Doctoral Degrees (Electrical and Electronic Engineering)

Permanent URI for this collection

https://scholar2.stb.sun.ac.za/handle/10019.1/327

Browse

Recent Submissions

Now showing 1 - 5 of 200
  • Thumbnail Image
    Item
    The development of a biosensor for the early detection of pancreatic cancer
    (Stellenbosch : Stellenbosch University, 2023-03) Ebrahim, Taskeen; Perold, Willem; Engelbrecht, Anna-Mart; Stellenbosch University. Faculty of Engineering. Dept. of Electrical and Electronic Engineering.
    ENGLISH ABSTRACT: Pancreatic cancer has one of the highest cancer mortality rates, as it is often detected in late stages. Current methods of detection include diagnostic imaging tests or laboratory intensive blood tests such as radioimmunoassays or ELISAs. Researchers have used biosensors for detection and monitoring of different medical conditions, as a more accessible and cheaper alternative. This project focuses on the design of a biosensor toward the early detection of pancreatic cancer, using Carbohydrate Antigen 19-9 (CA19-9) as the selected biomarker for the biosensor. Electrochemical impedance spectroscopy was identified as an appropriate transducer mechanism for this biosensor and uses gold interdigitated electrodes as a sensor transducer surface. Anti-CA19-9 antibodies were immobilized onto gold using covalent bonding and crosslinking chemistries, and binding was validated using fluorescence microscopy. After electrodes were electronically characterized to identify the appropriate impedance and frequency ranges, an impedance analyser was designed, fabricated, and tested, with added computation of complex capacitance. The impedance analyser was calibrated and tested relative to the PalmSens4 Electrochemical Interface. The designed impedance analyser showed mean impedance and phase errors of 0.538% and 0.381% respectively. Similarly, the accuracy of complex capacitance computation showed errors of 1.222% and 0.656% for real and imaginary components respectively. The impedance analyser differentiated between changes in concentration of phosphate buffered saline using complex capacitance. The biosensor was tested with five concentrations of CA19-9 and differentiated between concentrations above and below the thresholds for pancreatic cancer. The design of a novel impedance analyser utilizing complex capacitance computation for the quantification of CA19-9 using IDEs fabricated on an FR4 board is a unique contribution to biosensor research.
  • Thumbnail Image
    Item
    Approximate message passing algorithms for latent Dirichlet allocation
    (2023-03) Taylor, Rebecca; Du Preez, Johan; Stellenbosch University. Faculty of Engineering. Dept. of Electrical and Electronic Engineering.
    ENGLISH ABSTRACT: Latent Dirichlet allocation (LDA) is a hierarchical Bayesian model that is most well known for its ability to extract latent semantic topics from text corpora. Since exact inference of the posterior distribution is intractable, a variety of approximate inference techniques can be employed for LDA. The traditional variational Bayesian inference (VB) approach is sufficient for most text topic modelling problems. VB, however, does not perform topic extraction well where the corpora of documents are small, where document sizes are limited, or where topic overlap is large. In these cases, collapsed Gibbs sampling, a slower, computationally more expensive inference method, is typically preferred. In this dissertation, we present two variational message passing algorithms for inference in LDA. Our first algorithm, approximate LBU (designated by the acronym ALBU), is derived by applying loopy belief update (LBU) (also known as the Lauritzen-Spiegelhalter algorithm), where possible, and using ideas from hierarchical sampling to derive messages where conjugacy does not apply. The second algorithm, which we designate variational message passing plus (VMP+), is based on variational message passing (VMP), the message passing variant of VB. In VMP+, instead of following the VMP algorithm exactly, we relax the mean-field assumption between parent and child nodes that are in the same cluster of the cluster graph representation of LDA. This results in an algorithm that is similar to VB, but performs better at topic extraction. To evaluate ALBU and VMP+, we use four text corpora (20 Newsgroups, 6 Newsgroups, Covid tweets, and Bible verses), all of which contain short texts, to compare the performance of our two algorithms with VB and collapsed Gibbs sampling using topic coherence scores. Because VB typically performs well on large data sets, we apply these two new variants specifically to smaller data sets where topic extraction is expected to be difficult. In addition to the real-life text documents described above, we perform topic extraction on a variety of simulated corpora which are obtained by using different hyperparameter settings, to enable comparisons among the estimated results and the true distributions. Here, Kullback-Leibler divergence (KLD) is used to evaluate and compare inference algorithm performance. Based on results obtained for both the real-life and synthetic data, we show that both algorithms outperform VB, and that the performance of ALBU is similar to that of collapsed Gibbs sampling only when the sampler is given an extended period of time to converge. ALBU appears to be an efficient approach to inference in LDA in certain situations where topic extraction is expected to be difficult. VMP+ is an alternative inference algorithm for LDA that is simpler than VB and also typically extracts more coherent topics than VB. On the basis of our experiments, we summarise the differences between the variational algorithms, and propose reasons for these differences. We conclude by providing recommendations regarding the methods that we anticipate would be most suitable for inference on particular corpus types.
  • Thumbnail Image
    Item
    Modelling and optimization of linear-motion kinetic energy harvesters: two approaches
    (Stellenbosch : Stellenbosch University, 2023-03) Struwig, Michael; Niesler, Thomas; Stellenbosch University. Faculty of Engineering. Dept. of Electrical and Electronic Engineering.
    ENGLISH ABSTRACT: Energy harvesting is a well-established method for extending the life of battery-powered devices, such as wildlife tracking collars. However, the operating conditions of these devices provide a number of challenges, such as size and weight constraints. They are also typically exposed to the non-harmonic forms of mechanical motion associated with animal footsteps. This renders much of the existing literature inapplicable, because it applies only to harmonic excitation. We propose a microgenerator architecture that consists of a variable number of evenly spaced magnets, forming a fixed assembly that is free to move through a series of evenly spaced coils, and is supported by a magnetic spring. Based on this architecture, we develop two microgenerator design approaches, each with their own electro-mechanical system model and optimization philosophy. The first approach assumes idealized (constant velocity) motion as a proxy for optimizing for the true, non-harmonic motion. We applied this method to design an optimal energy harvester for impulsive motion, resulting in a device with a length of approximately 125 mm and tube diameter of 11 mm that generated an average power of 3.01 mW in a 40 Ω test load from a 2.2 g impact force from a walking human test subject. The same method was applied to design a 85 mm length-constrained device, which was subsequently field-tested on a wild rhinoceros. When the animal walked slowly, this device generated an average of 0.342 mW. The second approach is based on an extended time-domain system model that, after an evolutionary parameter search, can predict the temporal behaviour of a microgenerator to within 25% of the measured load voltage RMS, for any chosen input excitation. Utilizing this model, we propose an enhanced optimization process that selects a set of energy harvester design parameters that maximizes the power delivered to a resistive load, resulting in an optimized device that is specific to any choice of input excitation. The resulting optimal design has a length of approximately 135 mm and a tube diameter of 11 mm and was found to deliver an average power of 1.526 mW to a 30 Ω load when driven by a less vigorous human footstep-like motion with a 1.5 g impact force. Finally, we introduce an open-source declarative energy harvester framework, FFS, and demonstrate how it can be used to design, simulate and optimize their energy harvester models.
  • Thumbnail Image
    Item
    Estimation of external disturbance torques on CubeSats to enhance full state attitude estimation accuracy and mission success
    (Stellenbosch : Stellenbosch University, 2023-03) Habila, Asia Saeed Kajo; Steyn, Herman; Stellenbosch University. Faculty of Engineering. Dept. of Electrical and Electronic Engineering.
    ENGLISH ABSTRACT: CubeSats are becoming increasingly popular, and the number of launches is growing rapidly. This is due to their low manufacturing costs and ease of launch. CubeSats were initially utilized for educational purposes; however, over time, they began to be used for more complex missions, such as astronomy and remote sensing applications. For these complex missions to be successful, precise control is required, which is the role of the attitude determination and control subsystem (ADCS). ADCS, on the other hand, has challenges achieving precise control due to the CubeSat's small moment of inertia, which makes it susceptible to disturbance torques. This thesis focuses on the accurate estimation and development of models for disturbance torques that considerably improve the full-state estimate accuracy and, as a result, the CubeSats' control and, eventually, preserve the mission. As measuring the disturbance torque that affects an object is as simple as measuring its angular acceleration, the gyroscope is chosen as the main attitude sensor in this thesis to estimate attitude disturbance torques. However, the gyroscope was found to have shortcomings due to noise and errors that affect its output and performance, resulting in an inaccuracy of measurement. As a consequence, two noise drift estimation algorithms are developed, tested, and evaluated in order to determine which is best to use to boost the accuracy of disturbance torque estimation. The general models for attitude disturbance torques, along with the estimators, are developed, tested, and evaluated. The outcomes from the models are used to assist in the eventual development of the estimators. For the estimators, deterministic and nonlinear recursive estimation methods are developed, tested, and evaluated through simulations in order to determine which are best to be used in the thesis to estimate the attitude disturbance torques. The deterministic algorithms that are utilized are QUEST and Linear Least Square, while the recursive ones are Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF). In addition, a suggested method (algorithm) for estimating attitude disturbance is proposed, followed by simulations for assessment. Furthermore, an identification technique is proposed for determining which attitude disturbance torques are acting on the satellite body, and their magnitude is also identified and evaluated through simulations. A design and practical implementation of a disturbance torque estimator on the ground is developed. An air bearing platform is used to represent the satellite in terms of potential space disturbances and to evaluate the effectiveness of the disturbance torque estimator. Firstly, the mass properties of the air bearing are obtained using two different methods to ensure the air bearing system dynamics is correct, and then the estimator is developed and evaluated. The EKF was found to be the most suitable choice that can be utilized in this research to estimate the disturbance torque. It gives a more accurate estimate compared to QUEST and Weighted Least Square (WLS), is more precise, and requires less execution time compared to UKF. Accordingly, a bank of EKFs is employed to independently estimate each of the disturbance torques and used in combination with the identification function to accurately estimate the total disturbance torque in orbit.
  • Thumbnail Image
    Item
    A figure of merit for the x-band all-sky survey
    (Stellenbosch : Stellenbosch University, 2023-03) Kriel, Scott; De Villiers, Dirk; Stellenbosch University. Faculty of Engineering. Dept. of Electrical and Electronic Engineering.
    ENGLISH ABSTRACT: The X-Band All-Sky Survey (X-BASS), is an experiment aimed at mapping the sky, both in intensity and polarisation, in the frequency range 7—15 GHz. This is an extension of the C-Band All-Sky Survey (C-BASS), taking the form of a single-dish polarimeter located in Klerefontein, South Africa. Using Cassegrain-like reflector surfaces already in place as part of C-BASS, the work will focus on the design of a feed antenna forthe new frequency band of interest. In traditional design algorithms, generic Figures of Merit (FoMs) describing the radiation pattern variations are used as design goals. These multiple FoMs are heuristically determined, and it is normally difficult to ascertain the quantitative importance of each FoM when combining them to describe the actual system performance. This is especially true for X-BASS, where the specific role of the instrument poses a unique set of requirements, the trade-off of which is not immediately apparent. As such, this work establishes new FoMs which directly relate antenna properties to their effect on distinct science goals. In order to accomplish this, attention must be paid toward the specific purpose of X-BASS, which is to aid experiments observing the cosmic microwave Background (CMB). Radiation from the CMB is seen to fluctuate across the sky in both temperature and polarisation. It is in the measurement of these anisotropies that lies a wealth of information on the origins of the universe and the physical mechanisms governing its evolution. The next generation of experiments seek to observe the CMB polarisation with an extraordinary degree of accuracy, placing ever-increasing demands on the instruments used to perform them. As such, more tailored design approaches than those that currently exist are required to optimise antenna structures in a way which maximises the quality of these science outcomes. This work sets out to accomplish this, providing a FoM for the X-BASS antenna which is readily calculable and relates directly to the uncertainty seen in the determination of cosmological parameters from CMB observations. Furthermore, this description is not specific to X-BASS, but could be applied more generally to other instruments aimed at measuring the CMB. With the FoM developed here at hand, future work can focus on rigorous analysis and optimisation of the antenna structure, allowing the effect of specific design choices to be studied in detail. The prescription could also be extended to consider more non-ideal effects, adding to the complexity of the analysis and providing a tool by which the next generation of CMB instruments can be designed.