Energies 2015,8 11001 However,these software have some limitations for simulating occupant energy-use behavior.The main limitation is that they assume the same energy use pattern for all occupants in a building,and this pattern is constant over time [18,24,28,71-73].In fact,they are not able to account for dynamic aspects of occupancy.Due to these limitation,the energy use estimated by these software normally deviates from the real levels by up to 30 percent [5,28,74].Furthermore,in addition to traditional software,traditional building management systems also have limitations with real-time inputs of occupancy-related dynamic factors,such as the number of occupants and their preferences,actions,and decisions [63].This limitation is problematic since the inputs of real-time occupancy information can reduce HVAC and lighting energy consumption by up to 20 and 30 percent,respectively [56,66,67,75].In response to these limitations in modeling occupants'energy-use behaviors,a number of studies have recently worked on various simulation techniques to attempt to overcome these particular limitations. It is noteworthy that the developed energy-modeling and simulation tools for modeling occupants' energy-related characteristics and behaviors(discussed below)are mainly used during the early phase (i.e.,design phase)of buildings [73,76-78].Such tools could help users to choose the correct size and most energy-efficient building systems and the appliances that are proportionate to the number of occupants.These tools,therefore,help to improve overall building simulation capabilities.However, to achieve the best results,the application of these tools should be very sensitive to occupants'input parameters to accurately represent occupants'actions [73,79].In fact,these tools could be used to analyze the specific dynamics for all individual occupants,and could be calibrated to ensure that they can be used for all sizes of commercial buildings with different numbers of occupants.Researchers might also set the simulations to consider the decreased occupancy of after-hours and non-working days. To maximize the benefits of such software,the systems should be flexible enough to consider all possible occupant actions as well as all of the common practices of occupants. In addition to simulating the design phase of the buildings,simulation tools could also be used during other phases such as the construction and operation phases [28,63,75,80-83].For instance,within the renovation phase of buildings,such tools could help decision makers choose the most efficient appliances/systems when making a purchase.In addition,the use of such simulation techniques would help avoid the real resource-intensive process of testing which appliances and systems work well for a building.Time of a run,accuracy,and versatility(i.e.,solving different occupancy problems in any commercial building)are the main criteria that must be evaluated for occupancy simulation tools [50]. Many effective options are discussed below. 3.1.Agent-Based Modeling Simulation research has indicated that occupants'dynamic energy use patterns can result in significant variations in energy consumption in the commercial sector [28].In particular,a significant number of simulations employed Agent-Based Modeling (ABM)techniques to overcome software limitations in order to simulate actions and interactions between occupants and their built environment.These simulations sought to better predict building operational energy performance during the design phase.ABM is a kind of computational model that simulates the actions and interactions of agents with each other and their environments [84];in ABM,building occupants are agents in the built environment.Unlike most mathematical models,ABM agents have heterogeneous features and abilities [85]
Energies 2015, 8 11001 However, these software have some limitations for simulating occupant energy-use behavior. The main limitation is that they assume the same energy use pattern for all occupants in a building, and this pattern is constant over time [18,24,28,71–73]. In fact, they are not able to account for dynamic aspects of occupancy. Due to these limitation, the energy use estimated by these software normally deviates from the real levels by up to 30 percent [5,28,74]. Furthermore, in addition to traditional software, traditional building management systems also have limitations with real-time inputs of occupancy-related dynamic factors, such as the number of occupants and their preferences, actions, and decisions [63]. This limitation is problematic since the inputs of real-time occupancy information can reduce HVAC and lighting energy consumption by up to 20 and 30 percent, respectively [56,66,67,75]. In response to these limitations in modeling occupants’ energy-use behaviors, a number of studies have recently worked on various simulation techniques to attempt to overcome these particular limitations. It is noteworthy that the developed energy-modeling and simulation tools for modeling occupants’ energy-related characteristics and behaviors (discussed below) are mainly used during the early phase (i.e., design phase) of buildings [73,76–78]. Such tools could help users to choose the correct size and most energy-efficient building systems and the appliances that are proportionate to the number of occupants. These tools, therefore, help to improve overall building simulation capabilities. However, to achieve the best results, the application of these tools should be very sensitive to occupants’ input parameters to accurately represent occupants’ actions [73,79]. In fact, these tools could be used to analyze the specific dynamics for all individual occupants, and could be calibrated to ensure that they can be used for all sizes of commercial buildings with different numbers of occupants. Researchers might also set the simulations to consider the decreased occupancy of after-hours and non-working days. To maximize the benefits of such software, the systems should be flexible enough to consider all possible occupant actions as well as all of the common practices of occupants. In addition to simulating the design phase of the buildings, simulation tools could also be used during other phases such as the construction and operation phases [28,63,75,80–83]. For instance, within the renovation phase of buildings, such tools could help decision makers choose the most efficient appliances/systems when making a purchase. In addition, the use of such simulation techniques would help avoid the real resource-intensive process of testing which appliances and systems work well for a building. Time of a run, accuracy, and versatility (i.e., solving different occupancy problems in any commercial building) are the main criteria that must be evaluated for occupancy simulation tools [50]. Many effective options are discussed below. 3.1. Agent-Based Modeling Simulation research has indicated that occupants’ dynamic energy use patterns can result in significant variations in energy consumption in the commercial sector [28]. In particular, a significant number of simulations employed Agent-Based Modeling (ABM) techniques to overcome software limitations in order to simulate actions and interactions between occupants and their built environment. These simulations sought to better predict building operational energy performance during the design phase. ABM is a kind of computational model that simulates the actions and interactions of agents with each other and their environments [84]; in ABM, building occupants are agents in the built environment. Unlike most mathematical models, ABM agents have heterogeneous features and abilities [85]
Energies 2015,8 11002 Li et al.[86]employed ABM to simulate occupant load in HVAC design in order to optimize HVAC system size.By simulating the correct occupancy behavior characteristics,the model estimated a more accurate load and effectively designed an HVAC system that saved up to 43 percent of total energy.The number of occupants in each specific space at a given time became the main parameter of their proposed model.Erickson et al.[75]also used ABM to optimize HVAC loading and showed a total energy reduction of 14 percent at the room level of commercial buildings.They used wireless camera sensor networks to find occupants'mobility patterns in buildings.Then,they employed ABM to simulate the mobility patterns for various control strategies of HVAC.Li et al.'s [86]and Erickson et al.'s [75] approaches feed various dynamic occupants'information into the ABM simulation tools in order to directly calculate the HVAC loads.HVAC controls the indoor comfort;however,in their models,they did not clearly respond to the ventilation requirement that decreases CO2 levels inside the building. Lee and Malkawi [81]developed an ABM tool that simulates multiple occupant behaviors(i.e.,adjusted clothing levels,adjusted activity levels,window use,blind use,and space heater/personal fan use)in order to predict such behavior changes due to changes in climate and buildings topologies.Their proposed tool is an open architecture program that can adapt to different building functions and climate topologies, and that provides opportunities for an occupant to make decisions based on his/her thermal comfort level. However,this tool cannot track the thermal comfort conditions of individual occupants to fully understand whether they are satisfied with the thermal comfort level.Azar and Menassa [28,80]proposed an ABM technique to simulate the diverse and dynamic energy-use patterns of occupants and their behavior changes over time.This technique also considers various interactions among occupants.Compared to common energy software,their proposed model showed a 25 percent reduction in energy use at a small office due to the correct modeling of occupant behavior.However,this technique is limited to interactions of occupants within a room,and could not account for occupants'interactions in different rooms of a building.Such interactions may be considered to achieve more realistic results. Furthermore,social network type and structure can affect occupants'energy-use behaviors. The commercial sector frequently has complex social structures due to presence of multiple independent entities within the same building [87].In most commercial buildings in the United States,at least two companies (i.e.,entities)work in the same building [88].Some researchers recently employed ABM to simulate interactions of occupants in different entities within a commercial building.ABM can also differentiate the impact of various dynamic interactions of occupants from different social structures/networks [89],which greatly affect occupants'energy use behaviors [32,90].Anderson et al.[78] applied ABM to simulate the interactions of heterogeneous building occupants in their social networks to examine how social network type and structure can affect occupants'energy use behaviors.They considered four social network types:random graph,scale-free network,small-world network,and regular ring lattice.The results from their case study of a commercial building with different social network structures and connectivity levels proved that network type and structure hold significant influence over an occupant's energy-use behavior.Anderson and Lee [91]employed ABM to evaluate the effect of static and dynamic social networks on occupants'energy-use behavior.Their results indicated that dynamic networks increase the uncertainties of energy behavior and therefore have more influence on occupant energy behavior than static networks.However,Anderson et al.[78]and Anderson and Lee [91]did not mention at what rate occupants'energy-use behaviors can be affected.Finding a rate for behavioral change would better indicate how different social networks affect occupants
Energies 2015, 8 11002 Li et al. [86] employed ABM to simulate occupant load in HVAC design in order to optimize HVAC system size. By simulating the correct occupancy behavior characteristics, the model estimated a more accurate load and effectively designed an HVAC system that saved up to 43 percent of total energy. The number of occupants in each specific space at a given time became the main parameter of their proposed model. Erickson et al. [75] also used ABM to optimize HVAC loading and showed a total energy reduction of 14 percent at the room level of commercial buildings. They used wireless camera sensor networks to find occupants’ mobility patterns in buildings. Then, they employed ABM to simulate the mobility patterns for various control strategies of HVAC. Li et al.’s [86] and Erickson et al.’s [75] approaches feed various dynamic occupants’ information into the ABM simulation tools in order to directly calculate the HVAC loads. HVAC controls the indoor comfort; however, in their models, they did not clearly respond to the ventilation requirement that decreases CO2 levels inside the building. Lee and Malkawi [81] developed an ABM tool that simulates multiple occupant behaviors (i.e., adjusted clothing levels, adjusted activity levels, window use, blind use, and space heater/personal fan use) in order to predict such behavior changes due to changes in climate and buildings topologies. Their proposed tool is an open architecture program that can adapt to different building functions and climate topologies, and that provides opportunities for an occupant to make decisions based on his/her thermal comfort level. However, this tool cannot track the thermal comfort conditions of individual occupants to fully understand whether they are satisfied with the thermal comfort level. Azar and Menassa [28,80] proposed an ABM technique to simulate the diverse and dynamic energy-use patterns of occupants and their behavior changes over time. This technique also considers various interactions among occupants. Compared to common energy software, their proposed model showed a 25 percent reduction in energy use at a small office due to the correct modeling of occupant behavior. However, this technique is limited to interactions of occupants within a room, and could not account for occupants’ interactions in different rooms of a building. Such interactions may be considered to achieve more realistic results. Furthermore, social network type and structure can affect occupants’ energy-use behaviors. The commercial sector frequently has complex social structures due to presence of multiple independent entities within the same building [87]. In most commercial buildings in the United States, at least two companies (i.e., entities) work in the same building [88]. Some researchers recently employed ABM to simulate interactions of occupants in different entities within a commercial building. ABM can also differentiate the impact of various dynamic interactions of occupants from different social structures/networks [89], which greatly affect occupants’ energy use behaviors [32,90]. Anderson et al. [78] applied ABM to simulate the interactions of heterogeneous building occupants in their social networks to examine how social network type and structure can affect occupants’ energy use behaviors. They considered four social network types: random graph, scale-free network, small-world network, and regular ring lattice. The results from their case study of a commercial building with different social network structures and connectivity levels proved that network type and structure hold significant influence over an occupant’s energy-use behavior. Anderson and Lee [91] employed ABM to evaluate the effect of static and dynamic social networks on occupants’ energy-use behavior. Their results indicated that dynamic networks increase the uncertainties of energy behavior and therefore have more influence on occupant energy behavior than static networks. However, Anderson et al. [78] and Anderson and Lee [91] did not mention at what rate occupants’ energy-use behaviors can be affected. Finding a rate for behavioral change would better indicate how different social networks affect occupants’
Energies 2015,8 11003 behaviors.Such studies would also improve if they could find which types of networks are most common in commercial buildings.In addition,they could find whether there is any relationship between the building type and network type. Azar and Menassa [12,87]used ABM to model occupancy-related behaviors in social sub-networks to show how occupants'interactions impact the energy-use of buildings.They tested various numbers of sub-networks in a typical United States'commercial building,and concluded that traditional modeling techniques (such as single-network modeling and bounded confidence models)are not applicable to simulate social networks and sub-networks in commercial buildings.However,in their studies,they did not considered the four main social network types studied by Anderson et al.[78].In fact,they only considered the small-world and scale-free network.Studying all social network types could be more effective to show the limitations of traditional modeling techniques. 3.2.Multi Agent Systems Compared to ABM,Multi Agent Systems(MAS)provide the opportunity for agents(i.e.,occupants) to communicate more with each other as well as with their built environment.MAS divides a complex problem into sub-problems solved by representative agents [63];for this reason,this approach is employed to model complex problems with multiple cyber agents.ABM is related to,but clearly distinct from,the MAS concept [92].A MAS can contain combined ABM,and in cases where the problem of energy saving is a multi-dimensional problem,MAS is an appropriate application [92,93].MAS may balance between occupants'preferences and energy saving;ABM fails to achieve this aim.In fact, concerning the commercial sector,MAS typically helps make tradeoffs between both building demands and occupant comfort [94,95]. Qiao et al.[96]introduced some prospects to indicate how MAS can simulate occupant behaviors to adjust device control in commercial buildings.Dounis and Caraiscos [93]presented MAS architecture for energy efficiency and comfort in built environments.They indicated that various advanced techniques(e.g.,Fuzzy Logic,Markov Chain Model,and Neural Networks)are implementing methods used in order to develop a MAS tool for improving the efficiency of building control systems.In addition, their simulation results from implementing MAS on a building showed that this model can manage occupants'preferences for thermal and luminance comfort,indoor air quality,and energy conservation. However,they did not clearly respond to the balance between thermal comfort and energy conversation. In some cases,achieving a level of thermal comfort could lead to an increase in energy consumption. They proposed MAS architecture for managing both energy efficiency and occupant comfort,and conducted a tradeoff between these two parties is needed.Klein et al.[63]proposed a MAS tool to model the management and control of appliances and occupants in a building.Their model could simulate and predict how changes to the building,occupant behavior(i.e.,preferences and schedule),and operational policies affect energy use and occupant comfort.In fact,their model simulated occupancy behavior as well as building operational policies.Based on their results from employing the model on a case study of a three-story university building,an improvement in occupants'comfort level and a reduction in energy consumption were realized.For this model,some data needed to be manually input.However, since such models need a large group of input data to simulate and predict energy use and occupant
Energies 2015, 8 11003 behaviors. Such studies would also improve if they could find which types of networks are most common in commercial buildings. In addition, they could find whether there is any relationship between the building type and network type. Azar and Menassa [12,87] used ABM to model occupancy-related behaviors in social sub-networks to show how occupants’ interactions impact the energy-use of buildings. They tested various numbers of sub-networks in a typical United States’ commercial building, and concluded that traditional modeling techniques (such as single-network modeling and bounded confidence models) are not applicable to simulate social networks and sub-networks in commercial buildings. However, in their studies, they did not considered the four main social network types studied by Anderson et al. [78]. In fact, they only considered the small-world and scale-free network. Studying all social network types could be more effective to show the limitations of traditional modeling techniques. 3.2. Multi Agent Systems Compared to ABM, Multi Agent Systems (MAS) provide the opportunity for agents (i.e., occupants) to communicate more with each other as well as with their built environment. MAS divides a complex problem into sub-problems solved by representative agents [63]; for this reason, this approach is employed to model complex problems with multiple cyber agents. ABM is related to, but clearly distinct from, the MAS concept [92]. A MAS can contain combined ABM, and in cases where the problem of energy saving is a multi-dimensional problem, MAS is an appropriate application [92,93]. MAS may balance between occupants’ preferences and energy saving; ABM fails to achieve this aim. In fact, concerning the commercial sector, MAS typically helps make tradeoffs between both building demands and occupant comfort [94,95]. Qiao et al. [96] introduced some prospects to indicate how MAS can simulate occupant behaviors to adjust device control in commercial buildings. Dounis and Caraiscos [93] presented MAS architecture for energy efficiency and comfort in built environments. They indicated that various advanced techniques (e.g., Fuzzy Logic, Markov Chain Model, and Neural Networks) are implementing methods used in order to develop a MAS tool for improving the efficiency of building control systems. In addition, their simulation results from implementing MAS on a building showed that this model can manage occupants’ preferences for thermal and luminance comfort, indoor air quality, and energy conservation. However, they did not clearly respond to the balance between thermal comfort and energy conversation. In some cases, achieving a level of thermal comfort could lead to an increase in energy consumption. They proposed MAS architecture for managing both energy efficiency and occupant comfort, and conducted a tradeoff between these two parties is needed. Klein et al. [63] proposed a MAS tool to model the management and control of appliances and occupants in a building. Their model could simulate and predict how changes to the building, occupant behavior (i.e., preferences and schedule), and operational policies affect energy use and occupant comfort. In fact, their model simulated occupancy behavior as well as building operational policies. Based on their results from employing the model on a case study of a three-story university building, an improvement in occupants’ comfort level and a reduction in energy consumption were realized. For this model, some data needed to be manually input. However, since such models need a large group of input data to simulate and predict energy use and occupant
Energies 2015,8 11004 comfort in a commercial building,the process of inputting the data into these tools needs to be totally automated in order to facilitate the tool's operation. 3.3.Other Techniques In addition to ABM and MAS,some researchers have proposed other models and techniques aimed at simulating occupants'energy-related characteristics.Yamada et al.[97]developed a system that combines neural networks,fuzzy systems,and predictive control in order to control air-condition systems.Their system can predict the number of occupants in order to estimate building performance to achieve energy savings and high comfort levels for indoor conditions.However,neural network-and fuzzy system-based models typically need a training process,and for Yamada et al.'s [97]developed tool,this training process needs a considerable amount of time.Their proposed system therefore needs to be improved in its training level.Yamada et al.[97]also considered only the temperature as an indicator for comfort level.Such works on comfort level may consider other aspects of indoor comfort, such as humidity and air speed.Wang et al.[98]proposed a Markov chain-based model for building-occupancy simulations in commercial buildings;the model can simulate occupants'stochastic movements in order to predict each occupant's location.It can also produce nonsynchronous occupants' location-changes according to the time and distribution of occupants in space;such predictions become inputs for building management processes for energy savings.However,they validated the model by single offices,which is problematic since for such studies,more cases-especially multiple offices-need to be considered to study occupants'stochastic movements.Jazizadeh et al.[82,83]developed a framework that models occupants'thermal preference profiles into HVAC control logic in order to set room conditions at occupants'desired temperatures.They employed a fuzzy based model to put occupants'comfort profiles into the framework.The results from their test bed of a university building showed up to a 40 percent reduction in HVAC daily average airflow.However,similar to Dounis and Caraiscos [93],they did not clearly respond to the balance between thermal comfort and energy conversation,which is important since achieving a level of thermal comfort might lead to increasing total energy consumption of a building.Zhao et al.[99]developed a practical data-mining approach that collects the energy consumption data of various systems and appliances within office spaces to find occupants'passive energy behaviors.The proposed data-mining approach is based on nominal classification(ie.,C4.5 decision tree,locally weighted naive bayes,and support vector machine)and numeric regression algorithms (i.e., linear regression and support vector regression).The approach has the capability to separately find the behaviors of individual occupants and the schedule of an occupant groups and use this information to set various office appliances and systems in order to reduce the energy consumption.However,the validity of their proposed data-mining approach was limited to data that may have included some incorrect outcomes;such data-mining models require a considerable sample of validated data to test the models and show their effectiveness.Hong et al.[18]presented a framework,DNAs,to observe and simulate occupant energy use behaviors in built environments.This framework is developed based on four key components:(a)drivers of occupants'energy-related behaviors;(b)needs of occupants,(c)actions carried out by occupants;and (d)building's systems acted on by occupants.Such occupancy components directly and indirectly influence building's energy consumption,and therefore DNAs provide the
Energies 2015, 8 11004 comfort in a commercial building, the process of inputting the data into these tools needs to be totally automated in order to facilitate the tool’s operation. 3.3. Other Techniques In addition to ABM and MAS, some researchers have proposed other models and techniques aimed at simulating occupants’ energy-related characteristics. Yamada et al. [97] developed a system that combines neural networks, fuzzy systems, and predictive control in order to control air-condition systems. Their system can predict the number of occupants in order to estimate building performance to achieve energy savings and high comfort levels for indoor conditions. However, neural network- and fuzzy system-based models typically need a training process, and for Yamada et al.’s [97] developed tool, this training process needs a considerable amount of time. Their proposed system therefore needs to be improved in its training level. Yamada et al. [97] also considered only the temperature as an indicator for comfort level. Such works on comfort level may consider other aspects of indoor comfort, such as humidity and air speed. Wang et al. [98] proposed a Markov chain-based model for building-occupancy simulations in commercial buildings; the model can simulate occupants’ stochastic movements in order to predict each occupant’s location. It can also produce nonsynchronous occupants’ location-changes according to the time and distribution of occupants in space; such predictions become inputs for building management processes for energy savings. However, they validated the model by single offices, which is problematic since for such studies, more cases—especially multiple offices—need to be considered to study occupants’ stochastic movements. Jazizadeh et al. [82,83] developed a framework that models occupants’ thermal preference profiles into HVAC control logic in order to set room conditions at occupants’ desired temperatures. They employed a fuzzy based model to put occupants’ comfort profiles into the framework. The results from their test bed of a university building showed up to a 40 percent reduction in HVAC daily average airflow. However, similar to Dounis and Caraiscos [93], they did not clearly respond to the balance between thermal comfort and energy conversation, which is important since achieving a level of thermal comfort might lead to increasing total energy consumption of a building. Zhao et al. [99] developed a practical data-mining approach that collects the energy consumption data of various systems and appliances within office spaces to find occupants’ passive energy behaviors. The proposed data-mining approach is based on nominal classification (i.e., C4.5 decision tree, locally weighted naïve bayes, and support vector machine) and numeric regression algorithms (i.e., linear regression and support vector regression). The approach has the capability to separately find the behaviors of individual occupants and the schedule of an occupant groups and use this information to set various office appliances and systems in order to reduce the energy consumption. However, the validity of their proposed data-mining approach was limited to data that may have included some incorrect outcomes; such data-mining models require a considerable sample of validated data to test the models and show their effectiveness. Hong et al. [18] presented a framework, DNAs, to observe and simulate occupant energy use behaviors in built environments. This framework is developed based on four key components: (a) drivers of occupants’ energy-related behaviors; (b) needs of occupants, (c) actions carried out by occupants; and (d) building’s systems acted on by occupants. Such occupancy components directly and indirectly influence building’s energy consumption, and therefore DNAs provide the
Energies 2015,8 11005 opportunities to incorporate more energy-related behaviors into simulation tools.In addition, this framework has the capability to evolve into BIM Another approach,Relative Agreement(RA)modeling,is an extension of a Bounded Confidence model [100]that can take into account different energy use characteristics of occupants,uncertainties about their opinion dynamics,and their interactions to each other.RA was defined and introduced by Deffuant et al.[101-103],and it can consider occupants as a population of agents that are selected randomly to interact with each other.In addition,each occupant (i.e.,agent)is characterized by two variables:its opinion,and its uncertainty [100,101].These two variables change over time.The ABM model developed by Azar and Menassa [12,87]is based on an RA concept.Additionally,Verplanken and Wood [104]and Gockeritz et al.[105]employed RA concepts to simulate pre-environmental behaviors of occupants in order to understand occupants'responses to the new energy characteristics of their built environment.Their results shows that an occupant's energy-conserving behavior is highly connected to his/her belief regarding other occupants'energy-conserving behaviors. Figure 2 shows the framework of current research.Although MAS tools have potential to simultaneously integrate ABM and other techniques for simulating occupancy related behavior [93], such MAS tools have not been directly addressed by literature.In this context,hybrid simulation approaches could be proposed. [28[75][86][81] [63][93]1 Markov Logic [98] Predictive Control [97]I ABM [80[78][91] 1[96 MAS [93]1 Fuzzy Logie [97][83][82]1 [12][87] [93]i Neural Network [97] RA[104[105] Nominal and Numerical Classifications[99] Figure 2.Framework of current research. 4.Improving Occupant Energy-Consuming Behaviors Improving occupant energy-consuming behaviors is a more cost-effective technique for cutting energy consumption than improving building's physical properties [18-20].Failure to improve occupant behaviors undermines the investment in retrofitting building envelopes and appliances since occupants define the success of such sustainable retrofitting projects [13,31].Furthermore,if occupants learn appropriate energy-saving behaviors,they can practice such behaviors in all buildings.Therefore,adopting energy-saving behaviors among occupants would then provide an opportunity for general energy savings within all built environments. Changing energy-use behaviors and motivating occupants to have sustainable behaviors are typically achieved by providing intervention tools for their behaviors and habits in order to improve the occupant's intentions and beliefs [45].Such interventions have used several techniques (e.g.,prompts,providing information and feedback,goal setting,and motivations)to attempt to improve occupant behavior,and each technique has had a level of success in reducing energy consumption [91,106-109].Generally, there are two main occupancy-focused intervention approaches(see Figure 3)[12,79]:(1)continuous
Energies 2015, 8 11005 opportunities to incorporate more energy-related behaviors into simulation tools. In addition, this framework has the capability to evolve into BIM. Another approach, Relative Agreement (RA) modeling, is an extension of a Bounded Confidence model [100] that can take into account different energy use characteristics of occupants, uncertainties about their opinion dynamics, and their interactions to each other. RA was defined and introduced by Deffuant et al. [101–103], and it can consider occupants as a population of agents that are selected randomly to interact with each other. In addition, each occupant (i.e., agent) is characterized by two variables: its opinion, and its uncertainty [100,101]. These two variables change over time. The ABM model developed by Azar and Menassa [12,87] is based on an RA concept. Additionally, Verplanken and Wood [104] and Göckeritz et al. [105] employed RA concepts to simulate pre-environmental behaviors of occupants in order to understand occupants’ responses to the new energy characteristics of their built environment. Their results shows that an occupant’s energy-conserving behavior is highly connected to his/her belief regarding other occupants’ energy-conserving behaviors. Figure 2 shows the framework of current research. Although MAS tools have potential to simultaneously integrate ABM and other techniques for simulating occupancy related behavior [93], such MAS tools have not been directly addressed by literature. In this context, hybrid simulation approaches could be proposed. Figure 2. Framework of current research. 4. Improving Occupant Energy-Consuming Behaviors Improving occupant energy-consuming behaviors is a more cost-effective technique for cutting energy consumption than improving building’s physical properties [18–20]. Failure to improve occupant behaviors undermines the investment in retrofitting building envelopes and appliances since occupants define the success of such sustainable retrofitting projects [13,31]. Furthermore, if occupants learn appropriate energy-saving behaviors, they can practice such behaviors in all buildings. Therefore, adopting energy-saving behaviors among occupants would then provide an opportunity for general energy savings within all built environments. Changing energy-use behaviors and motivating occupants to have sustainable behaviors are typically achieved by providing intervention tools for their behaviors and habits in order to improve the occupant’s intentions and beliefs [45]. Such interventions have used several techniques (e.g., prompts, providing information and feedback, goal setting, and motivations) to attempt to improve occupant behavior, and each technique has had a level of success in reducing energy consumption [91,106–109]. Generally, there are two main occupancy-focused intervention approaches (see Figure 3) [12,79]: (1) continuous