Student projects99

Document Sample
Student projects99 Powered By Docstoc
					         Student Projects by Dr. Shrivastava for Semester 2, 2006
[YS1] Neural network for speed estimation and control of a DC motor
(Software Project)(Number of Students: 2)

Artificial neural networks (ANN) are becoming increasingly popular for control applications due to
their ease in training. They can be trained to approximate any linear or nonlinear function. This
project requires implementing an ANN (using Neural Network Toolbox in MATLAB) to estimate
the speed of a DC motor and using the speed estimator in turn to control the speed of the DC motor.
The algorithm can be designed, simulated and tested in MATLAB/Simulink environment.
Successful speed estimator design can then be tested on the real time data obtained from the actual
voltage and current measurements (using Data Acquisition Board) from the DC motor driven by a
DC chopper.

[YS2] Design of web pages using MATLAB for teaching signals
(Software Project)(Number of Students: 4)

The project requires implementing web pages to illustrate various concepts in ELEC 2302: Signals
& Systems and ELEC 3305: Digital Signal Processing. The web pages would be generated by Web
Server Toolbox for MATLAB. This toolbox provides an interface for MATLAB and Simulink
applications to get the data from HTML forms and generates dynamic HTML pages to display your
results including sophisticated graphics produced by MATLAB programs. The examples can be
negotiated based on the students’ taste.

[YS3] Control the speed of Induction motor using sliding mode control
(Software Project)(Number of Students: 2)

The induction motor is becoming more popular in industrial applications. Its simplest model is a
fifth order nonlinear system. Many different speed control algorithms for the induction motor can
be found in the literature. All of these manipulate the voltage or current applied to the motor to
achieve the speed control. These schemes are implemented by using power electronics to switch a
DC supply appropriately. Sliding mode control is a nonlinear control technique that involves
switching control signal to drive the system on a sliding trajectory to a desired equilibrium point.
This feature makes it particularly attractive for the induction motor speed control. The project
involves designing, simulating and testing the sliding mode control for induction motor in
MATLAB. The long-term view is to implement such a scheme on a DSP based system.

[YS4] Harmonic elimination for 2-level inverter having a ripple in the DC link
(Hardware Project)(Number of Students: 2)

DC to AC power conversion in many applications is performed by a 2-level inverter. The pulse
width modulated (PWM) output of the inverter is filtered to remove high frequency components.
Harmonic elimination is one of the techniques to reduce the constraints on the filter design and is
especially useful for low switching frequencies. In the computation of harmonic elimination scheme
one normally assumes that the DC link has no ripple on it. This can be practically achieved by using
a large capacitor to filter the rectifier voltage. However for high power applications the use of a
large capacitor increases the cost substantially. In this project we will trade off the cost of the
capacitor by allowing a larger ripple in the DC link and explicitly account for this ripple in the
design of harmonic elimination scheme. The students are required to design such a harmonic
elimination scheme and implement it using a Texas Instruments DSP (TMS320C240).

[YS5] DC-DC power converter – sliding mode control
(Hardware / Software Project)(Number of Students: 2)

The switch mode power supply (SMPS) is used in all electronic equipment, which operates from
the DC supply obtained by rectifying the AC voltage filtered by a capacitor. The design of the
power supply assumes a constant DC input, which can be practically achieved by using a large
capacitor to filter the rectifier voltage. However for high power applications the use of a large
capacitor increases the cost substantially.
In this project we will trade off the cost of the capacitor by allowing a larger ripple in the DC link
and with the use of sliding mode closed loop control, and achieve a good regulation of output

The project involves simulation of the proposed sliding mode control scheme using MATLAB and
then implementing it on a DSP based system.

[YS6] Implementing Linear prediction-based vocoders on a DSP board
(Software Project)(Number of Students: 2)

Vocoders are commonly used devices in telecommunication for analyzing and synthesizing speech.
Their figures of merit include bit rate and the resultant speech quality. Linear Predictive Coding
(LPC) is a popular model for speech synthesis. LPC has also been used in speech compression and
reasonable quality LPC based vocoders have been designed with bit rates as low as 2400 bit/s. The
project involves implementing and testing LPC based vocoders on the Texas Instruments DSP
boards (presently being used in ELEC 3305: Digital Signal Processing).

[YS7] Model of Spinal Cord Reflexes
(Software Project)(Number of Students: 2)(Requires credit average and appropriate background)
(Supervisors: P. M. Nickolls and Y. Shrivastava)

There are sensors in the muscles that detect their length and rate of change of their length. These
have a positive feedback effect on the nerves in the spinal cord that causes these same muscles to
contract. Sensors in the tendons of muscles detect the force or tension induced by muscle
contraction. These have a negative feedback effect on the muscle nerves. Other connections to the
muscle nerves in the spinal cord come from sensors in the skin and also from the brain. All of these
connections give rise to automatic movements called reflexes. These spinal cord circuits provide
much of the control of walking, with the brain having some control over the frequency of the
walking cycle. After spinal cord injury, as well as paralysis and loss of sensation, patients develop
abnormal reflexes that interfere with any walking ability they may retain. The task is to model these
circuits and also the changes with spinal cord injuries. A long-term aim of this type of work is to
determine sites of electrical stimulation that might be used to restore normal reflexes and walking
ability. The project would build upon existing models written in C++ and Delphi Pascal and be run
in MATLAB, C++, Java or Pascal.
[YS8] Design and Construction of a Portable DC-DC Inverter to Power a Muscle Stimulator
(Hardware / Software Project)(Number of Students: 2) (Requires credit average and appropriate
(Supervisors: P. M. Nickolls and Y. Shrivastava)

To restore walking in people paralyzed after spinal cord injuries it is necessary to stimulate their
muscles using 8 high voltage stimulators. These devices have been designed and are undergoing
testing. A portable power supply that can be worn on a belt is required. The power supply would
itself be powered by a small 12V gel battery and would have a DC-DC inverter to deliver the
required +/- 130 V supplies to the stimulators which would draw an average power of 80W. The
people involved in this project would need a good knowledge and enthusiasm for power electronics
and would need to grasp the complexities of the magnetic circuits that are peculiar to the inductors
and transformers used in these types of inverters.

[YS9] Labview driver for a GPIB controlled Capacitance Meter
(Hardware / Software Project)(Number of Students: 1) (Requires appropriate background)
(Supervisors: D. Debuf and Y. Shrivastava)

Labview provides a user-friendly interface for the control and visualization of instrumentation and
data respectively. The Labview environment has features that include calling matlab routines for
data manipulation and an SDK (Software Development Kit) for the development of low-level
drivers for instrumentation. This project entails writing a low level driver for a Capacitance Meter
under GPIB control using the SDK. The effectiveness of the driver is to be tested by measuring the
capacitance versus time of a diode under pulsed forward and reverse bias conditions. This may also
include writing some Labview code to control a pulse generator and a sampling card. The pulse
generator and sampling card drivers may be supplied. This procedure forms the basis of a method to
detect electrically active defects in semiconductors, which are typically detrimental to the
performance of semiconductor devices.

Shared By:
Description: Student projects99