Grant: Discovery Launch Supplement
University of Calgary
$12,500.00
Nonlinear Network System Identification with Applications in Monitoring
FY 2024-25
Summary
The long term objective of my research program is to develop a data driven monitoring approach that is able to track changes in the physics of each component of a machine or industrial plant. Consider the following sketch that highlights the challenges for a data driven monitoring approach. A gearbox is attached to a motor in a robotic arm and we wish to detect wear of the teeth of the gear box. Sensors do not directly measure wear on the teeth of a gear box, but, a worn out gear box will affect the vibrations and position of a motor which can be measured using accelerometers and encoders. The motor is subjected to external vibrations caused by heavy machinery around the motor, and other motors on the robotic arm; the motor is operating at highly variable speeds and accelerations, which cause highly variable vibrations; and the temperature of the robot is constantly changing, which affects the accuracy of the sensors. In short, the data is dominated by changes caused by the ambient environment and the properties of the sensors, not the change that we wish to detect. To further complicate matters, a gearbox is governed by nonlinear dynamics due to the presence of hysteresis. Currently there is no data driven monitoring approach that would be able to accurately track changes in the physics of the gearbox under these conditions. In my research program I am proposing to develop a nonlinear network system identification approach to tackle this challenge.
In this approach complex systems are modelled as interconnected components. Under certain conditions, the dynamics of the components can be estimated irrespective of the ambient noise affecting the system, or sensor noise. The novelty of this approach is that I am combining of linear network system identification and nonlinear system identification.
The short term objectives of the program are designed to gradually transition linear network identification results to a network consisting of both linear and nonlinear components. The short term objectives are (1) develop methods where linear network identification tools can be used to estimate a linear transfer function embedded in a network that contains both linear and nonlinear components, (2) determine variance expressions for these estimated transfer functions, (3) develop nonlinear system identification tools to estimate any component embedded in the network.
The impact of my research is better fault detection, which will mean earlier detection of faults or wear. This will have a large impact in the case of infrastructure or large industrial plants in terms of implementing a predictive/preventative maintenance program; and increasing efficiency, safety, and throughput of a plant. I plan to disseminate results to industry by outreach and by releasing a software package implementing the algorithms of my research.
Program
Discovery Launch Supplement
Recipient
Location
Alberta, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
DGECR-2024-00136
Name nom
Dankers, Arne
Department département
Electrical and Software Engineering
Organization id
11
Institution établissement
University of Calgary
Province en
Alberta
Province fr
Alberta
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
12500
Program id
DGECR
Program name en
Discovery Launch Supplement
Program name fr
Supplément Tremplin vers la découverte
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1510
Committee name en
Electrical and Computer Engineering
Committee name fr
Génie électrique et informatique
Area of application code
601
Area of application group en
Construction, urban and rural planning
Area of application group fr
Construction, urbanisme et aménagement rural
Area of application en
Structural engineering
Area of application fr
Génie des charpentes
Research subject code
2509
Research subject group en
Electrical and electronic engineering
Research subject group fr
Génie électrique et électronique
Research subject en
Control systems
Research subject fr
Asservissements
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Nonlinear Network System Identification with Applications in Monitoring
Keywords
system identification; data driven modelling; distributed control systems; model based monitoring; machine learning
Application summary
The long term objective of my research program is to develop a data driven monitoring approach that is able to track changes in the physics of each component of a machine or industrial plant. Consider the following sketch that highlights the challenges for a data driven monitoring approach. A gearbox is attached to a motor in a robotic arm and we wish to detect wear of the teeth of the gear box. Sensors do not directly measure wear on the teeth of a gear box, but, a worn out gear box will affect the vibrations and position of a motor which can be measured using accelerometers and encoders. The motor is subjected to external vibrations caused by heavy machinery around the motor, and other motors on the robotic arm; the motor is operating at highly variable speeds and accelerations, which cause highly variable vibrations; and the temperature of the robot is constantly changing, which affects the accuracy of the sensors. In short, the data is dominated by changes caused by the ambient environment and the properties of the sensors, not the change that we wish to detect. To further complicate matters, a gearbox is governed by nonlinear dynamics due to the presence of hysteresis. Currently there is no data driven monitoring approach that would be able to accurately track changes in the physics of the gearbox under these conditions. In my research program I am proposing to develop a nonlinear network system identification approach to tackle this challenge.
In this approach complex systems are modelled as interconnected components. Under certain conditions, the dynamics of the components can be estimated irrespective of the ambient noise affecting the system, or sensor noise. The novelty of this approach is that I am combining of linear network system identification and nonlinear system identification.
The short term objectives of the program are designed to gradually transition linear network identification results to a network consisting of both linear and nonlinear components. The short term objectives are (1) develop methods where linear network identification tools can be used to estimate a linear transfer function embedded in a network that contains both linear and nonlinear components, (2) determine variance expressions for these estimated transfer functions, (3) develop nonlinear system identification tools to estimate any component embedded in the network.
The impact of my research is better fault detection, which will mean earlier detection of faults or wear. This will have a large impact in the case of infrastructure or large industrial plants in terms of implementing a predictive/preventative maintenance program; and increasing efficiency, safety, and throughput of a plant. I plan to disseminate results to industry by outreach and by releasing a software package implementing the algorithms of my research.