Comprehensive review of modeling, structure, and integration techniques of smart buildings in the cyber-physical-social system

Abstract

Smart buildings have been proven to be a kind of flexible demand response resources in the power system. To maximize the utilization of the demand response resources, such as the heating, ventilating and air-conditioning (HVAC), the energy storage systems (ESSs), the plug-in electric vehicles (PEVs), and the photovoltaic systems (PVs), their controlling, operation and information communication technologies have been widely studied. Involving human behaviors and cyber space, a traditional power system evolves into a cyber-physical-social system (CPSS). Lots of new operation frameworks, controlling methods and potential resources integration techniques will be introduced. Conversely, these new techniques urge the reforming requirement of the techniques on the modeling, structure, and integration techniques of smart buildings. In this paper, a brief comprehensive survey of the modeling, controlling, and operation of smart buildings is provided. Besides, a novel CPSS-based smart building operation structure is proposed, and the integration techniques for the group of smart buildings are discussed. Moreover, available business models for aggregating the smart buildings are discussed. Furthermore, the required advanced technologies for well-developed smart buildings are outlined.

Graphical abstract

Keywords

smart buildings / cyber-physical-social-system / optimization / modeling / demand response / virtual power plant

1 Introduction

As the largest energy consumer in the world, mega-cities almost account for 75% of global carbon emissions [1]. With the enormous attractions and convenient environment, an amazing 70% of the world population will come to live and work in mega-cities by 2050 [2]. The energy consumption of commercial and residential buildings (CRBs) contributes to more than 70% of the electricity use among the grid scale [3]. Equipped with smart devices for fast and controllable adjustment in energy saving and human comfort, CRB has gradually become a fundamental component of the smart grid, named smart building.

In a typical smart building, there are several smart building equipment, such as HVAC units, ESSs, PEVs, PVs, and controlled lights. Numerous researches on modeling [4–7], control strategy [8–11], energy management [12–15], and energy transaction [16–20] of smart buildings have been conducted. In modeling, Efkarpidis etal. [4] proposed a “grey-box”-based model of heat and cooling energy in smart buildings to hourly accurately compute the thermal energy needs. Zhang etal. [5] proposed a plug-and-play learning framework based on the Internet of Things (IoT) technologies to automatically identify the thermal model of each thermal zone. The aforementioned methods require domain knowledge and lengthy building-by-building configuration. Differently, the following thermal models are established by artificial intelligence (AI) techniques. Zhang etal. [6] utilized the IoT technologies and deep neural network (DNN) to derive an accurate thermal model considering the environmental thermal comfort. Similarly, supported by the IoT-based pervasive sensing technologies, Hu etal. [7] proposed a learning-based solution for thermal comfort modeling via machine learning technologies. The IoT technologies and AI methods are playing an important role in thermal modeling of CRBs.

In the aspect of control strategy, Yu etal. [8] proposed a real-time HVAC control algorithm based on Lyapunov optimization techniques for cost-efficient commercial buildings. Similarly considering economical costs, Ostadijafari etal. [9] developed a nonlinear economic model predictive controller to minimize the cost of energy consumption of HAVC, PVs, and ESSs in smart buildings. Different from the one-zone thermal control, a HVAC control algorithm was proposed in a multi-zone commercial building considering thermal comfort, random zone occupancy and indoor air quality comfort in Ref. [10]. Sadid etal. [11] formulated a discrete-event system-based admission control of thermal appliances in smart buildings. The control strategy research in smart buildings mostly focused on the HVAC control algorithm for minimizing energy cost or energy admission.

As for energy management, a practical application of economical building management system for commercial buildings Singapore was presented in Ref. [12]. Razmara etal. [13] proposed a bilevel optimization framework for smart building-to-grid (B2G) system, which benefited both entities involved in operation of smart buildings and distribution grid. Mocanu etal. [14] and Pinzon etal. [15] developed a building energy management optimization model that could be easily incorporated into online optimization schemes.

The aim of building energy management is not only for the minimum energy consumption and costs, but also for maximum profits in electricity markets. Chouikhi etal. [16] introduced two electricity management mechanisms and proposed a game-theoretic distributed energy demand scheduling approach for smart buildings. The peer-to-peer (P2P) energy transaction frameworks for smart buildings were presented in Refs. [17,19]. Participating in demand response (DR) to improve power system reliability and efficiency by adjusting the load profiles is another common business model for smart building. Zhou etal. [18] proposed an agent-based modeling and techniques for different types of commercial buildings to simulate an electricity market. Buildings in the microgrids are also associated with the renewable energy to participate in day-ahead market [20]. The advanced control strategies make building devices more controllable from the device level, and the building energy managements and energy trading strategies make the building more economical and environmentally-friendly.

The framework of smart buildings contains the electrical devices, the information transmission, and human society. As illustrated in Fig. 1, the smart building is a small but complete cyber-physical-social-system (CPSS), an extended version of cyber-physical system (CPS). The context of CPSs is first completely presented in Ref. [21] and three layers were introduced, i.e., the perception layer, the transport layer, and the application layer [22]. Compared with CPSs, CPSSs consider human behavior and society factors as a part of the system through human-computer interactions. Kalluri etal. [23] analyzed the connection between CPSs and the design of smart buildings in urban cities. Jin etal. [24] proposed a user-oriented investment decision toolset, considering smart buildings as human-cyber-physical-systems. With the further development of CPSS, the effects of human behaviors should not be neglected.

Up to the present, there have been several breakthroughs or innovations in modeling, management, and integration of smart buildings, but few comprehensive reviews are presented to illustrate the techniques concerning the development of smart buildings. Therefore, this paper focuses on the review of the technical development of smart buildings in CPSS. Several reviews of CPSS or smart buildings have been conducted. Zhou etal. [25] made a comprehensive survey about the differences between CPSs and CPSSs. Samad etal. [26] reviewed the automated demand response for smart buildings and outlined the architectural models, technology infrastructure, and control and communication protocols. Minoli etal. [27] organized some technical opportunities and the technical challenges for smart buildings in the field of IoT. Rashidizadeh-Kermani etal. [28] proposed a risk-based scheduling optimization of VPP considering the demand response of smart buildings. Royapoor etal. [29] assessed the potential of demand response which is provided by commercial buildings integrated as a VPP. However, most of these reviews just focused on the practical applications and challenges of CPSS. Few of them considered the impact of CPSS on smart buildings. The smart building framework using the concept of CPSS and its integration techniques have not been summarized yet. This paper presents a comprehensive review on the framework and orientation of smart buildings in CPSSs concerning modeling of smart buildings, energy management/optimal operation of smart buildings, architecture and integration of smart buildings in CPSSs, and opportunities and challenges for smart buildings, as shown in Fig. 2.

The paper is contributive because various modeling techniques for smart buildings are summarized, including power device modeling, information transmission modeling, human behavior description modeling, embodied energy and carbon of smart buildings, and cost-benefit modeling under IoT technologies. In addition, published researches on management, optimization models, and control strategies of smart buildings are listed and introduced. Moreover, the framework of smart buildings under the concept of CPSS is proposed and the integration techniques for multi smart buildings are compared. Ultimately, smart building integrated as VPPs is discussed. The relevant key components and techniques are also outlined. Furthermore, available business model for smart buildings integrated as virtual power plant is expounded. Meanwhile, the research frontiers and challenges of smart buildings integrated as virtual power plant are outlined.

2 Modeling of smart building

Many kinds of electrical devices are deployed for environmental comfort and protection, such as HVACs for suitable living temperature, PVs for low-carbon environment, ESSs for emergency reserve, and PEVs for convenient transportation. The operation states of these devices are perceived and delivered by the sensors network. After receiving the information from the physical world, human in social layer can get a good command of these devices through the actuation network. Accordingly, smart building modeling consists of electrical device modeling, information transmission modeling, and human behavior modeling. Related references are included in Table 1.

2.1 Electrical device modeling

2.1.1 Heating, ventilating, and air-conditioning (HVAC)

A typical HVAC system consists of an air handling unit for the whole building and a set of variable air volume boxes for n zones in the building [10,30]. A physical-based modeling technique for heat and cooling demand is a major research direction, which is also called “white-box” techniques; another major research direction is based on data-driven techniques, also known as “black-box” techniques [4]. In the literature related to white-box techniques, numerical techniques are mainly utilized for the modeling of thermal heating and cooling demand [31–33]. A finite-difference-method based thermal conduction transfer function through walls was developed by implementing it in TRNSYS [31]. In Ref. [32], the authors proposed a heating and cooling demand estimation model using the finite element method in a residential building. To reduce the computational complexity of simulation tools based on finite elements or finite difference methods, a multiparameter model order reduction was proposed for thermal modeling [33]. Although numerical techniques can explicitly present the physical principles, expensive calculation, high computation time requirement, and high amount of building information requirement for parameter identification limit the scope of numerical techniques. Data-driven techniques perform well in the weak areas of numerical techniques, such as support vector regression (SVR) [5], the deep learning (DL) approach [6], and the reinforcement learning (RL) approach [7]. However, data-driven techniques totally depend on the input data from the building system, which cannot explicitly express the physical meanings. Besides, the scope of data-driven techniques is limited, since the slight difference from the input data might leads to completely different parameters. Combining prior knowledge and mitigating the drawbacks of the two kinds of techniques, hybrid techniques (also called “grey-box” techniques), emerge to develop the thermal modeling and present the positive interpretability and model flexibility [4,34].

2.1.2 Building integrated photovoltaic systems (BIPVs)

Decentralized photovoltaic modules attached or integrated to buildings gradually have become a significant development trend in urban areas. In this context, the particular integration of PVs was termed as building integrated photovoltaic system (BIPV) [35], or building attached photovoltaic system (BAPV) [36]. Generally, the techno-logies of BIPV modeling mostly depend on AI approaches. In Ref. [36], convolutional neural network (CNN) was employed to present the nonlinear relationship between meteorological information and BAPV output. Xu etal. [37] utilized a dynamic long short-term memory (LSTM) network for the modeling of rooftop photovoltaic generations. In fact, there is no difference in modeling technologies between BIPVs and distributed PVs. Nevertheless, due to the integration mode of the BIPVs, the mathematical modeling of BIPV is quite different from that of distributed PV. In Ref. [38], a PEV charging scheduling algorithm was proposed for smart buildings based on the predicted PV output and electricity consumption. In Ref. [39], the configuration of ESSs was optimized for rooftop PV to improve the voltage profile of the low voltage distribution system. Overall, the modeling of BIPVs considers the prediction of output of PVs using AI technologies and the coexisting devices, such as PEVs, ESSs, and other controlled loads.

2.2 Information transmission modeling

Information transmission systems play a significant role in virtual world. Qolomany etal. [40] detailly introduced three main types of smart building network and home gateway technology, listed as powerline communication, busline, and wireless interconnection. Nevertheless, in this section, the impact of information transmission in the aspect of smart buildings operation is concentrated on rather than the protocol and standards of smart building information communication. All the state information of the devices is captured and transmitted through wireless or wired network to smart building energy management system for controlling. During the information transmission process, the data transmission error of loss, which is called imperfect communication, will result in the distortion of the transmitted information. The general approach for considering the impact of information transmission distortion is imperfect communications modeling.

Based on the fact of unrealistic perfect bidirectional communication, the probability of imperfect communication was calculated in Refs. [41–43]. In Ref. [41], the uncertainty of imperfect communication was considered and a spectrum resource allocation scheme was proposed to minimize the communication cost while maximizing the demand response management performance. Zhou etal. [42] utilized the absorbing Markov chain process for modeling the state transition probability of data packet communication to get the forecasting accuracy ratio of PVs and wind power plants (WPPs). Ko and Sung [43] proposed a co-simulation platform to evaluate the communication delay effect on the regulation performance provided by an EV aggregator. Imperfect communication mainly affects system operation in a short time scale, such as the real-time dispatch. Therefore, the smart building operation with the uncertainty of imperfect communication should be appropriately taken into consideration. In Fig. 3, smart devices, such as controlled loads (CLs), HVACs, PEVs, and charging piles (CPs), are controlled by a building energy management system (BEMS) by using WIFI or bluetooth [40]. Multiple BEMSs are integrated as a smart building aggregator (SBA) by utilizing wireless network, where the probability of the imperfect communication needs to be considered and modeled. In the upper hierarchy, high data reliability is guaranteed between multiple SBAs with urban VPP control center using optical fiber core network. Therefore, the imperfect communication in this hierarchy is ignored.

2.3 Human behavior description modeling

Due to the development of the configuration of IoT-based devices in smart buildings [44], the interactions between the building environment and human behavior gradually increase. As discussed in Section 1, smart building is a small but complete CPSS, thus investigating the modeling of human behavior in the loop of CPSS is meaningful for the development of smart buildings.

The common approach to analyze the connection between human factor and the CPSS is the data-driven technology [45–48]. Abrol etal. [45] presented a human-in-the-loop data-enabled feedback framework for residential building energy saving. In Ref. [46], data-driven correlation models were proposed between the energy consumption and human behavior, such as turning on lights, opening windows, and optimal energy management strategies were obtained based on the correlation models to increase energy efficiency. A Markov chain was proposed in Ref. [47] to identify the relationship between residential energy consumption and household occupant behavior. A decision tree in Ref. [48] is utilized to predict the appliance usage at homes and offices for optimal energy management.

In spite of the efficient prediction and significant energy savings with energy management, the data-driven based human behavior description approaches cannot formally model the human factor, which results in its limited scope of applications. For instance, large-scale estimations cannot explicitly represent the individual energy usage behaviors. Several recent researches offered some new insights. Aksanli and Rosing [49] presented a user-behavior model based on detailed activity sequences of household appliances to estimate the house energy consumption. Then, the flexibility of residential building was formally modeled to represent the user willingness. Yu etal. [50] quantified human perception on indoor air quality (IAQ) and offered a flexible trade-off between building cost and thermal discomfort while avoiding the out-of-limit of IAQ. Gupta etal. [51] elicited a true thermal comfort feedback from the occupants presented by proposing an incentive-based mechanism design framework.

Human behavior description modeling is of great importance for smart building operation, and the pivotal point is the way to effectively represent the relationship between human behavior and energy consumption. The following research aspects can be considered: the definition and classification of human behavior, the interactive mode/task between human and smart building, the quantification for quality of accomplished tasks by human, the uncertainty brought by human nature, such as sudden interruptions, or jumping into another task. Furthermore, the mechanism for stimulating human to perform the tasks is worthy to be taken into consideration. Through human behavior description modeling, BEMS can assign human appropriate tasks for energy savings, thus realizing optimal energy consumption by putting human in the loop.

2.4 Embodied energy and carbon of smart buildings

To make buildings smarter, hundreds of materials and components are manufactured and connected with each other. Along with the construction process, large amounts of energy are consumed with massive carbon dioxide emission.

To clearly represent the amount of energy consumption and carbon emission derived from material production and building construction, the concept of embodied energy and carbon was defined in Refs. [52–55]. Minunno etal. [56] created a benchmark to discuss the impact of buildings on the environment, revealing the embodied energy and carbon of the buildings. Alwan etal. [57] proposed a single parametric building information model and a whole life-cycle-analysis-based tool to estimate both the operational and embodied energy of UK buildings. Abd Alla etal. [58] considered the embodied energy from the perspective of practical estimations in the design process of buildings. Besides, the lifecycle energy savings of energy efficiency interventions were quantified.

With the future advancement of the low-carbon energy revolution, the embodied energy and carbon estimation of smart buildings will attract more attention in designing smart buildings in a more environment-friendly and scientific way.

2.5 Cost-benefit modeling using IoT

Since 1990s, the IoT technology has aimed to connect all physical devices by advanced information technologies. Differently, CPSS aims to operate optimally on a system level, which combines the physical, informational, and social factors into the control loop. In other words, IoT is an effective method to enable CPSS to control devices.

Before the proposal of the IoT technology, the majority of home and buildings cannot be controlled and managed. Only few commercial buildings can afford to install the BEMS. With the advancement of the IoT technology, cost-effective solutions for buildings become available, such as OpenHAB [59], HomeAssistant [60], and BEMOSS [61].

In the field of cost-benefit modeling for building devices, Zhang etal. [5] and Hu etal. [7] proposed IoT-based thermal models in smart buildings and realized accurate indoor temperature prediction and control. Cui etal. [62] focused on the cost-benefit analysis of the developed fast demand response strategy for building demand management. Zhang etal. [63] introduced an IoT-based efficient energy management considering green energy resource incorporation.

Overall, the IoT technique enables the physical devices to communicate and share information with each other. Using the collected information, the cost-benefit analysis for building devices can be conducted in a real-time manner, which ultimately facilitates prompt decision-making for smart buildings energy management.

3 Energy management and optimization

3.1 Related existing research on management and optimization models

A smart building energy management and optimization model is pivotal to achieve the holistic goals of SBA. There are primarily two broad types: building internal operation, which refers to energy savings and energy cost reduction [13–15,64–68], and building external interaction, which refers to demand response management [16,18], ancillary service provider [69–73], and peer to peer (P2P) energy sharing [17,19]. Among these two types, a coupled relationship exists between the building internal operation and external interaction. The impacts of electricity price fluctuation are considered in building internal operation, and the internal operation cost is included in building external interaction models.

3.1.1 Building internal operation

The minimum energy cost is the primary target in this kind of management model in which the optimal operations and energy savings of different controlled devices are generally considered. The procedure of energy management and optimization of this model can be summarized as follows.

1) Prediction. In a BEMS, the energy consumption prediction of different zones are the primary tasks, such as the temperature requirements of office areas, the electricity demands of the rest area, etc. The precise prediction of multi timescales can guarantee the optimal energy scheduling and device control.

2) Scheduling. After obtaining the predicted results of energy consumption, the scheduling schemes are further taken into consideration. The decisions of schemes depend on the objective of BEMS, such as energy saving, which would result in the designed shutdown of HVACs. Different scheduling schemes are decided under different time scales.

3) Transmission. When the scheduling is decided, the time series demand of the electricity of the smart building is transmitted to the distributed system operator (DSO) before the operation day. Then, in the operation day, real-time operation instructions are transmitted by BEMS to each device and various control methods are used to make the device run at the specified operation point.

4) Monitoring and response. The perception and monitoring of the operation state of devices are essential during the real-time operation of the smart building. On the other hand, BEMS needs to response for human thermal comfort and other requirements while following the decided scheduling scheme. To achieve this requirement, the real time energy consumption of controllable loads and sources are rescheduled by BEMS.

By following the above procedures, building internal operation energy management model can minimize the energy cost and realize energy savings. Nevertheless, in CPSS, this kind of energy management model gradually evolves into the foundation of the building external interaction energy management model.

3.1.2 Building external interaction

Active interaction is the emphasis in building external interaction energy management model, such as participating in electric market [74]. Different from the procedure in building internal operation, it needs to take more external factors into consideration in building external interaction energy management model. The details are summarized as follows.

1) Prediction. In building an external interaction energy management model, the electric prices in different markets are included in prediction requirements, such as the real-time market [75], the ancillary service market [76], and the demand side management [77]. Furthermore, from the DSO perspective, the electricity demand variation and ramping rate extend can be also predicted for improving the DSO scheduling flexibility.

2) Scheduling. Based on the aim of maximized profit in an electric market, the scheduling schemes are drawn up considering the operation cost and available service type of devices provided.

3) Transmission. Similar to the transmission content in building internal energy management model, the demand of electricity is transmitted to DSO in building external interaction energy management model. Furthermore, the offered service types and corresponding prices are also transmitted.

4) Monitoring and response. Compared with the content of building internal operation energy management model, the additional content in building external interaction energy management model is to follow the requirement of DSO ancillary services, such as the required response accuracy, response speed, and response delay. It is hard to follow the single smart building due to limited controlled devices, and the feasible way is the integration of smart buildings as the SBA which is ultimately controlled by the VPP control center.

There is no superiority or inferiority between these building energy management model. The interaction mode depends on the environment where the smart building is located. Nevertheless, the stable energy management relies on the remarkable control strategy.

3.2 Related, existing researches on control strategies

Typical controllable building-level loads consist of HVAC, PEV, and smart light. As a conclusion to the current researches, the control strategy within each building-level load can be summarized as follows.

3.2.1 HVAC

The methods to control a group of HVACs are extensively discussed in recent years. References [78–84] made great efforts on aggregated HAVC control for regulation service. As is identified as the potential resource for regulation service, HAVCs in smart buildings play an important role in building external interaction energy management model, thus the control strategies considering dynamic prices are also attractive to researchers, such as model predictive control (MPC) [85–87], the event-based approach [88], and the learning-based approach [10,89].

3.2.2 PEV

The researches of PEV control strategy mainly focus on the energy management of EV aggregators. In general, the EV aggregator has the full permission of the charging control if EV owners sign with the aggregator. However, the EV owners have only a temporary consumer relationship with the smart buildings so that BEMS may not be able to fully control the charging strategy of EVs. Consequently, the control strategies of the EV aggregator cannot be applied in BEMS. Nazari etal. [90] introduced recent findings on the control algorithm of EVs integrated with smart buildings. References [91,92] proposed a real-time control strategy of EV for smart buildings.

3.2.3 Smart lighting system

Combining advanced light sources, color sensors, and control algorithms, smart lighting systems can finish complex commissions with the use of sophisticated feedback control methods for the purpose of energy efficiency, lighting quality, and energy consumption minimum [93,94]. Meanwhile, dimmable lighting loads can also provide demand response [95] and frequency regulation service [96] through changing the illumination within the acceptable variation of users.

3.2.4 Virtual energy storage system (VESS)

VESS, as an excellent alternative for traditional energy storage equipment, can reduce peak-to-valley rate effectively through shifting loads. Practically, HVAC, PEV, smart lighting system, and all controllable loads in smart buildings, are aggregated coordinately as a VESS to provide grid service [97]. Hao etal. [98] proposed a first order VESS energy dynamics model including the energy state and power dispatch bounds constraints. References [99–104] developed energy-based VESS models that operate in the same way as traditional battery storage systems. Basically, the system operating with VESS has obvious advantages. Zhu etal. [105] proposed an integrated energy system scheduling method with VESS and the results showed that VESS could reduce the integrated energy system operation costs. Zhao etal. [106] proposed a novel business model to enable VESS sharing among users. The simulations indicate that VESS can reduce the physical energy storage investment of the aggregator by 54.3%. In the future, VESS will play an important role in both the demand side and the system side.

Along with the precise control strategies for building-level loads, different operation aims of BEMS come to reality. However, although smart buildings have coordinated with a variety of energy equipment on the demand side, the large number and complex management of smart buildings still remains a difficult scheduling problem for DSO. On the other hand, the existing smart building management, which only considers electricity dispatch but ignores human behavior, is insufficient. Therefore, the adjustment of smart building management to the CPSS is needed and the integration techniques for smart buildings also remain to be discussed.

4 Architecture and integration for smart buildings

Several researches have been conducted on the architecture and framework for smart buildings. Hansen etal. [107] focused on the design approach for smart houses to balance the needs of the occupants for power use and beauty of daylight. Jia et al. [108] proposed a platform-based methodology for the design of smart building applications, and then maps to their physical implementations. Li [12] briefly introduced the connectivity for building demand monitoring in a commercial building case in Singapore. Although the framework designs for the energy management of a smart building are mentioned in the above literature, they may not be suitable for the concept of CPSS. Existing studies on smart building framework design rarely discuss the design of building devices and information communication from the perspective of CPSS. Moreover, little research is conducted on the united scheduling framework for a large number of smart buildings. Duerr etal. [109] emphatically introduced a simulation framework for multiple smart buildings integrated in the district-level network. However, a top-down scheduling framework not only involves the distribution network, but also considers the coordinated operation of multiple areas. Therefore, in this section, the CPSS-framework-based design of smart building framework (termed CPSS-SBF) is presented and integration techniques of multiple smart buildings are discussed.

4.1 Proposed CPSS-SBF and its introduction

4.1.1 Introduction to CPSS-SBF structure

In Fig. 4, the framework of a CPSS-SBF is introduced. Five layers are included in the CPSS-SBF, i.e., the society layer, the cloud layer, the cyber layer, the perception layer, and the physics layer. In the society layer, human decision making affects the operation of the smart building, such as the need for energy savings, participating in demand response, and bidding in electricity market. These human actuations are collected into the cloud layer, which consists of a provincial customer side energy service platform. The advantage of the provincial platform lies in the large information throughput, which is suitable for the large-scale users accessing in the smart building.

The cyber layer mainly conveys the data from the cloud layer and the perception layer. The techniques of the Internet, the Wireless Internet, and the Powerline carrier are utilized in the cyber layer. The optimization model and control strategies mentioned in Section 3 are integrated in the perception layer. In the perception layer shown in Fig. 4, three functional areas are included, which are the computing and decision area, the information transmission area, and the sensors actuation and acquisition area, respectively. In the computing and decision area, the energy optimization model and cooperation control are the main functions while the data administration, protocol interpretation, edge computing, and real-time communication are secondary functions. The information transmission area is divided into two kinds of techniques, which are wireless transmission (such as WiFi, Zigbee, LoraWAN, Zwave, GPRS) and wire transmission (such as RS485, HPLC, KNX, DO/DI). In the sensors actuation and acquisition area, many kinds of meters are installed with building devices, such as electricity meter, water meter, gas meter, hygrothermograph, flow meter, and controllers. The physics layer contains every component in the smart building, such as HVAC, BIPV, and all power source equipment.

The information flow of CPSS-SBF is demonstrated in Fig. 5. Blue box represents the enabling function of the smart building, the yellow box represents the data interacted with human beings, and the orange box represents the underlying interaction data or instructions. The green box is the brain of the smart building, which is responsible for task scheduling and allocation.

The proposed CPSS-SBF enhances the automation, energy efficiency, human comfort, and energy savings of smart buildings. Meanwhile, from the DSO perspective, the CPSS-SBF provides more chances and convenience for smart buildings to participate in demand response service by taking the advantages of ubiquitous connectivity, comprehensive awareness, real-time control, and optimal decision making.

4.1.2 Advantage of CPSS-SBF structure

As the intellectualized level of the building is not high, the CPSS-SBF can greatly help to improve it. Nevertheless, it is of great significance to extensively spread the CPSS-SBF in many aspects.

1) Realization of intellectualized control. Currently, more than 90% of the buildings just have the function of energy use monitoring and remote manipulation. At most, 7% of the buildings have automatic systems to coordinately optimize and control all kinds of the devices. CPSS-SBF can realize the intelligent collaborative control of the devices by utilizing wireless transmission techniques. Besides, it brings the individuals into the control loop, fully reflecting the impact of human behavior on building operation. Through the frequent human-computer interaction, CPSS-SBF makes buildings more intelligent, personalized, and humanized.

2) Interactive unification of various systems in buildings. There are several independent subsystems in conventional buildings, i.e., the self-control subsystem, the smart lighting subsystem, the energy consuming monitoring subsystem, and the transformer and distribution management subsystem. Due to the nature of their independence, the communication between different subsystems requires various protocols, leading to the low efficiency of information transmission. The CPSS-SBF integrates various subsystems through the cloud interconnection technology. Different operation information is uniformly processed in the cloud layer, thus realizing the unified management from the cloud layer and the self-optimization of subsystems from the physical layer.

3) More flexible interactions with the power grid. Energy trading is generally not considered in conventional buildings. There are few interactions between the buildings and the power grid. The CPSS-SBF designs the market transaction module to enable the buildings to fully participate in the electricity market, thus changing the power consumption feature of the buildings.

In general, the CPSS-SBF reflects the interactive relationship among human, the power grid, and various devices, bringing all controllable information variables into the optimization and control loop, which changes the traditional operation mode dominated by power dispatch.

4.1.3 Disadvantage of CPSS-SBF structure

On the one hand, the information variables improve the energy management ability of buildings; on the other hand, cybersecurity threatens the normal operation of buildings. Buildings under CPSS-SBF are more vulnerable to cyber-attacks due to the high informatization degree.

Cyber-attack can be classified into denial-of-service attack, false data injection attack, and man-in-the-middle attack [110–112]. Risk assessment for cyber-attack is essential for operation strategy decision of buildings, including the cyber model assessment and the physical model assessment [113–116]. In addition, by utilizing the supervisory control and data acquisition (SCADA) and phasor measurement unit (PMU) system, researches on cyber-attack detection and mitigating were conducted in Refs. [117–120]. In the future, many efforts are to be made to further improve the security operation of buildings against cyber-attacks, such as the physical layer protection, protocol-level security principle, etc.

4.2 Integration techniques of multiple smart buildings

4.2.1 Smart buildings integrated as VPP

Though having the advantages mentioned above, it is difficult to dispatch and control large amounts of smart buildings in the grid coordinately. Therefore, a top-down dispatch framework for integration of smart buildings is needed. Up to the present, several energy integration techniques have been proposed, such as microgrid (MG) [121,122], active distribution network (ADN) [123,124], load aggregator (LA) [125,126], and virtual power plant (VPP) [127,128]. Based on Ref. [128], the integration techniques were compared in detail and listed in Table 2.

1) VPP. The concept of VPP stemmed from the framework of virtual utility in 1997 [129], defined as a flexible cooperation between public utilities, such as the distributed generators and smart buildings. VPP focuses on virtualized integrations. In other words, it mainly depends on the information interaction to coordinate resources. Three characteristics highlight the concept of VPP, i.e., information gathering and processing, geographical influence ignorance, and dynamic operation and optimization, which enable VPP to have the nature of profit-seeking.

2) MG. The requirement on components to formulate a MG is more critical and complex due to its two kinds of operation modes. When MG operates in an island mode, energy storage devices and distributed generators are essential to maintain the active power balance [130]. When MG operates in a grid-connected mode, the power exchange with the grid is restricted by its location and grid topology. Compared with VPP, MG has limited influences on the whole grid due to its regional operation features.

3) ADN. With the help of optimization techniques, communication device and monitoring system, the traditional distribution network is changed into ADN [131]. ADN can actively manage its own resources such as the distributed energy resources and accomplish its bid/offer strategies in electricity market. ADN does not operate flexibly due to the limitation of topology.

4) LA. LA mainly aggregates loads on the demand side including unregulated loads and flexible loads [132]. Load curtailment and load shifting are the main task of LA. Different from VPP, ADN, and MG, LA focuses on the management of the demand side and the generation unit is beyond the scope of LA.

Limited by the location and the topology, the management framework of MG and ADN are not suitable for smart buildings. LA and VPP have the appropriate framework for smart building management. However, the limited service provided by LA in electricity market decreases the profitability of smart buildings. On the contrast, smart buildings integrated as VPP are much more competitive in the day ahead market and the ancillary service market. Therefore, in this paper, the concept of VPP was utilized as the basis of the proposed top-down scheduling framework for the integration of smart buildings, as presented in Fig. 6.

4.2.2 Key components and techniques in SBI-VPP

Key techniques of SBI-VPP are categorized into device operation and monitoring techniques, the virtual power plant control center, and commercial strategy decision techniques. First, device operation and monitoring are important in the scheduling of SBI-VPP. The relevant content is presented in Section 4.1.1. In addition, the control center collects the information from the monitoring system and transfers control instructions to devices. Apparently, the control instructions are formulated according to the different commercial bidding strategies of the building. Therefore, commercial strategy decision techniques are needed to cope with the optimal bidding strategy in the electricity market.

1) SBI-VPP control center

In the proposed scheduling framework, both the low-level voltage area (or distribution network) and the high-level voltage area (or transmission network) are included. In the low-level voltage area, large numbers of smart buildings are randomly distributed. Then, smart buildings are basically classified and integrated by blocks. The SBA integrates the smart buildings in the same block, and the block-based SBA is formed, as manifested in Fig. 7.

After forming the block-based SBAs, the urban VPP control center integrates the SBAs together and participates in the electricity market. The hierarchical top-down scheduling framework helps reduce the difficulty in the dispatch and control of smart buildings. For independent system operator (ISO), the scheduling profiles are only needed to be sent to the urban VPP control center.

2) Commercial strategy decision techniques: external characteristic analysis techniques

In an urban VPP, there are generally four kinds of sources and loads, i.e., uncontrollable sources, uncontrollable loads, controllable sources, and controllable loads. Different loads and sources have its own operation characteristics. When they are integrated as a VPP, it is impossible for VPP to offer the kinds of operation characteristics to the ISO. Therefore, a holistic external operation characteristic is to be obtained. Traditionally, VPP only offers the adjustment range of loads to the ISO which can make a day-ahead operation scheme. However, with the increase of renewable energy connected to the grid, the ramping capacity and ramping rate of the power system become insufficient. If the VPP can offer more information to the ISO, such as its ramping rate, response time, and ramping capacity in Fig. 8, ISO can similarly consider the VPP as a generator and the operation scheme can flexibly meet the challenge from the lacking of ramping capacity and ramping rate.

In general, point-prediction techniques are frequently used on the demand side [133–135]. The red point in Fig. 9 represents the predicted results each time, and the black dotted line represents the real-time power consumption of the demand side. It is hard to obtain the global optimal scheduling plan through the small amount of predictable information. Interval prediction techniques [136] can provide more information, such as the power consumption range of the demand side which is represented by blue dotted line. Combining the two prediction techniques, the information availability on the demand side is effectively improved, thus providing more flexibility to ISO scheduling.

3) Commercial strategy decision techniques: profit allocation techniques

In general, SBI-VPP brings extra profits, compared with the sum of the individual payoffs in the market. The traditional profit allocation based on the marginal payoff increment of an aggregator is not suitable for this situation, which means that the budget will not be balanced using the traditional allocation technique. Therefore, it is essential for SBI-VPP to research the new profit allocation techniques.

The Shapley-value [137] effectively handles the profit allocation with the advantage of regardless of the entrance orders of participants. However, its calculation complexity will dramatically increase when facing a large number of participants. Subsequently, the Aumann-Shapley procedure [138] was proposed which overcame the isonomy and computation burden of the Shapley-value. Li etal. [139] utilized the Aumann-Shapley procedure to cope with the profit allocation among cooperative demand side resource aggregators.

The cost-benefit model of the participant is needed in the Aumann-Shapley procedure. In the future, the cost-benefit will be considered as the private information which is unavailable. Therefore, advanced profit allocation techniques need to be further studied.

4.3 Introduction to the practice of SBI-VPP

The practice of commercial buildings integrated as a VPP in Shanghai is constructed in 2018. Five hundred and 50 flexible resources (74% are HVAC) are concluded in the VPP which consists of 130 buildings. Since 2018, the VPP has been dispatched for more than 1200 times, and the cumulative load curtailment are more than 200 MW. The operation and monitoring platform of the VPP control center is constructed. Through the platform, the operation state of devices and total energy usage can be observed.

This engineering practice of SBI-VPP plays an important role in load curtailment during the tropical storm Rumbia landing in Shanghai. In Aug. 17, 2018, the SBI-VPP decreased the loads to about 20.12 MW in one hour, which contributed to the security operation of the grid.

SBI-VPP also contributes to the energy conservation and emission reduction. Based on the condition of the 50 MWh capacity of SBI-VPP, each building can save 30000 to 40000 kWh of electricity consumption, which is equivalent to nearly 10 tec (tons equivalent coal). The buildings can nearly reduce the emission of 2700 t carbon dioxide.

4.4 Summary and insights

In this section, the existing framework for smart buildings were discussed and a CPSS-framework-based design of smart building framework was proposed. The proposed framework is similar to the CPSS framework, which makes the smart building more intelligent and ubiquitously connected. With the development of CPSS, the smart building becomes the minimum energy management unit. To effectively manage these smart buildings, the integration techniques of multiple smart buildings were introduced and a top-down scheduling framework for smart buildings integration was proposed for less directly scheduled individuals and less computational cost. In addition, the key components and techniques for smart buildings integrated as virtual power plant were discussed and the new practice of smart buildings integrated as virtual power plant in Shanghai was introduced.

5 Opportunities and challenges for smart buildings

5.1 Available business model for SBI-VPP

The renovation investment in residential and commercial buildings according to the CPSS-SBF requires considerable profits. With the rise of the concepts of peer-to-peer (P2P) and blockchain energy trading framework [140–143], more business models for the integration of smart building will be proposed. By dividing the business model of smart buildings into two main parts, the details are summarized in the following twofold subsections.

5.1.1 Centralized energy trading

In a centralized transactive energy market, the spot market and the ancillary market are the main forms of energy transaction. In Fig. 10, the paradigm of centralized energy trading is presented. All the urban VPPs that integrate smart buildings submit their bidding prices and provide energy capacity. Then, the ISO sends to the VPPs trading profiles considering the minimum of trading adjustment and network configuration. For an urban VPP, it is necessary to identify the operation characteristic of its own response resources due to the different response requirements of different markets.

5.1.2 Decentralized energy trading

To release more trading flexibility of the power market, decentralized energy trading can be implemented in the urban VPPs by eliminating the interaction with the ISO, which is illustrated in Fig. 11. At a lower level, urban VPPs trade energy or ancillary service and determine the corresponding price and amount. At a higher level, the ISO calculates the equivalent load of each urban VPP and reconfigures the network. It is noted that the ISO is still in charge of the management of the transmission network although each urban VPP can directly sign an energy trading contract. Like the P2P decentralized energy trading, the blockchain-based energy trading is also suitable for the following management flow paradigm. Nevertheless, the difference is that the smart contracts and private/public keys are utilized among the urban VPPs and ISO.

5.2 Open research issues for SBI-VPP

Although smart buildings have been studied for many years, further researches are still needed due to the multi-disciplines intersecting. In this section, the potential research directions to reach smart buildings integrated as VPP are outlined from the perspective of awareness-optimization-control, big data, edging computing, and resilience.

5.2.1 Awareness-optimization-control

The awareness ability for device condition, human environment/behavior, and subentry energy consumption is an important feature of the smart building. Therefore, the advanced image identification for device condition, the advanced quantity techniques for human environment/behavior, and the elaborate metering techniques for energy consumption should be further studied. With the collected information, the optimization model needs to synthetically consider the energy interactive strategy for building to the grid, the energy saving and energy efficiency, and the human comfort. Furthermore, considering the different types of users in the smart building, diversified control methods should be studied for the flexible control in real time.

5.2.2 Big data

With so many metering devices installed in smart buildings, a large amount of data are collected and stored. Big data techniques help the BEMS to dig out the potential characteristics of users’ energy use and to draw users’ portraits which are helpful for flexible scheduling. In addition, from the perspective of urban VPP, data-driven evolutionary approaches [144–146] can be used to build the surrogate model instead of the practical operation model for smart buildings, with the advantage of non-intrusive data acquisition and flexible online optimization. As the data becomes more accessible, data mining and reinforcement learning techniques will be widely used in operation optimization of smart buildings.

5.2.3 Edging computing

In the developed CPSS, the smart edging computing technique is the basis for smart buildings which are considered as the terminal users. The open, extensible, multilingual, efficient and safe edging computing techniques can provide a faster response time and release the bandwidth. Some of the existing representative edging computing techniques can be utilized in smart buildings such as fog computing, mobile edge computing, cloudnet, and mist computing.

5.2.4 Resilience

Power system resilience is an important indicator in power system security and reliability [147,148]. The interaction among the cyber system, the physical system, and the social system brings new security threats. For instance, the hacking attack in the cyber system will cause the faults of devices in the physical system. It is necessary to study the resilience enhancement strategy in CPSS. With larger amount of the schedulable resources in smart buildings and its integration, the resilience enhancement, pre-fault optimal operation, and the post-fault demand response of smart buildings are the main research direction. Additionally, an emergency demand response mechanism considering resilience enhancement remains to be studied.

5.2.5 5G technology

With the overwhelming advantage of transfer speed, reliability, security, and power usage, the 5G technology attracts the attention all over the world [149]. There are also applications for 5G in the power system. The 5G technology helps electric vehicles to participate in the ancillary service and demand response by implementing massive connectivity and fast communication [150–152]. In terms of smart buildings, the possibility and implementation methods for the connection between independent subsystems through the bridge of 5G are explored and discussed [153]. Kumar etal. [154] made a comprehensive review on the 5G advanced technology which is utilized in smart hospital. With the further popularization of 5G technology, smart buildings will realize more accurate energy consumption and devices control. At that time, the corresponding control strategy and optimization methodology needs to be further studied.

5.3 Risks of SBI-VPP in economic, security and other aspects

The changes of electricity market policy, the progress of technology, the abrupt changes of global climate environment, and other practical situations will pose challenges to a built VPP. The various risks the VPP may counter in the future in terms of technology, security, economic and impact of extreme weather events are summarized.

1) Technology. Demand side management techniques for the VPP will not be applicable in the future scenarios where large numbers of users can dynamically aggregate or quit. The online demand side management technique is a potential research topic. Moreover, network security risks brought by the innovation of wireless technologies and power supply stability risks brought by the diversification of plug-in technologies still need constant attention.

2) Security. The VPP mainly faces the risks from power supply security and information transmission security which are not independent but mutually coupled. The power supply security affects the continuity of the information transmission, and the information transmission security leads to the unreliability of power supply.

3) Economic. The current commercial management model of VPP cannot adapt to the future trend of diversified power product trading, such as the flexible ramping product. In terms of market policy, the changes of permissible market participant pose economic risks to VPP. With the advancement of low carbon energy revolution, the limitation of carbon emission and the carbon trading market will bring both challenges to and opportunities for VPP.

4) Extreme weather events. In recent years, there has been a surge in EI Nino events around the world, and the extreme weather events have become more frequent. Although the VPP owns the advantage of geographical influence ignorance, the equipment failure caused by extreme weather still poses risk to the operation and information transmission of the VPP. Line hardening and uninterruptible power supply are effective solutions for enhancing the system resilience against the extreme weather.

6 Conclusions

In this paper, existing researches on the modeling of building devices and the optimization and energy management of smart buildings were reviewed. Besides, the architecture and framework for smart building in CPSS were particularly discussed. The CPSS-based operation framework for smart buildings was proposed, providing suggestions and requirements for the construction of smart buildings. In addition, an integration framework was proposed for multi smart buildings on the concept of virtual power plant after comparing various integration techniques. The key components and techniques for smart buildings integrated as VPP were outlined. Moreover, the practice of VPP in Shanghai was introduced. Furthermore, centralized and decentralized energy trading frameworks for smart buildings were outlined. A number of open research issues were also posed for the future development of smart buildings, smart building integration, as well as CPSSs. The risks for smart buildings integrated as VPP in terms of economic, technology, security, and the impact of extreme weather were alerted.

References

[1] Swilling M, Robinson B, Marvin S, City-level decoupling: urban resource flows and the governance of infrastructure transitions. A Report of the Working Group on Cities of the International Resource Panel. United Nations Environment Programme, Nairobi, Kenya, 2013

[2] Wilson M. By 2050, 70% of world’s population will be urban. Is that a good thing? 2012-3-12, available at

[3] Taha A F, Gatsis N, Dong B, Buildings-to-grid integration framework. IEEE Transactions on Smart Grid, 2019, 10(2): 1237–1249

[4] Efkarpidis N A, Christoforidis G C, Papagiannis G K. Modeling of heating and cooling energy needs in different types of smart buildings. IEEE Access: Practical Innovations, Open Solutions, 2020, 8: 29711–29728

[5] Zhang X, Pipattanasomporn M, Chen T, An IoT-based thermal model learning framework for smart buildings. IEEE Internet of Things Journal, 2020, 7(1): 518–527

[6] Zhang W, Hu W, Wen Y. Thermal comfort modeling for smart buildings: a fine-grained deep learning approach. IEEE Internet of Things Journal, 2019, 6(2): 2540–2549

[7] Hu W, Wen Y, Guan K, iTCM: toward learning-based thermal comfort modeling via pervasive sensing for smart buildings. IEEE Internet of Things Journal, 2018, 5(5): 4164–4177

[8] Yu L, Xie D, Jiang T, Distributed real-time HVAC control for cost-efficient commercial buildings under smart grid environment. IEEE Internet of Things Journal, 2018, 5(1): 44–55

[9] Ostadijafari M, Dubey A, Yu N. Linearized price-responsive HVAC controller for optimal scheduling of smart building loads. IEEE Transactions on Smart Grid, 2020, 11(4): 3131–3145

[10] Yu L, Sun Y, Xu Z, Multi-agent deep reinforcement learning for HVAC control in commercial buildings. IEEE Transactions on Smart Grid, 2021, 12(1): 407–419

[11] Sadid W H, Abobakr S A, Zhu G. Discrete-event systems-based power admission control of thermal appliances in smart buildings. IEEE Transactions on Smart Grid, 2017, 8(6): 2665–2674

[12] Li W. Application of economical building management system for Singapore commercial building. IEEE Transactions on Industrial Electronics, 2020, 67(5): 4235–4243

[13] Razmara M, Bharati G R, Shahbakhti M, Bilevel optimization framework for smart building-to-grid systems. IEEE Transactions on Smart Grid, 2018, 9(2): 582–593

[14] Mocanu E, Mocanu D C, Nguyen P H, On-line building energy optimization using deep reinforcement learning. IEEE Transactions on Smart Grid, 2019, 10(4): 3698–3708

[15] Pinzon J A, Vergara P P, da Silva L C P, Optimal management of energy consumption and comfort for smart buildings operating in a microgrid. IEEE Transactions on Smart Grid, 2019, 10(3): 3236–3247

[16] Chouikhi S, Merghem-Boulahia L, Esseghir M, A game-theoretic multi-level energy demand management for smart buildings. IEEE Transactions on Smart Grid, 2019, 10(6): 6768–6781

[17] Cui S, Wang Y, Shi Y, A new and fair peer-to-peer energy sharing framework for energy buildings. IEEE Transactions on Smart Grid, 2020, 11(5): 3817–3826

[18] Zhou Z, Zhao F, Wang J. Agent-based electricity market simulation with demand response from commercial buildings. IEEE Transactions on Smart Grid, 2011, 2(4): 580–588

[19] Cui S, Wang Y, Xiao J. Peer-to-peer energy sharing among smart energy buildings by distributed transaction. IEEE Transactions on Smart Grid, 2019, 10(6): 6491–6501

[20] Nguyen D T, Le L. Optimal bidding strategy for microgrids considering renewable energy and building thermal dynamics. IEEE Transactions on Smart Grid, 2014, 5(4): 1608–1620

[21] Yu X, Xue Y. Smart grids: a cyber–physical systems perspective. Proceedings of the IEEE, 2016, 104(5): 1058–1070

[22] Burg A, Chattopadhyay A, Lam K Y. Wireless communication and security issues for cyber–physical systems and the Internet-of-Things. Proceedings of the IEEE, 2018, 106(1): 38–60

[23] Kalluri B, Chronopoulos C, Kozine I. The concept of smartness in cyber-physical systems and connection to urban environment. Annual Reviews in Control, 2021, 51: 1–22

[24] Jin M, Jia R, Das H P, BISCUIT: building intelligent system customer investment tools. Energy Procedia, 2019, 158: 6152–6157

[25] Zhou Y, Yu F R, Chen J, Cyber-physical-social systems: a state-of-the-art survey, challenges and opportunities. IEEE Communications Surveys and Tutorials, 2020, 22(1): 389–425

[26] Samad T, Koch E, Stluka P. Automated demand response for smart buildings and microgrids: the state of the practice and research challenges. Proceedings of the IEEE, 2016, 104(4): 726–744

[27] Minoli D, Sohraby K, Occhiogrosso B. IoT considerations, requirements, and architectures for smart buildings—energy optimization and next-generation building management systems. IEEE Internet of Things Journal, 2017, 4(1): 269–283

[28] Rashidizadeh-Kermani H, Vahedipour-Dahraie M, Shafie-Khah M, A stochastic short-term scheduling of virtual power plants with electric vehicles under competitive markets. International Journal of Electrical Power & Energy Systems, 2021, 124: 106343

[29] Royapoor M, Pazhoohesh M, Davison P J, Building as a virtual power plant, magnitude and persistence of deferrable loads and human comfort implications. Energy and Building, 2020, 213: 109794

[30] Kim Y J. Optimal price based demand response of HVAC systems in multizone office buildings considering thermal preferences of individual occupants buildings. IEEE Transactions on Industrial Informatics, 2018, 14(11): 5060–5073

[31] Delcroix B. Modeling of thermal mass energy storage in buildings with phase change materials. Dissertation for the Doctoral Degree. Montréal: Université de Montréal, 2015

[32] Koo C, Park S, Hong T, An estimation model for the heating and cooling demand of a residential building with a different envelope design using the finite element method. Applied Energy, 2014, 115: 205–215

[33] Dong X, Griffo A, Wang J. Multiparameter model order reduction for thermal modeling of power electronics. IEEE Transactions on Power Electronics, 2020, 35(8): 8550–8558

[34] Beneventi F, Bartolini A, Tilli A, An effective gray-box identification procedure for multicore thermal modeling. IEEE Transactions on Computers, 2014, 63(5): 1097–1110

[35] Sechilariu M, Wang B, Locment F. Building integrated photovoltaic system with energy storage and smart grid communication. IEEE Transactions on Industrial Electronics, 2013, 60(4): 1607–1618

[36] Du L, Zhang L, Tian X. Deep power forecasting model for building attached photovoltaic system. IEEE Access: Practical Innovations, Open Solutions, 2018, 6: 52639–52651

[37] Xu X, Xu Z, Zhang R, Data-driven-based dynamic pricing method for sharing rooftop photovoltaic energy in a single apartment building. IET Generation, Transmission & Distribution, 2020, 14(24): 5720–5727

[38] Wi Y M, Lee J U, Joo S K. Electric vehicle charging method for smart homes/buildings with a photovoltaic system. IEEE Transactions on Consumer Electronics, 2013, 59(2): 323–328

[39] Tang J, Cai D, Yuan C, Optimal configuration of battery energy storage systems using for rooftop residential photovoltaic to improve voltage profile of distributed network. Journal of Engineering (Stevenage, England), 2019, 2019(16): 728–732

[40] Qolomany B, Al-Fuqaha A, Gupta A, Leveraging machine learning and big data for smart buildings: a comprehensive survey. IEEE Access: Practical Innovations, Open Solutions, 2019, 7: 90316–90356

[41] Yang C, Yao J, Lou W, On demand response management performance optimization for microgrids under imperfect communication constraints. IEEE Internet of Things Journal, 2017, 4(4): 881–893

[42] Zhou B, Zhang K, Chan K W, Optimal coordination of electric vehicles for virtual power plants with dynamic communication spectrum allocation. IEEE Transactions on Industrial Informatics, 2021, 17(1): 450–462

[43] Ko K, Sung D K. The effect of cellular network-based communication delays in an EV aggregator’s domain on frequency regulation service. IEEE Transactions on Smart Grid, 2019, 10(1): 65–73

[44] Pan J, Jain R, Paul S, An Internet of Things framework for smart energy in buildings: designs, prototype, and experiments. IEEE Internet of Things Journal, 2015, 2(6): 527–537

[45] Abrol S, Mehmani A, Kerman M, Data-enabled building energy savings (D-E BES). Proceedings of the IEEE, 2018, 106(4): 661–679

[46] Zhao T, Zhang C, Xu J, Data-driven correlation model between human behavior and energy consumption for college teaching buildings in cold regions of China. Journal of Building Engineering, 2021, 38: 102093

[47] Johnson B J, Starke M R, Abdelaziz O A, A method for modeling household occupant behavior to simulate residential energy consumption. In: Innovative Smart Grid Technologies Conference, Washington, DC, USA, 2014

[48] Basu K, Hawarah L, Arghira N, A prediction system for home appliance usage. Energy and Building, 2013, 67: 668–679

[49] Aksanli B, Rosing T S. Human behavior aware energy management in residential cyber-physical systems. IEEE Transactions on Emerging Topics in Computing, 2020, 8(1): 45–57

[50] Yu L, Xie D, Huang C, Energy optimization of HVAC systems in commercial buildings considering indoor air quality management. IEEE Transactions on Smart Grid, 2019, 10(5): 5103–5113

[51] Gupta S K, Kar K, Mishra S, Incentive-based mechanism for truthful occupant comfort feedback in human-in-the-loop building thermal management. IEEE Systems Journal, 2018, 12(4): 3725–3736

[52] Ajayi S O, Oyedele L O, Ilori O M. Changing significance of embodied energy: a comparative study of material specifications and building energy sources. Journal of Building Engineering, 2019, 23: 324–333

[53] Jia Wen T, Chin Siong H, Noor Z Z. Assessment of embodied energy and global warming potential of building construction using life cycle analysis approach: case studies of residential buildings in Iskandar Malaysia. Energy and Building, 2015, 93: 295–302

[54] Robati M, Daly D, Kokogiannakis G. A method of uncertainty analysis for whole-life embodied carbon emissions (CO₂-e) of building materials of a net-zero energy building in Australia. Journal of Cleaner Production, 2019, 225: 541–553

[55] Monahan J, Powell J C. An embodied carbon and energy analysis of modern methods of construction in housing: a case study using a lifecycle assessment framework. Energy and Building, 2011, 43(1): 179–188

[56] Minunno R, O’Grady T, Morrison G M, Investigating the embodied energy and carbon of buildings: a systematic literature review and meta-analysis of life cycle assessments. Renewable & Sustainable Energy Reviews, 2021, 143: 110935

[57] Alwan Z, Nawarathna A, Ayman R, Framework for parametric assessment of operational and embodied energy impacts utilising BIM. Journal of Building Engineering, 2021, 42: 102768

[58] Abd Alla S, Bianco V, Tagliafico L A, Life-cycle approach to the estimation of energy efficiency measures in the buildings sector. Applied Energy, 2020, 264: 114745

[59] Kreuzer K. The open home automation bus. 2019, available at the

[60] Schoutsen P. Home assistant. 2019, available at the

[61] Zhang X, Adhikari R, Pipattanasomporn M, Deploying IoT devices to make buildings smart: performance evaluation and deployment experience. In: 2016 IEEE 3rd World Forum on Internet of Things (WF-IoT), Reston, VA, USA, 2016

[62] Cui B, Gao D, Wang S, Effectiveness and life-cycle cost-benefit analysis of active cold storages for building demand management for smart grid applications. Applied Energy, 2015, 147: 523–535

[63] Zhang X, Wang D, Zhang Y, IoT driven framework based efficient green energy management in smart cities using multi-objective distributed dispatching algorithm. Environmental Impact Assessment Review, 2021, 88: 106567

[64] Yoon Y B, Seo B, Koh B B Heating energy savings potential from retrofitting old apartments with an advanced double-skin façade system in cold climate. Frontiers in Energy, 2020, 14(2): 224–240

[65] Paul S, Padhy N P. Real-time bilevel energy management of smart residential apartment building. IEEE Transactions on Industrial Informatics, 2020, 16(6): 3708–3720

[66] Lee S, Kwon B, Lee S. Joint energy management system of electric supply and demand in houses and buildings. IEEE Transactions on Power Systems, 2014, 29(6): 2804–2812

[67] Xu Z, Guan X, Jia Q, Performance analysis and comparison on energy storage devices for smart building energy management. IEEE Transactions on Smart Grid, 2012, 3(4): 2136–2147

[68] Wang J, Chen B, Li P, Distributionally robust optimization of home energy management system based on receding horizon optimization. Frontiers in Energy, 2020, 14(2): 254–266

[69] Zhao P, Henze G P, Plamp S, Evaluation of commercial building HVAC systems as frequency regulation providers. Energy and Building, 2013, 67: 225–235

[70] Lin Y, Barooah P, Meyn S P. Low-frequency power-grid ancillary services from commercial building HVAC systems. In: 2013 IEEE International Conference on Smart Grid Communications (SmartGridComm), Vancouver, Canada, 2013

[71] Blum D H, Zakula T, Norford L K. Opportunity cost quantification for ancillary services provided by heating, ventilating, and air-conditioning systems. IEEE Transactions on Smart Grid, 2017, 8(3): 1264–1273

[72] Kim Y J, Blum D H, Xu N, Technologies and magnitude of ancillary services provided by commercial buildings. Proceedings of the IEEE, 2016, 104(4): 758–779

[73] Qureshi F A, Lymperopoulos I, Khatir A A, Economic advantages of office buildings providing ancillary services with intraday participation. IEEE Transactions on Smart Grid, 2018, 9(4): 3443–3452

[74] La Q D, Chan Y W E, Soong B H. Power management of intelligent buildings facilitated by smart grid: a market approach. IEEE Transactions on Smart Grid, 2016, 7(3): 1389–1400

[75] Yoon J H, Baldick R, Novoselac A. Dynamic demand response controller based on real-time retail price for residential buildings. IEEE Transactions on Smart Grid, 2014, 5(1): 121–129

[76] Bilgin E, Caramanis M C, Paschalidis I C, Provision of regulation service by smart buildings. IEEE Transactions on Smart Grid, 2016, 7(3): 1683–1693

[77] Arun S L, Selvan M P. Dynamic demand response in smart buildings using an intelligent residential load management system. IET Generation, Transmission & Distribution, 2017, 11(17): 4348–4357

[78] Lu N, Zhang Y. Design considerations of a centralized load controller using thermostatically controlled appliances for continuous regulation reserves. IEEE Transactions on Smart Grid, 2013, 4(2): 914–921

[79] Vanouni M, Lu N. Improving the centralized control of thermostatically controlled appliances by obtaining the right information. IEEE Transactions on Smart Grid, 2015, 6(2): 946–948

[80] Muhssin M T, Cipcigan L M, Jenkins N, Dynamic frequency response from controlled domestic heat pumps. IEEE Transactions on Power Systems, 2018, 33(5): 4948–4957

[81] Hu J, Cao J, Chen M Z Q, Load following of multiple heterogeneous TCL aggregators by centralized control. IEEE Transactions on Power Systems, 2017, 32(4): 3157–3167

[82] Ma K, Yuan C, Yang J, Controller design and parameter optimization of aggregated thermostatically controlled loads for frequency regulation. In: 2016 35th Chinese Control Conference (CCC), Chengdu, China, 2016

[83] Hao H, Lin Y, Kowli A S, Ancillary service to the grid through control of fans in commercial building HVAC systems. IEEE Transactions on Smart Grid, 2014, 5(4): 2066–2074

[84] Adhikari R, Pipattanasomporn M, Rahman S. Heuristic algorithms for aggregated HVAC control via smart thermostats for regulation service. IEEE Transactions on Smart Grid, 2020, 11(3): 2023–2032

[85] Mantovani G, Ferrarini L. Temperature control of a commercial building with model predictive control techniques. IEEE Transactions on Industrial Electronics, 2015, 62(4): 2651–2660

[86] Ma Y, Matuško J, Borrelli F. Stochastic model predictive control for building HVAC systems: complexity and conservatism. IEEE Transactions on Control Systems Technology, 2015, 23(1): 101–116

[87] Wang Z, Hu G, Spanos C J. Distributed model predictive control of bilinear HVAC systems using a convexification method. In: 2017 11th Asian Control Conference (ASCC). Gold Coast, QLD, Australia, 2017

[88] Wu Z, Jia Q, Guan X. Optimal control of multiroom HVAC system: an event-based approach. IEEE Transactions on Control Systems Technology, 2016, 24(2): 662–669

[89] Zhang Z, Chong A, Pan Y, Whole building energy model for HVAC optimal control: a practical framework based on deep reinforcement learning. Energy and Building, 2019, 199: 472–490

[90] Nazari S, Borrelli F, Stefanopoulou A. Electric vehicles for smart buildings: a survey on applications, energy management methods, and battery degradation. Proceedings of the IEEE, 2021, 109(6): 1128–1144

[91] Zhang G, Tan S T, Wang G G. Real-time smart charging of electric vehicles for demand charge reduction at non-residential sites. IEEE Transactions on Smart Grid, 2018, 9(5): 4027–4037

[92] Liu Z, Wu Q, Shahidehpour M, Transactive real-time electric vehicle charging management for commercial buildings with PV on-site generation. IEEE Transactions on Smart Grid, 2019, 10(5): 4939–4950

[93] Afshari S, Mishra S. A plug-and-play realization of decentralized feedback control for smart lighting systems. IEEE Transactions on Control Systems Technology, 2016, 24(4): 1317–1327

[94] Al-Ghaili A M, Kasim H, Al-Hada N M, A review: buildings energy savings-lighting systems performance. IEEE Access : Practical Innovations, Open Solutions, 2020, 8: 76108–76119

[95] Lee C K, Liu H, Fuhs D, Smart lighting systems as a demand response solution for future smart grids. IEEE Journal of Emerging and Selected Topics in Power Electronics, 2020, 8(3): 2362–2370

[96] Liu J, Zhang W, Liu Y. Primary frequency response from the control of LED lighting loads in commercial buildings. IEEE Transactions on Smart Grid, 2017, 8(6): 2880–2889

[97] Amini M, Almassalkhi M. Optimal corrective dispatch of uncertain virtual energy storage systems. IEEE Transactions on Smart Grid, 2020, 11(5): 4155–4166

[98] Hao H, Sanandaji B M, Poolla K, Aggregate flexibility of thermostatically controlled loads. IEEE Transactions on Power Systems, 2015, 30(1): 189–198

[99] Mathieu J L, Kamgarpour M, Lygeros J, Energy arbitrage with thermostatically controlled loads. In: 2013 European Control Conference, Zurich, Switzerland, 2013

[100] Hao H, Sanandaji B M, Poolla K, Aggregate flexibility of thermostatically controlled loads. IEEE Transactions on Power Systems, 2015, 30(1): 189–198

[101] Martínez G, Liu J, Li B, Enabling renewable resource integration: the balance between robustness and flexibility. In: 2015 53rd Annual Allerton Conference on Communication, Control, and Computing, Allerton, USA, 2015

[102] Vrakopoulou M, Li B, Mathieu J L. Chance constrained reserve scheduling using uncertain controllable loads part I: formulation and scenario-based analysis. IEEE Transactions on Smart Grid, 2019, 10(2): 1608–1617

[103] Chakraborty I, Nandanoori S P, Kundu S. Virtual battery parameter identification using transfer learning based stacked autoencoder. In: 2018 17th IEEE International Conference on Machine Learning and Applications, 2018: 1269–1274

[104] Madjidian D, Roozbehani M, Dahleh M A. Energy storage from aggregate deferrable demand: fundamental trade-offs and scheduling policies. IEEE Transactions on Power Systems, 2018, 33(4): 3573–3586

[105] Zhu X, Yang J, Liu Y, Optimal scheduling method for a regional integrated energy system considering joint virtual energy storage. IEEE Access: Practical Innovations, Open Solutions, 2019, 7: 138260–138272

[106] Zhao D, Wang H, Huang J, Virtual energy storage sharing and capacity allocation. IEEE Transactions on Smart Grid, 2020, 11(2): 1112–1123

[107] Hansen E K, Hammershøj Olesen G G, Mullins M. Home smart home: a Danish energy-positive home designed with daylight. Proceedings of the IEEE, 2013, 101(11): 2436–2449

[108] Jia R, Jin B, Jin M, Design automation for smart building systems. Proceedings of the IEEE, 2018, 106(9): 1680–1699

[109] Duerr S, Ababei C, Ionel D M. SmartBuilds: an energy and power simulation framework for buildings and districts. IEEE Transactions on Industry Applications, 2017, 53(1): 402–410

[110] Li Y, Wang Y, Hu S. Online generative adversary network based measurement recovery in false data injection attacks: a cyber-physical approach. IEEE Transactions on Industrial Informatics, 2020, 16(3): 2031–2043

[111] Patel A, Purwar S. Switching attacks on smart grid using non-linear sliding surface. IET Cyber-Physical Systems: Theory & Applications, 2019, 4(4): 382–392

[112] Khalid H M, Muyeen S M, Peng J C H. Cyber-attacks in a looped energy-water nexus: an inoculated sub-observer-based approach. IEEE Systems Journal, 2020, 14(2): 2054–2065

[113] Lyu X, Ding Y, Yang S. Bayesian network based C2P risk assessment for cyber-physical systems. IEEE Access: Practical Innovations, Open Solutions, 2020, 8: 88506–88517

[114] Cardenas D J S, Hahn A, Liu C C. Assessing cyber-physical risks of IoT-based energy devices in grid operations. IEEE Access : Practical Innovations, Open Solutions, 2020, 8: 61161–61173

[115] Bhuiyan M Z A, Anders G J, Philhower J, Review of static risk-based security assessment in power system. IET Cyber-Physical Systems: Theory & Applications, 2019, 4(3): 233–239

[116] Lyu X, Ding Y, Yang S. Safety and security risk assessment in cyber-physical systems. IET Cyber-Physical Systems: Theory & Applications, 2019, 4(3): 221–232

[117] Roberts C, Scaglione A, Jamei M, Learning behavior of distribution system discrete control devices for cyber-physical security. IEEE Transactions on Smart Grid, 2020, 11(1): 749–761

[118] Venkataramanan V, Hahn A, Srivastava A C P S A M. Cyber-physical security assessment metric for monitoring microgrid resiliency. IEEE Transactions on Smart Grid, 2020, 11(2): 1055–1065

[119] Zhang Y, Krishnan V V G, Pi J, Cyber physical security analytics for transactive energy systems. IEEE Transactions on Smart Grid, 2020, 11(2): 931–941

[120] Oozeer M I, Haykin S. Cognitive risk control for mitigating cyber-attack in smart grid. IEEE Access: Practical Innovations, Open Solutions, 2019, 7: 125806–125826

[121] Liu G, Xu Y, Tomsovic K. Bidding strategy for microgrid in day-ahead market based on hybrid stochastic/robust optimization. IEEE Transactions on Smart Grid, 2016, 7(1): 227–237

[122] Yang Z, Wu R, Yang J, Economical operation of microgrid with various devices via distributed optimization. IEEE Transactions on Smart Grid, 2016, 7(2): 857–867

[123] Ding T, Liu S, Yuan W, A two-stage robust reactive power optimization considering uncertain wind power integration in active distribution networks. IEEE Transactions on Sustainable Energy, 2016, 7(1): 301–311

[124] Zheng W, Wu W, Zhang B, A fully distributed reactive power optimization and control method for active distribution networks. IEEE Transactions on Smart Grid, 2016, 7(2): 1021–1033

[125] Chen S, Chen Q, Xu Y. Strategic bidding and compensation mechanism for a load aggregator with direct thermostat control capabilities. IEEE Transactions on Smart Grid, 2018, 9(3): 2327–2336

[126] Hu J, Cao J, Guerrero J M, Improving frequency stability based on distributed control of multiple load aggregators. IEEE Transactions on Smart Grid, 2017, 8(4): 1553–1567

[127] Zhang G, Jiang C, Wang X, Bidding strategy analysis of virtual power plant considering demand response and uncertainty of renewable energy. IET Generation, Transmission & Distribution, 2017, 11(13): 3268–3277

[128] Zhang G, Jiang C, Wang X. Comprehensive review on structure and operation of virtual power plant in electrical system. IET Generation, Transmission & Distribution, 2019, 13(2): 145–156

[129] Awerbuch S, Preston A. The Virtual Utility: Accounting, Technology and Competitive Aspects of the Emerging Industry. Boston, MA: Springer US, 1997

[130] Yang Z, Wu R, Yang J, Economical operation of microgrid with various devices via distributed optimization. IEEE Transactions on Smart Grid, 2016, 7(2): 857–867

[131] Lin J, Wan C, Song Y, Situation awareness of active distribution network: roadmap, technologies, and bottlenecks. CSEE Journal of Power and Energy Systems, 2016, 2(3): 35–42

[132] Xu Y, Xie L, Singh C. Optimal scheduling and operation of load aggregators with electric energy storage facing price and demand uncertainties. In: 2011 North American Power Symposium. Boston, MA, USA, 2011

[133] Wang L, Zhu Z, Jiang C, Bi-level robust optimization for distribution system with multiple microgrids considering uncertainty distribution locational marginal price. IEEE Transactions on Smart Grid, 2021, 12(2): 1104–1117

[134] Wang L, Jiang C, Gong K, Data-driven distributionally robust economic dispatch for distribution network with multiple microgrids. IET Generation, Transmission & Distribution, 2020, 14(24): 5712–5719

[135] Sharma S, Verma A, Xu Y, Robustly coordinated Bi-level energy management of a multi-energy building under multiple uncertainties. IEEE Transactions on Sustainable Energy, 2021, 12(1): 3–13

[136] Wen S, Zhang C, Lan H, A hybrid ensemble model for interval prediction of solar power output in ship onboard power systems. IEEE Transactions on Sustainable Energy, 2021, 12(1): 14–24

[137] Zolezzi J M, Rudnick H. Transmission cost allocation by cooperative games and coalition formation. IEEE Transactions on Power Systems, 2002, 17(4): 1008–1015

[138] Molina Y P, Saavedra O R, Amarís H. Transmission network cost allocation based on circuit theory and the aumann-shapley method. IEEE Transactions on Power Systems, 2013, 28(4): 4568–4577

[139] Li B, Wang X, Shahidehpour M, Robust bidding strategy and profit allocation for cooperative DSR aggregators with correlated wind power generation. IEEE Transactions on Sustainable Energy, 2019, 10(4): 1904–1915

[140] Zhang Z, Li R, Li F. A novel peer-to-peer local electricity market for joint trading of energy and uncertainty. IEEE Transactions on Smart Grid, 2020, 11(2): 1205–1215

[141] Nezamabadi H, Vahidinasab V. Arbitrage strategy of renewable-based microgrids via peer-to-peer energy-trading. IEEE Transactions on Sustainable Energy, 2021, 12(2): 1372–1382

[142] AlAshery M K, Yi Z, Shi D, A blockchain-enabled multi-settlement quasi-ideal peer-to-peer trading framework. IEEE Transactions on Smart Grid, 2021, 12(1): 885–896

[143] Hamouda M R, Nassar M E, Salama M M A. A novel energy trading framework using adapted blockchain technology. IEEE Transactions on Smart Grid, 2021, 12(3): 2165–2175

[144] Jin Y, Wang H, Chugh T, Data-driven evolutionary optimization: an overview and case studies. IEEE Transactions on Evolutionary Computation, 2019, 23(3): 442–458

[145] Huang P, Wang H, Ma W. Stochastic ranking for offline data-driven evolutionary optimization using radial basis function networks with multiple kernels. In: 2019 IEEE Symposium Series on Computational Intelligence, Xiamen, China, 2019

[146] Huang P, Wang H, Jin Y. Offline data-driven evolutionary optimization based on tri-training. Swarm and Evolutionary Computation, 2021, 60: 100800

[147] Wang X, Shahidehpour M, Jiang C, Resilience enhancement strategies for power distribution network coupled with urban transportation system. IEEE Transactions on Smart Grid, 2019, 10(4): 4068–4079

[148] Wang X, Li Z, Shahidehpour M, Robust line hardening strategies for improving the resilience of distribution systems with variable renewable resources. IEEE Transactions on Sustainable Energy, 2019, 10(1): 386–395

[149] Gong K, Wang X, Jiang C, Security-constrained optimal sizing and siting of BESS in hybrid AC/DC microgrid considering post-contingency corrective rescheduling. IEEE Transactions on Sustainable Energy, 2021, 12(4): 2110–2122

[150] Tao M, Ota K, Dong M. Foud: integrating fog and cloud for 5G-enabled V2G networks. IEEE Network, 2017, 31(2): 8–13

[151] Zhang Y, Li J, Zheng D, Privacy-preserving communication and power injection over vehicle networks and 5G smart grid slice. Journal of Network and Computer Applications, 2018, 122: 50–60

[152] Zhang Y, Zhao J, Zheng D. Efficient and privacy-aware power injection over AMI and smart grid slice in future 5G networks. Mobile Information Systems, 2017, 2017: 1–11

[153] Zhou Y, Li L. The 5G communication technology-oriented intelligent building system planning and design. Computer Communications, 2020, 160: 402–410

[154] Kumar A, Dhanagopal R, Albreem M A, A comprehensive study on the role of advanced technologies in 5G based smart hospital. Alexandria Engineering Journal, 2021, 60(6): 5527–5536

RIGHTS & PERMISSIONS

Higher Education Press