Analysis and Modeling of Piezoelectric Transformers This thesis presents an equivalent circuit model for a Rosen type piezoelectric transformer. Transmission line equations for various vibration modes with a clamped boundary condition on one end of each section are developed. The partial expansion theorem is used to convert the transmission line model into an equivalent circuit model. Losses are introduced in a post-processing step using the experimental value of mechanical quality factor. L'Hospitals rule is used to simplify the circuit near a desired resonance frequency. Experimental results confirm the existence of multimode resonance and anti - resonance behaviour. The modeling methodology is then extended to multi-layered rectangular and circular geometries. The applicability of two DC-DC converter topologies in connection with a piezoelectric transformer i.e. Class-E and an Asymmetrical converter are discussed. Also, experiments on a three winding multi - layered device connected to an asymmetrical converter are performed. Results indicate that the third winding cannot be used to provide a feedback voltage since the voltage does not track the secondary output voltage.
A Contribution to the understading of a PEM fuel cell transient model The purpose of this thesis is to build a foundation from which a transient model of a proton exchange membrane (PEM) fuel cell (FC) can be constructed. The FC model was developed taking into consideration basic physical and mathematical laws that describe the electrochemical mechanisms that govern performance of a proton exchange membrane fuel cell (PEMFC). Our modeling approach considers the charge distribution in the whole FC (including the electrodes). This information is required in one is create a realistic transient model of a FC. This thesis takes into account the presence to two mobile charged species in the (PEM). These charges are responsible for shaping the variable charge distribution profile across the double layers and the membrane. Additionally, due to the very well known impossibility to measure the absolute potential (or inner potential) of a half-cell, the need for a reformulation of the reference voltage is presented, and constitutes one of the major findings of this thesis.
Stand-alone equipment such as relays, meters and static
power converters utilize the fundamental sequence components of power
system signals. These applications require an initial data acquisition
stage followed by signal processing stages whose properties are application
dependent. This thesis proposes an integrated signal acquisition and filtering
scheme based on the "system on a chip" concept. Various filtering techniques
are investigated and several novel filtering schemes suitable for power
applications are presented to achieve the objective. A fast extraction
of sequence components algorithm is presented. The variable sampling multirate
system realizes frequency tracking, disturbance/fault detecting, delayless
filtering with good noise attenuation and with fast response, respectively.
The thesis also explores hardware implementation issues. The prototyped
models in VHDL are simulated with the MAX+PLUS II platform and tested
in-circuit. Experimental results are in conformity with the simulations.
This thesis presents a digital signal processing system for synchronization
to AC utility network signals in power electronics applications. The
proposed system is suitable in applications where the signal to which
synchronization must be accomplished contains severe disturbances and
where the signal frequency and the signal amplitude are variable. The
system is based on a multirate phase-locked loop (PLL). The main advantages
of the multirate approach are that it relaxes the implementation of
the antialiasing filter and it enables one to accommodate the varying
amplitude of the input signals. The antialiasing filter, which is in
this case a high order bandpass filter, is implemented in the digital
part of the system. This feature is achieved by applying the oversampling
technique to the input signal. The antialiasing filter automatically
adapts to the input signal frequency through the system's variable sample
rate operation. The thesis deals with modeling, analysis, control and
implementation issues of the proposed system. Nonlinear and linear state-space
models and a transfer function model of the system are derived. The
system's control aspects are discussed using the derived transfer function
model. The system is implemented on a platform based on the Texas Instruments
TMS320C31 floating point digital signal processor. Tracking properties
of the implemented system are verified with realistic signals such as
a sinusoidal signal containing notch type disturbance and noise and
an arc furnace voltage signal.
For my thesis, I am going to do some researches on the
power supply of microprocessors. The power supply has to deliver a high
current at a low voltage requiring a very high slew rate in current change
and a high voltage regulation.