The Systems Control Group at the University of Toronto has students working in both theoretical and experiment control. Currently there are three laboratories working on experimental control.
The topics of research illustrate the diversity of control theory. Experiments with positioning using magnetic levitation may result in new manufacturing techniques for semiconductor devices with high yield due to the absence of moving parts. The work being done with autonomous multivehicle systems could lead to flocks of intelligent robots exploring Mars. Techniques developed within the optical network control lab have applications in advanced high-performance communication networks.
The systems control group is interested in the application of magnetic levitation to develop high-precision positioning systems. A real time Matlab interface controlling a group of four permanent magnet linear synchronous motors (PMLSM) implements the non-linear control.
The current experiment performs nonlinear setpoint stabilization and sinusoidal tracking for a large levitating disk, or mover, with three degrees of freedom. The mover can be seen sitting on linear guides in the photo on the right in between the table and the four PMLSM's.More information on this research
The systems control group has 10 mobile robots to perform multivehicle experiments. Using Bluetooth for communication and an overhead camera to sense position data, the robot experiment provides a testbed to validate theories in time optimal control, path planning, multivehicle systems, and collision avoidance.
The robot control Java code is highly configurable and can currently implement closed loop control of up to four robots.
This testbed has been used to implement experiments in circulant pursuit and collision avoidance.More information on this research.
Optical Network Control
The systems control group is interested in the application of control and optimization theories to optical communication networks. The system control group has a number of experimental devices used in the optical industry that allowed us to develop a realistic testbed for implementing closed-loop control algorithms in multi-channel optical networks. A real time Labview and Matlab interface controls individually the power level of three optical transmitters, each emitting light of a specified wavelength. The optical wavelengths are then optically multiplexed and transmitted together through the same optical fiber, and then amplified by an optical amplifier and dynamic gain equalizer. The testbed is flexible and can be used to setup various single or two-link configurations, with different noise loadings.
The current experiment performs channel power balancing and optical signal-to-noise ratio equalization. The control loop operates based on real-time channel measurements provided by an optical spectrum analyzer interfaced to the control setup, as seen in the photo on the right.More information on this research.
Funding for the Systems Control Graduate Laboratory was provided by Canada Foundation for Innovation. Additional financial support was provided the Province of Ontario through the Ontario Innovation Trust. The Trust invests in research infrastructure at Ontario's universities, hospitals, colleges, and research institutes.