since solving this problem with classical methods is challenging and prone to suboptimal solutions, a developed whale optimization algorithm is suggested. In the proposed developed approach, efforts are made to enhance its capabilities and prevent it from converging to local points using local search. Finally, the proposed method and model are discussed and examined under different scenarios. The results demonstrate that when consumers have a positive attitude towards energy consumption, the net cost significantly decreases compared to a conservative attitude. The battery is used during times of favourable prices when users perceive more solar energy availability. Moreover, when users adopt different price risk indicators, these indicators have varying effects on the electricity purchasing strategy of the user.

Smart building energy management and monitoring system based on artificial intelligence in smart city

In the present scenario, the fastest-growing environmental concerns are energy management and monitoring. In-efficient energy recycling, energy consumption, energy utilization, and drain characteristic are smart building energy management challenges. Hence to examine the connection between smart city management policies and energy management, this research proposed an Artificial Intelligence Technique for Monitoring Systems in Smart Buildings (AIMS-SB) to manage energy consumption and produce and recycle energy required for a smart building. AIMS-SB helps to predict energy analysis, renewable energy production, and recycling evaluation based on prediction model strategies. AIMS-SB developed eco-design monitoring systems for smart buildings to optimize energy consumption, utilization, and drain characteristics. These efficient implementation strategies and methods for harnessing renewable energy help to improve the safety process, recycling, and reuse of our energy resources for smart building energy management. AIMS-SB provides viable solutions to the growing number of challenges associated with smart city energy management. Therefore, the system's findings demonstrate increased accuracy and efficiency compared to conventional methods.

Predictive Energy Management in Internet of Things: Optimization of Smart Buildings for Energy Efficiency

As energy efficiency and sustainability become paramount in the face of growing urbanization and environmental concerns, predictive energy management in smart buildings has emerged as a promising avenue for mitigating energy consumption and optimizing resource utilization. In this paper, we investigate the application of advanced machine learning techniques, particularly a multi-layer Long Short-Term Memory (LSTM) model, within the framework of the Internet of Things (IoT), to predict and manage energy consumption. We rigorously evaluate our approach against a suite of machine learning baselines, including Linear Regression, Random Forest, Support Vector Machine, and Gradient Boosting, utilizing a comprehensive dataset encompassing power consumption data from smart home appliances and associated weather variables. Our experimental results demonstrate the superior predictive capabilities of the LSTM model, showcasing its ability to outperform traditional machine learning baselines across various metrics, including Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE). These findings underscore the potential of deep learning models in capturing intricate temporal dependencies within energy consumption data, contributing to improved energy efficiency, cost savings, and environmental sustainability in smart building environments. The integration of predictive energy management models into IoT-enabled smart buildings holds the promise of a more intelligent and sustainable future in urban development and resource management.

A Machine Learning Approach to Passive Human Motion Detection Using WiFi Measurements From Commodity IoT Devices

Human motion is a primary indicator of indoor occupancy and activity. Motion sensing has paramount importance in the energy management of modern smart buildings and is employed for automated controls of lighting and heating, ventilation and air conditioning (HVAC) equipments. The all-pervasive WiFi infrastructure in urban buildings offers an opportunistic method of human motion detection through passive sensing of WiFi received signal strength indicator (RSSI) and channel state information (CSI). This technique unfolds a plethora of Building IoT related services, in addition to sustainable energy utilization and reduced emission of greenhouse gases. In this paper, a device free human motion detection method through WiFi RSSI and CSI collected from commercial-off-the-shelf (COTS) IoT devices, is proposed. Utilizing a laptop, a smartphone and an ESP32 as receivers, WiFi RSSI and CSI samples were collected from two residential buildings, to constitute 6 datasets. A four dimensional feature vector that exploits the data spread in time domain, is extracted from the collected samples and utilized to train a two stage ensemble machine learning model. A comparison of various RSSI based datasets indicates a mean cross validation accuracy of up to 99.7% and 97.8% in line of sight (LoS) and through-the-wall scenarios, respectively. The detection accuracy in non LoS environments can be enhanced using CSI based features, enabling motion detection in different rooms using a single WiFi router.

GridLearn: Multiagent reinforcement learning for grid-aware building energy management

Increasing amounts of distributed generation in distribution networks can provide both challenges and opportunities for voltage regulation across the network. Intelligent control of smart inverters and other smart building energy management systems can be leveraged to alleviate these issues. GridLearn is a multiagent reinforcement learning platform that incorporates both building energy models and power flow models to achieve grid level goals, by controlling behind-the-meter resources. This study demonstrates how multi-agent reinforcement learning can preserve building owner privacy and comfort while pursuing grid-level objectives. Building upon the CityLearn framework which considers RL for building-level goals, this work expands the framework to a network setting where grid-level goals are additionally considered. As a case study, we consider voltage regulation on the IEEE-33 bus network using controllable building loads, energy storage, and smart inverters. The results show that the RL agents nominally reduce instances of undervoltages and reduce instances of overvoltages by 34%. • Feasibility study of multiagent reinforcement learning for voltage regulation. • Multiagent reinforcement learning for demand side management with power flow analysis. • Open source building energy management simulation.

A location-based fog computing optimization of energy management in smart buildings: DEVS modeling and design of connected objects

A location-based fog computing optimization of energy management in smart buildings: DEVS modeling and design of connected objects