|
|
|
|
|
|
|
Victor
M. Becerra |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|||||||||||||||
|
|
Student projectsBrief information and videos about some of the undergraduate student projects that I have supervised can be found below. Magnetic levitation ballThis project (2001-2002), carried out by Phillip Elliot and Daniel Faulkner, consisted of the design of a magnetic levitation system. Watch videos of the maglev system under real time fuzzy control: Video 1 (AVI format, 30mb); Video 2 (AVI format, 30mb). Computed torque control of a PUMA 560 manipulator robotThis project (2002-2003),
carried out by Callum Cage,
consisted of the retrofit and PC interface of a PUMA 560 manipulator
robot. This type of manipulator has six degrees of freedom. The robot
is now controlled in real time using modern Matlab/Simulink based
software from a PC. An AVI format video
shows the robot under computed torque control, a model based technique
that employs feedback linearisation to improve closed loop performance
over the operating range of the robot.
Magnetic levitation platformThis project has been carried out by four generations of part 4 MEng Cybernetics students (P. Elliot, C. Sulzbacher, R. Haines, J. O'Hea) between 2002 and 2006. The project consisted of the design, construction and control of a magnetic levitation vehicle. The vehicle consisted of a 40kg cart equipped with four electromagnets for levitation and four for lateral guidance. The cart is attracted by magnetic force to two rails located above the vehicle. Two further rails are used for guidance. Spring based contact sensors were used to measure the gap between the magnets and the rails. The vehicle was controlled in real time from Matlab/Simulink and the magnets driven by PWM circuits. Successful levitation was achieved using a number of methods, including LQR control. The picture below illustrates the vehicle construction. An AVI format video shows the cart being lifted by the magnets. A further video shows the platform during levitation under LQR control.
Magnetic bearings deviceThis project, carried out by Daniel
Aldridge, a part 4 MEng
Cybernetics student in 2005-2006, consisted of the design, construction
and control of a magnetic bearings device. The device consisted of a
physical construction which included a
motor coupled to a shaft on one end, magnetic proximity sensors and
four electromagnets arranged in pairs in the vertical and horizontal
axes at the other end of the shaft. In addition, the required
electronic circuit to drive the electromagnets was designed and
implemented by the student. Control was
performed from a PC equipped with Matlab/Simulink software and a data
acquisition board. The control was
implemented using linear state feedback designed with the LQR
(linear-quadratic-control) technique, where the electromagnets are
coupled along the same axis. This
is a challenging control problem due to its inherent nonlinearities,
instability and fast time constants.
A video of the magnetic bearings device in
action can be found here.
|
|
|||||||||||||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|