Prospective graduate students are invited to contact me if they are interested in any of the projects below.
Dynamics and Control of a Multi-Tethered Aerostat System
Astronomers at the National Research Council are interested in using a multi-tethered aerostat to support the receiver of a very large radio telescope. The system consists of a Helium balloon connected by a tether to a payload (the telescope receiver) below it. The payload is, in turn, attached to the ground by multiple tethers (3 to 6). At the base of each tether, a winch is used to reel cable in and out, thereby allowing precise positioning of the payload.
We are studying the dynamics and control of this sytem using simulations developed in our laboratory. Among the questions we are investigating: how precisely can the receiver be positioned in the presence of turbulent winds ? What kind of controllers should be used to adjust the winches ? Would other means of actuation (e.g., a winch between the payload and the balloon, or an articulated mechanism for the receiver) substantially improve the positioning capability?
We are also deploying a one-third scale prototype of the multi-tethered aerostat system in order to validate our simulations and as a proof-of-concept for the idea.
Trajectory Planning for Autonomous Underwater Vehicles
As part of a larger project based at Memorial University in Newfoundland, we are studying the dynamics and control of an Autonomous Underwater Vehicle (AUV) named C-SCOUT. So far, we have investigated, both in simulation and experimentally, the characteristics of the through-body thrusters which allow the vehicle to be maneuvered at low speeds. We are now interested in incorporating these results into an existing simulation of the C-SCOUT.
The simulation would then be used to plan the trajectory of the AUV through the exhaust plume of an offshore platform, to allow an onboard mass spectrometer to collect samples of the produced water. To properly plan these trajectories, the characteristics of the vehicle and the mass spectrometer must be considered.
Over the past few years, we have developed extensive cable models for use in the modelling of underwater systems (ROVs, towed vehicles) and airborne systems. Our models are based on the lumped mass approximation, and we have extended these, for example, to the modelling of slack cables.
Further work will include: (a) an investigation (experimental and numerical) of internal damping characteristics, (b) modelling of varying-length cables, (c) modelling of cables in contact with their environment, and (d) improved numerical integration techniques.
Distance Determination Algorithms
Working with MD Robotics in Toronto, we have developed a series of distance determination algorithms. These have been used by MDR as part of their simulations of space robotics systems. The algorithms calculate separation/interference distances between objects in a scene, and allow the calculation of contact forces. The main requirements for these algorithms are that they be fast, accurate, and robust.
Recently, we have developed algorithms which calculate distances between concave objects, without partitioning. Future work will involve coupling these algorithms with our earlier convex algorithms so allow them to work together seamlessly. As well, we are interested in developing pruning algorithms to reduce the burden of the distance calculations to only those objects which are likely to come into contact.
Design and Control of a Biomimetic Underwater Vehicle
Most amphibious animals propel themselves through the water using a paddling motion. This is motivated by the fact that webbed feet, or paddles can be used on land or in the water. With this inspiration, the Ambulatory Robotics Laboratory has developed a robotic vehicle that propels itself using a paddling motion. We are interested in developing control algorithms for (a) operator interface, (b) stability augmentation, and (c) autopilot functionality for this unusual vehicle.
Dynamics and Control of Modern Airships
The golden age of the large airship ended over 60 years ago. Since then, there have been important advances in modern materials and computer control methods which may allow a resugence of these impressive machines. In particular, we expect that new advancements will permit the development of unmanned airships for surveillance or communications relay. We are interested in the dynamics modelling of airships, and the application of modern control methods to develop intelligent autonomous vehicles capable of sustained long duration flight.
Control of Unmanned Aerial Vehicles
Building upon rapid advances in robotics, control, communications and computer technology, autonomous intelligent vehicles are being called upon to play an ever-increasing role in civilian and military environments. Uninhabited Air Vehicles (UAVs) may assume a variety of roles, such as surveillance of coastal areas, monitoring of icebergs or pipelines, and communications relay. Jointly with researchers at Defense Research and Development Canada (Valcartier), we are investigating the following topics: (a) Real-time reconfigurable control for autonomous UAV flight, (b) Multi-objective guidance and control of autonomous helicopters, and (c) Cooperative path planning and trajectory generation for multi-UAV systems.