PhD Oral Exam - Soroush Samareh Abolhassani, Information and Systems Engineering

When studying for a doctoral degree (PhD), candidates submit a thesis that provides a critical review of the current state of knowledge of the thesis subject as well as the student’s own contributions to the subject. The distinguishing criterion of doctoral graduate research is a significant and original contribution to knowledge.

Once accepted, the candidate presents the thesis orally. This oral exam is open to the public.

Abstract

One of the main required actions for moving forward in zero-carbon cities’ goals is to increase building energy efficiency. Urban Building Energy Modeling (UBEM) can help optimize the built environment's energy efficiency and improve the design and operation of building energy systems. Moreover, building modeling provides key information for HVAC control, building energy management, different building energy efficiency scenarios, and feedback to the occupants. In order to parametrize Urban Building Energy Modeling (UBEM), individual buildings' characteristics, such as constructions, internal loads, energy systems, etc., are required. Many dynamic UBEMs have already been developed, yet most use simplified assumptions for some of the parameters needed. The capabilities required for having a comprehensive UBEM are creating detailed 3D urban building geometry, creating a comprehensive building attributes library, conducting detailed archetype selection, and assignment of this construction information to the building surfaces. More importantly, occupancy behavior is another important factor that influences the modeling results, and it is the root of high uncertainty in UBEM results. Most existing tools have a significant weakness in dealing with occupant related monitoring data or estimation of likely schedules.

Many studies have been conducted to develop and introduce different occupancy behavior measurements and methods for urban building energy modeling. However, none of the previous studies introduced scalable and practical occupancy behavior monitoring on an urban scale. Some of the main challenges that prevent the scalability of the already developed occupancy behavior methods and measurements are the fact that they suffer from threatening the occupant's privacy and the high cost of their infrastructure and deployment.

This thesis aims to leverage passive WiFi sensing methods for occupant behavior monitoring and pattern analysis for UBEM using commercial off-the-shelf WiFi devices, for which the infrastructure is already available in many buildings. Passive and device-free WiFi sensing does not threaten the occupant's privacy, and its deployment cost is low. Therefore, this method of occupancy measurement has the potential to be used on an urban scale and to play an essential role to manage energy in zero carbon and smart cities in the future.

In order to pave the way for the application of WiFi sensing technology in occupancy monitoring and behavior estimation for UBEM, the following approaches are proposed:

A comprehensive UBEM workflow is designed, which can create the 3D geometry of buildings and take advantage of comprehensive building characteristics and attributes already developed in the Tool4Cities urban simulation platform, designed and developed at the Next Generation Cities Institute, Concordia University, Canada. Tools4Cities aims to model cities in a holistic approach, including buildings, transportation, energy systems, and networks, as well as waste management. Its workflows automatically perform the archetype selection and assignment to the building surfaces and internal loads assignment to the buildings and their floors.

The developed UBEM workflow is integrated as an extension into the Tool4Citites platform using a step-by-step comprehensive framework.

A comprehensive occupancy and behavior monitoring approach is designed, using data derived from WiFi sensing technology, to estimate the occupancy behavior. Case studies were done in a residential building and a large institutional building case study located in Montreal, Canada. The developed workflows apply different statistical and machine learning algorithms to the collected, denoised, and preprocessed Channel State Information (CSI) data to estimate occupants’ behavior and number. Moreover, five schedules of occupancy, occupant activity, lighting, electrical equipment usage, and thermostat set point temperature are extracted and estimated from CSI data.

In order to show the capability of the developed workflows, they are applied in two case studies of residential and institutional buildings, and the impact of measured occupancy data on the outcomes of building energy modeling is analyzed. Moreover, the UBEM performance is analyzed using the developed occupancy monitoring method. To this end, all estimated schedules are fed into the Tool4Citites platform, along with other required data for accurate urban building energy demand calculation to estimate heating and cooling demand.

This thesis addresses the UBEM limitations in occupancy monitoring and pattern analysis by leveraging the data derived from WiFi sensing technology. This technology has fewer privacy issues, low cost of devices, installation, and maintenance, and its infrastructure is already available in many buildings, which were the main barriers to occupancy monitoring on the urban scale. It is expected that this emerging technology will reshape the future of urban building energy modeling, and energy management and accelerates carbon emission reduction in the future by using it in more and more buildings.