Thesis (MScEng (Electrical and Electronic Engineering))--University of Stellenbosch, 2006. / This thesis is about the dynamic control of a permanent magnet assisted reluctance synchronous machine (PMA RSM). The PMA RSM in this thesis is a 110 kW traction machine and is ideal for the use in electrical rail vehicles. An application of the dynamic control of the PMA RSM in electrical rail vehicles is to reduce wheel slip. The mathematical model of the PMA RSM is derived and explained in physical terms. Two methods of current control for the PMA RSM are investigated, namely constant field current control (CFCC) and constant current angle control (CCAC). It is shown that CCAC is more appropriate for the PMA RSM. A current controller for the PMA RSM that guarantees stability is derived and given as an analytic formula. This current controller can be used for any method of current control, i.e. CFCC or CCAC. An accurate simulation model for the PMA RSM is obtained using results from finite element analysis (FEA). The accurate model is used in a simulation to verify CCAC. A normal proportional integral speed controller for the PMA RSM is designed and the design is also verified by simulation. Practical implementation of the current and speed controllers is considered along with a general description of the entire drive system. The operation of the resolver (for position measurement) is given in detail. Important safety measures and the design of the electronic circuitry to give protection are shown. Practical results concerning current and speed control are then shown. To improve the dynamic performance of the PMA RSM, a load torque observer with compensation current feedback is investigated. Two observer structures are considered, namely the reduced state observer and the full state observer. The derivation of the full state observer and the detail designs of the observer elements are given. The accurate simulation model of the PMA RSM is used to verify the operation of the observer structures and to evaluate the dynamic performance. Both observer structures are implemented practically and practical results are shown. One method of position sensorless control, namely the high frequency voltage injection method, is discussed in terms of the PMA RSM. This work is additional to the thesis but it is shown, because it raises some interesting questions regarding the dynamic control of the PMA RSM.

Thesis (MScEng (Electrical and Electronic Engineering))--Stellenbosch University, 2008. / The topic for this project entailed the development of a ‘Low Phase Noise – Microwave – Voltage Controlled Oscillator’ for use in radar applications. First of all, a low phase noise oscillator was designed. In order to minimise the phase noise of the oscillator, a high-Q, transmission line – cavity resonator was developed. By derivation it was confirmed that an optimal point for minimum phase noise does exist. The latter was done by evaluating the equation for the output power spectral density of the oscillator phase noise (as defined by Leeson’s Phase Noise Model) at its minimum point. Subsequently, the amount of power that needed to be dissipated inside the resonator could be compared to that dissipated in the source and the load. This identified the amount of coupling to the resonator allowed, assuring minimum phase noise. Since a specific amount of coupling to the resonator was sought after, it had to be practically feasible. Therefore several coupling techniques were investigated to ensure the most user-friendly way of tuning the amount of coupling to the resonator, and hence easily reaching the optimum point of minimum phase noise. After successful completion of the low phase noise oscillator design, it was modified for voltage controlled oscillator (VCO) use by means of variable tuning diodes. These varactor diodes were situated inside the cavity of the resonator. Again the most suitable position to place the diodes had to be determined. The latter was done through considerably detailed transmission line theory; where the loaded Q, the tuning bandwidth (amount of change in frequency reached) and the amount of power dissipated inside the resonator were measured against each other. By means of the necessary phase noise measurements, it was confirmed that in order to keep the phase noise to a minimum, the tuning bandwidth had to be kept small and the amount of power dissipated inside the resonator maximised; so as to keep the overall loaded Q-value of the circuit as high as possible.

Thesis (MScEng (Electrical and Electronic Engineering))--University of Stellenbosch, 2005. / The aim of the thesis is to design a Control Moment Gyro (CMG) actuator which can be used in small satellite applications. The hardware and software of the CMG has to be designed according to specifications given. A satellite fitted with these CMGs has to be able to do a 30 degree rotation within 10 seconds.

Thesis (MScEng (Electrical and Electronic Engineering))--University of Stellenbosch, 2007. / Clear procedures for designing dielectric resonator oscillators (DROs) are presented in this thesis, including built examples to validate these design procedures. Both series and parallel feedback DROs are discussed and the procedures for building them are presented. Two examples at different frequencies for each type of DRO are constructed and tested with the results shown. The first is at a frequency of approximately 6.22 GHz and the second for the higher frequency of 11.2 GHz. The DROs for the desired frequencies are designed using the Microwave Office (MWO) software by AWR with the design based on the small-signal model (scattering parameters). Oscillators are produced using the negative resistance method. The circuit achieves low noise by using a dielectric resonator with a high Q factor. Both the series and parallel feedback DRO circuits can be mechanically tuned around the resonant frequency to maximize performance.

Thesis (MScEng (Electrical and Electronic Engineering))--Stellenbosch University, 2008. / This thesis describes the design and implementation of a software platform for software defined radio (SDR). This platform was to form part of an experimental satellite payload. Several other experiments were also housed on this platfrom and subsequently had to be incorporated into the design. The hardware components of the payload were already de- termined at the start of the project but firmware had to be created as part of the project. The software platform was based on the Linux kernel. Device drivers had to be designed for the hardware and firmware components. These drivers were designed so that standard Unix utilities could be used to interact with them. This allowed for easy testing of the system and the programs running on it. The use of the platform for modulation and demodulation of analogue signals was demonstrated using a proof-of-concept SDR application.

Thesis (MScEng (Electrical and Electronic Engineering))--University of Stellenbosch, 2006. / An analysis of the attitude determination and control system required for a small geostationary satellite is performed in this thesis. A three axis quaternion feedback reaction wheel control system is the primary control system used to meet the stringent accuracy requirements. A momentum bias controller is also evaluated to provide redundancy and to extend actuator life. Momentum dumping is preformed by magnetic torque rods using a crossproduct controller. Performance of three axis thruster control is also evaluated. A full state Extended Kalman filter is used to determine attitude and body angular rates during normal operation whereas a Multiplicative Extended Kalman Filter is used during attitude manoeuvres. An analytical orbit control study is also performed to calculate the propellant required to perform station-keeping, for a specific sub-satellite location over a ten year period. Finally an investigation on the effects caused by thruster misalignment, on satellite attitude is also performed.

Thesis (MScEng (Electrical and Electronic Engineering))--University of Stellenbosch, 2006. / This project investigates two problems related to satellites that contain cold gas thrusters and reaction wheels. The first problem concerns the estimation of the thruster force response to a pulse command. An optimal data integration method is developed and implemented to estimate the thruster force output from a number of noisy sensors. The second problem occurs when the output force vector of a thruster mounted on a satellite is misaligned with the centre of mass with the satellite. As a result a disturbance torque is generated on the satellite during thruster operation.

Thesis (MScEng (Electrical and Electronic Engineering))--Stellenbosch University, 2008. / In this thesis the design of a low-cost ad hoc network for short distance data acquisition applications with low data arrival intervals will be presented. The focus is on cost reduction by replacing the traditional high power radios with low-power RF transceivers. The conventional way of using multiple stationary repeater towers (depending on the network) is also replaced by using an ad hoc con guration, where each individual station also serves as a repeater station to adjacent stations. This approach reduces network design time enormously, seeing that the network is able to con gure itself. By using this auto-routing multi-hop approach, data acquisition points are no longer restricted to the reception areas of base stations. A CSMA contention protocol is used for the data communication. Current models used to model this protocol are dependent on various assumptions. In the research reported in this thesis, a statistical study of the collision probability is performed and the results used to expand the current CSMA models. Inter-dependent characteristics of this model are also further enhanced to provide a more realistic model. A simulink model of the particular CSMA protocol is also designed. Both the mathematical- and the simulink models provide relatively good predictions when compared to actual measured results

Thesis (MScEng (Electrical and Electronic Engineering))--Stellenbosch University, 2008. / Automatic bird song recognition and transcription is a relatively new field. Reliable automatic recognition systems would be of great benefit to further research in ornithology and conservation, as well as commercially in the very large birdwatching subculture. This study investigated the use of Hidden Markov Models and duration modelling for bird call recognition. Through use of more accurate duration modelling, very promising results were achieved with feature vectors consisting of only pitch and volume. An accuracy of 51% was achieved for 47 calls from 39 birds, with the models typically trained from only one or two specimens. The ALS pitch tracking algorithm was adapted to bird song to extract the pitch. Bird song synthesis was employed to subjectively evaluate the features. Compounded Selfloop Duration Modelling was developed as an alternative duration modelling technique. For long durations, this technique can be more computationally efficient than Ferguson stacks. The application of approximate string matching to bird song was also briefly considered.

Thesis (MScEng (Electrical and Electronic Engineering))--University of Stellenbosch, 2005. / Most of the relevant and severe congenital cardiac malfunctions can be recognized in the neonatal period of a child’s life. The delayed recognition of a congenital heart defect may have a serious impact on the long-term outcome of the affected child. Experienced cardiologists can usually evaluate heart murmurs with a high sensitivity and specificity, although non-specialists, with less clinical experience, may have more difficulty. Although primary care physicians frequently encounter children with heart murmurs most of these murmurs are innocent. The aim of this project is to design an automated algorithm that can assist the primary care physician in screening and diagnosing pediatric patients with possible cardiac malfunctions. Although attempts have been made to automate screening by auscultation, no device is currently available to fulfill this function. Multiple indicators of pathology are nonetheless available from heart sounds and were elicited using several signal processing techniques. The three feature extraction algorithms (FEA’s) developed respectively made use of a Direct Ratio technique, a Wavelet analysis technique and a Knowledge based neural network technique. Several implementations of each technique are evaluated to identify the best performer. To test the performance of the various algorithms, the clinical auscultation sounds and ECG-data of 163 patients, aged between 2 months and 16 years, were digitized. Results presented show that the De-noised Jack-Knife neural network can classify 163 recordings with a sensitivity and specificity of 92 % and 92.9 % respectively. This study concludes that, in certain conditions, the developed automated auscultation algorithms show significant potential in their use as an alternative evaluation technique for the classification of heart sounds in normal (innocent) and pathological classes.