Concordia University

https://www.concordia.ca/content/shared/en/events/offices/vprgs/sgs/2020/04/29/phd-oral-exam-seyed-amirhosain-sharif-arani-building-engineering.html

LATEST INFORMATION ABOUT COVID-19

READ MORE

Thesis defences

PhD Oral Exam - Seyed Amirhosain Sharif Arani, Building Engineering

Optimizing Energy Performance of Building Renovation Using Traditional and Machine Learning Approaches

Date & time

Wednesday, April 29, 2020
1 p.m. – 4 p.m.

Cost

This event is free

Organization

School of Graduate Studies

Contact

Jennifer Sachs

Where

Online

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

International Energy Agency (IEA) studies show that buildings are responsible for more than 30% of the total energy consumption and an equally large amount of related greenhouse gas emissions. Improving the energy performance of buildings is a critical element of building energy conservation. Furthermore, renovating existing buildings envelopes and systems offers significant opportunities for reducing Life-Cycle cost (LCC) and minimizing negative environmental impacts. This approach can be considered as one of the key strategies for achieving sustainable development goals at a relatively low cost, especially when compared with the demolition and reconstruction of new buildings. One of the main methodological and technical issues of this approach is selecting a desirable renovation strategy among a wide range of available options.

The main motivation behind this research relies on trying to bridge the gap between building simulation, optimization algorithms, and Artificial Intelligence (AI) techniques, to take full advantage of the value of their couplings. Furthermore, for a whole building simulation and optimization, current simulation-based optimization models, often need thousands of simulation evaluations. Therefore, the optimization becomes unfeasible because of the computation time and complexity of the dependent parameters. To this end, one feasible technique to solve this problem is to implement surrogate models to computationally imitate expensive real building simulation models.

The aim of this research is three-fold: (1) to propose a Simulation-Based Multi-Objective Optimization (SBMO) model for optimizing the selection of renovation scenarios for existing buildings by minimizing Total Energy Consumption (TEC), LCC and negative environmental impacts considering Life-Cycle Assessment (LCA); (2) to develop surrogate ANNs for selecting near-optimal building energy renovation methods; and (3) to develop generative deep MLMs to generate renovation scenarios considering TEC and LCC. This study considers three main areas of building renovation, which are the building envelope, Heating, Ventilation and Air-Conditioning (HVAC) system, and lighting system; each of which has a noteworthy influence on building energy performance.

On this premise, this research initially develops a framework for data collection and preparation to define the renovation strategies and proposes a comprehensive database including different renovation methods. Using this database, different renovation scenarios can be compared to find the near-optimal scenario based on the renovation strategy. Each scenario is created from the combination of several methods within the applicable strategy. The SBMO model simulates the process of renovating buildings by using the renovation data in energy analysis software to analyze TEC, LCC, and LCA and identifies the near-optimal renovation scenarios based on the selected renovation methods. Furthermore, an LCA tool is used to evaluate the environmental sustainability of the final decision.

It is found that, although the proposed SBMO is accurate, the process of simulation is time consuming. To this end, the second objective focuses on developing robust Machine Learning Models (MLMs) to explore vast and complex data generated from the SBMO model and develop a surrogate building energy model to predict TEC, LCC, and LCA for all building renovation scenarios. The main advantage of these MLMs is improving the computing time while achieving acceptable accuracy. More specifically, the second developed model integrates the optimization power of SBMO with the modeling capability of Artificial Neural Networks (ANNs). While, the proposed ANNs are found to provide satisfactory approximation to the SBMO model in a very short period of time, they do not have the capability to generate renovation scenarios.

Finally, the third objective focuses on developing a generative deep learning building energy model using Variational Autoencoders (VAEs). The proposed semi-supervised VAEs extract deep features from a whole building renovation dataset and generate renovation scenarios considering TEC and LCC of existing institutional buildings. The proposed model also has the generalization ability due to its potential to reuse the dataset from a specific case in similar situations.

The proposed models will potentially offer new venues in two directions: (1) to predict TEC, LCC, and LCA for different renovation scenarios, and select the near-optimal scenario, and (2) to generate renovation scenarios considering TEC and LCC. Architects and engineers can see the effects of different materials, HVAC systems, etc., on the energy consumption, and make necessary changes to increase the energy performance of the building. The proposed models encourage the implementation of sustainable materials and components to decrease negative environmental impacts. The ultimate impact of the practical implementation of this research is significant savings in buildings’ energy consumption and having more environmentally friendly buildings within the predefined renovation budget.

Back to top Back to top

© Concordia University