Zixiao Shi

Zixiao Shi
Shopify inc.

Doctor of Philosophy

About

32
Publications
8,592
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
301
Citations
Introduction
Fault detection, data imputation, metadata inference, sensor placement
Additional affiliations
February 2020 - present
Shopify
Position
  • Analyst
June 2019 - present
Carleton University
Position
  • Professor (Associate)
May 2018 - February 2020
National Research Council Canada
Position
  • Research Associate
Education
September 2014 - July 2018
Carleton University
Field of study
  • Civil Engineering
September 2011 - December 2012
Purdue University
Field of study
  • Civil Engineering
September 2007 - May 2011
Purdue University
Field of study
  • Civil Engineering

Publications

Publications (32)
Article
Faults in heating, ventilation, and air-conditioning control networks substantially affect energy and comfort performance in commercial buildings. As these control networks are comprised of many sensors and actuators, it is challenging to identify, often subtle, anomalies caused by these faults. In this paper, we develop a cluster analysis method f...
Article
Full-text available
This paper describes an open data set of 3,053 energy meters from 1,636 non-residential buildings with a range of two full years (2016 and 2017) at an hourly frequency (17,544 measurements per meter resulting in approximately 53.6 million measurements). These meters were collected from 19 sites across North America and Europe, with one or more mete...
Preprint
Full-text available
This paper describes an open data set of 3,053 energy meters from 1,636 non-residential buildings with a range of two full years (2016 and 2017) at an hourly frequency (17,544 measurements per meter resulting in approximately 53.6 million measurements). These meters are collected from 19 sites across North America and Europe, and they measure elect...
Preprint
This paper introduces a robust data preprocessing framework for time series building energy data. Once configured, the proposed framework will provide an automated pipeline to prepare, screen and correct energy consumption data for subsequent analysis. The framework adopts an ensemble data imputation approach which automatically selects the best im...
Article
Full-text available
Fault detection, diagnostics, and prognostics (FDD&P) ensure the operation efficiency and safety of engineering systems. In the building domain, they can help significantly reduce energy consumption and improve occupant comfort. Specifically, prognostics are becoming increasingly important as a pro-active fault prevention strategy through continuou...
Article
While understanding end-uses and their flow within a building is essential for energy management, many existing buildings still do not have adequate submetering for important end-uses such as chillers, fans, lighting and plug loads. To this end, this paper presents a new end-use disaggregation method for commercial buildings. Unlike previous disagg...
Article
Metadata in most existing building automation systems (BASs) is inconsistent, incomplete, and nondescriptive. This situation is a major obstacle to the widespread use of data analytics to improve the operation of buildings. In this article, we put forward a method to infer zone-level metadata from features derived from BAS data. The method includes...
Article
Zone-level sensor and actuator faults can substantially affect the energy and comfort performance of heating, ventilation, and air conditioning (HVAC) systems in commercial buildings. As these faults leave their fingerprints on a building automation system (BAS), algorithms which use features generated from BAS data streams as symptoms to detect an...
Conference Paper
The morning start time of heating and cooling equipment plays an important role in the energy and comfort performance of buildings. Existing algorithms to guide this decision require either many data types with a consistent labelling nomenclature or a detailed calibrated model. In this paper, a model-free clustering-based morning start time estimat...
Conference Paper
Full-text available
Metadata inference for building automation system (BAS) is an increasingly important topic to promote wider adoption of smart building technologies. Metadata inference is used to automatically discover semantics within the BAS, such as labelling sensors, discover control variable relationships, etc. Clustering analysis has been applied in many prev...
Article
Full-text available
Short-term forecasts of energy demand in buildings serve as key information for various operational schemes such as predictive control and demand response programs. Despite this, developing forecast models for heating and cooling loads has received little attention in the literature compared to models for electricity load. In this paper, we present...
Article
This article reviews the current research on the development and implementation of automated fault detection and diagnostics (AFDD) technology for building systems. This article first examines the fundamentals of AFDD and its special requirements on building systems to provide a theoretical formulation of the problem. Some infrastructural barriers...
Preprint
Full-text available
This paper details the methodologies for creating and training grey-box thermal models for low-rise residential buildings. This paper covers different aspects of model development such as model vectorization, cost function definition and parameter estimation. Different computing strategies such as GPU accelerated calculation are also explored. This...
Article
This article introduces a parameter estimation and state prediction technique called constrained dual Extended Kalman Filter (EKF) that can be used in model prediction controls (MPC) and fault detection and diagnostics (FDD) in building systems. The proposed method is an improvement over the existing nonlinear filter-based methods such as the joint...
Conference Paper
Full-text available
Optimization in building performance simulation (BPS) has become increasingly important due to the growing need for high-performance building design and operation. Numerous research efforts have been dedicated to decreasing optimization runtime by introducing improved optimization algorithms and advanced sampling techniques. This paper presents a n...
Article
Full-text available
This paper introduces a distributed system for building fault detection, diagnostic, and evaluation (FDDE). The design of the distributed system aims to address computation and network limitations on a common commercial building automation system (BAS). This system also aims to be adaptable to different fault detection and fault diagnostics algorit...
Conference Paper
Full-text available
Automated building fault detection and diagnostics (AFDD) has become a valuable tool to maintain the efficiency of high-performance buildings. Still, when faults are detected or diagnosed, they are often presented to the building operators as-is with little explanation of their potential impacts. This could lead to higher workload and knowledge req...
Article
This paper introduces a novel building energy model reduction pipeline called ‘model-cluster-reduce’. It is centred around using clustering techniques to identify archetypes and eliminate redundant zones. An experiment was conducted in this paper using a detailed EnergyPlus model generated from building information modelling directly. A total numbe...
Article
Full-text available
Modern commercial buildings' resource consumption is metered at various levels of spatial and temporal resolution to track and reduce energy use and greenhouse gas emissions. However, not all data that could be used to detect faults or identify efficiency improvements are available due to the cost of meters and inaccessibility of the data they prod...
Article
One of the crucial elements to ensure the efficiency of building operations is to understand the dynamics of energy flows, control strategies, and occupant behaviour in buildings. Currently, abundant sensors and sub-meters are installed in modern buildings to measure resource consumption at various levels of spatial and temporal resolution to help...
Conference Paper
Full-text available
This paper introduces a building energy model reduction method by using one exemplar zone to represent a group of similar thermal zones. The procedure involves using principal component analysis to model and capture the thermal behaviors of the zones, and then use affinity propagation clustering technique to group similar zones together and identif...
Conference Paper
Full-text available
This project created a computer information visualization tool designed for displaying building operation data. The design goal is to improve the user's efficiency of monitoring building performance with the available building information. This tool was formalized based on the existing computer information visualization theories. This tool is capab...
Conference Paper
Full-text available
Fault detection and diagnostics (FDD) is important to maintain proper operation of the building systems. However, most research is focused on heating, ventilation and air conditioning (HVAC) systems FDD, while little work has been done on zone level diagnostics. A building zone is a complex system, but usually equipped with fewer sensors, making th...
Article
Two control-oriented models that can predict the temperature of a perimeter office space were developed by using the data gathered from light intensity, motion and temperature sensors, and terminal heating and cooling units. One model had five unknown parameters while the second had ten unknown parameters and an immeasurable state. The models’ para...
Technical Report
Full-text available
Although modern buildings are using increasingly sophisticated energy management and control systems that have tremendous control and monitoring capabilities, building systems routinely fail to perform as designed. More advanced building control, operation, and automated fault detection and diagnosis (AFDD) technologies are needed to achieve the go...
Conference Paper
Full-text available
Building Information Modelling (BIM) has emerged as a powerful technology that creates a central hub for managing building energy and resources at all phases of the building life cycle. Without it, many tools that lack interoperability are used, thus massively under-exploiting the efforts of other building design and management parties; this largel...

Network

Cited By

Projects

Projects (5)
Archived project
Funded by NIST 10D243 grant, this project aims at developing tools that could provide simulation capabilities to develop and evaluate advanced control, operation, and AFDD technologies for these less studied secondary systems. In this study, HVACSIM+ is selected as the simulation environment. Simulation testbeds for secondary systems, such as fan coil units, fan power units, and dual duct systems, are developed which can simulate system operation when the system contains faults and when fault free. The testbeds are validated using real building data under both fault free conditions and when artificial faults are implemented in the building.
Project
Many buildings have data points from legacy DDC are not properly labelled, or do not use a standard nomenclature. This project aims to use the available textual and numeric data from these points to infer their actual meta information.
Project
Most time series energy data in buildings, plants and utilieis are noisy. This tool aims to provide scalable solution to automatically perform anomaly detection, model selection and value estimation for time series energy data. It can be used as a off-line or online solution.