Technologies for Reliable Data Acquisition

In the next two sections, a comprehensive review is provided on state-of-the-art technologies that are used for relevant reliable data acquisition, followed by modeling methodologies of how building occupants react to indoor environment and affect energy use.

1.1  Data acquisition techniques

With the rapid development of advanced technology in the field of electrical and computer engineering in the past decade, numerous tools have been created to help facilitate data collection and enhance data accuracy. These tools are usually applied with the methods and objectives of the research. Therefore, various types of devices such as sensors, cameras or meters are selected to assist the process of the modeling.

Wireless Sensor Networks (WSN) are the most common and popular tool for monitoring occupant-related variables such as temperature, humidity, carbon dioxide, sounds, and illumination, etc. [1]. WSN consist of sensor nodes that can be distributed throughout the buildings. By using wireless technology, operation and maintenance costs are reduced as no cabling is required. Wireless sensors can be deployed in a remote place where some of the wired devices may not be able to reach. In addition, by forming a network, sensor nodes will be able to communicate and exchange information with each other, and the data could be logged in a more organized way simultaneously.

Previous research showed advantages by using WSN in building control systems. The merit of the implantation of WSN can reflect the presence and absence of people in individual office, therefore improvise the heating, ventilation and air conditioning (HVAC) control system based on the real-time occupancy data for improving the energy efficiency. Yang et al. [2] used ambient sensors to build occupancy models to support demand-response HVAC controls that are intended for automation intelligence and energy conservation in commercial buildings. Twelve ambient sensor variables were selected to evaluate real-time occupancy conditions by monitoring the environmental parameters including humidity, temperature, carbon dioxide concentration, light, sound, and motion. Six machine learning algorithms were performed for estimating occupancy.  The decision-tree technique performed the best overall accuracy and can count numbers of occupants at room level. The results demonstrated that the implementation of occupancy-based demand-responses HVAC control can save energy consumption effectively, such as 20% of gas and 18% of electricity. Agarwal et al. [3] conducted a study which presented the design and implementation of a presence sensor platform in the building management system (BMS). The sensor was low-cost and incrementally deployable using battery operated wireless sensor nodes. Then, an algorithm was developed to estimate occupancy at the level of individual offices. The results revealed that the measurements can detect occupancy with a high accuracy. Furthermore, using a building simulation framework and the occupancy information from their testbed, the system showed potential energy saving from 10% to 15%. 

Other types of devices such as smart phone and camera also widely used in the research of monitoring occupant behavior. Chen and Ahn [4] used Wi-Fi as occupant tracking technology in their research with an aim to find the correlation between Wi-Fi connection event and electricity consumption. The ponied out that Wi-Fi network worked better than multiple sensors. The explanation was that Wi-Fi would be able to track occupancy and the locations of occupants based on access points and signal strength. In addition, there is nearly no cost for this technology because of the wide use of smartphones nowadays. The only concern of Wi-Fi is that the occupant positions can be easily obtained however other detailed information is hardly captured. Wireless camera networks can solve the problem, Erickson et al. [5] developed a system consists of 16 sensor nodes on the testbed building. Each sensor node was comprised of a camera interfaced with an adapter board. Then the captured images by the system were processed based on the object detection algorithm. The complex image processing algorithm finally generated an array of data containing information needed for occupancy modeling.

1.2  Modeling methodologies

Due to the advanced data collection technologies, a variety of important data can be collected to model occupancy and occupant behavior that can be further integrated in energy simulation software. Jia and Srinivasan [1] categorized the occupant behavior modeling methodologies into four areas: agent-based modeling, statistical analysis, data mining approaches, and stochastic process modeling.

1.2.1        Agent-based modeling

Agent-based modeling (ABM) is a computational model for simulation of objects interaction with each other and the external environment. The model is based on regulated rules which enable assessing the effects on the whole system [1].

Zimmermann [6] is one of the earliest researchers who proposed the idea that occupant activities in office could be modeled base on the communicating agents. But the research did not emphasize the occupant interaction with building control and service strategies. Klein et al. [7] carried out a research to investigate the interaction of occupants and building systems by using ABM. They developed their own multi-agent comfort and energy system to compare four distinct control management strategies. Compared with results of energy and comfort from baseline, reactive, proactive strategies, they stated that proactive-Markov Decision Problems (MDPs) showed the best result of building operation and intelligent coordination with devices and occupants. 

ABM was demonstrated as a good way to mimic occupant behaviors and to model interaction between different occupants. Lee and Malkawi [8] used ABM to explore how different behaviors affect building energy use. They assumed there were three categories of beliefs that could cause different occupant behaviors. The cost function for each agent was expressed in a function consists of beliefs, time, and weight coefficients. The model was able to adjust relevant parameters to get different results. Azar and Menassa [9] applied ABM in their research to identify the energy consuming behaviors and the related factors that cause behavioral change exclusively. Their model was developed based on an interactive tool for data input. After inputting energy consumption rate by occupant behavior type, the model could simulate for energy use variation. Langevin et al. [10] presented a detailed ABM using thermal comfort and behavior data from a field study in an office building. This model assigned building occupant agents dynamics for clothing, metabolic rate, thermal acceptability and behavior choice hierarchy. The approach provided a platform for more flexible simulations based on the interactions between occupants and surrounding built environments.

References

1. Jia, M. and R.S. Srinivasan, Occupant behavior modeling for smart buildings: A critical review of data acquisition technologies and modeling methodologies, in 2015 Winter Simulation Conference (WSC). 2015, IEEE: Huntington Beach, CA, USA.
2. Yang, Z., N. Li, B. Becerik-Gerber, and M. Orosz, A systematic approach to occupancy modeling in ambient sensor-rich buildings. Simulation, 2014. 90(8): p. 960-977.
3. Agarwal, Y., B. Balaji, R. Gupta, J. Lyles, M. Wei, and T. Weng, Occupancy-driven energy management for smart building automation, in Proceedings of the 2nd ACM workshop on embedded sensing systems for energy-efficiency in building. 2010.
4. Chen, J. and C. Ahn, Assessing occupants’ energy load variation through existing wireless network infrastructure in commercial and educational buildings. Energy and Buildings, 2014. 82: p. 540-549.
5. Erickson, V., Y. Lin, A. Kamthe, R. Brahme, A. Surana, and A. Cerpa, Energy efficient building environment control strategies using real-time occupancy measurements, in Proceedings of the first ACM workshop on embedded sensing systems for energy-efficiency in buildings. 2009.
6. Zimmermann, G. Modeling and simulation of individual user behavior for building performance predictions. in Society for computer simulation international. 2007.
7. Klein, L., J.-y. Kwak, G. Kavulya, F. Jazizadeh, B. Becerik-Gerber, P. Varakantham, and M. Tambe, Coordinating occupant behavior for building energy and comfort management using multi-agent systems. Automation in Construction, 2012. 22: p. 525-536.
8. Lee, Y.S. and A.M. Malkawi, Simulating multiple occupant behaviors in buildings: An agent-based modeling approach. Energy and Buildings, 2014. 69: p. 407-416.
9. Azar, E. and C.C. Menassa. Impact of occupants behavior on building energy use: An agent-based modeling approach. in 10th International Conference on Modeling and Applied Simulation, MAS 2011, Held at the International Mediterranean and Latin American Modeling Multiconference, I3M 2011. 2011.
10. Langevin, J., J. Wen, and P.L. Gurian, Simulating the human-building interaction: Development and validation of an agent-based model of office occupant behaviors. Building and Environment, 2015. 88: p. 27-45.