Browsing by Subject "Electrical Engineering"

Now showing 1 - 20 of 748
  • Thumbnail Image
    Item
    Open Access
    3-Phase gate-turn off thyristor inverter
    (1986) Kleyn, D A
    The requirements of a standard 3-phase Induction Motor driven by a Voltage Source Inverter (VSI) are studied. A full 3-phase Variable Speed Drive (VSD) and its controller have been designed, constructed and tested. Gate Turn-Off Thyristors (GTO's) are used as the main switching elements in the Inverter stage of the Drive. The drive requirements of GTO's are studied in detail.
  • Thumbnail Image
    Item
    Open Access
    A 3-phase Z-source inverter driven by a novel hybrid switching algorithm
    (2007) Malengret, Jean-Claude; Braae, Martin
    A 3-phase Z-source inverter has been researched, designed, simulated, builtand tested. The purpose of the inverter is to deliver 3-phase 400 VAC from aDC supply that can vary over a range of 20 to 70 Vdc. This is done with a Zsourceinverter topology which is a single conversion method with no additionalDC to DC boost converter. A novel DSP control algorithm allows the inverter toachieve the following:· Run Space Vector Pulse Width Modulation (SV-PWM) for maximum DCbus voltage utilization while boosting the DC bus during zero space vectorstates using shoot through.· Seamless transition between modulation control and modulation / shootthrough control.· Optimised efficiency and DC bus utilisation using Hybrid Space VectorBoost Pulse Width Modulation (HSVB PWM) which is unique to thisdissertation.Such a system is particularly suited to fuel cell and particularly wind turbineapplications where the DC bus voltage is varies over a wide range resulting inthe need for a DC to DC buck/boost to regulate the DC bus to maintain a steady3-phase sinusoidal output. A further application could be for general purpose 3-phase inverter capable of operating on different DC standard bus voltages ( e.g.24, 36, 48 VDC).The benefits of a Z-source topology for the above purposes are a reduction inhigh power semi-conductor components (e.g. power MOSFET). There is also areduction in switching losses and inherent shoot through protection.Furthermore, the inverter is more robust in the sense that it is not vulnerable to spurious shoot through, which could be disastrous in the case of a traditionalvoltage fed inverter.
  • No Thumbnail Available
    Item
    Open Access
    3D intrawall imaging using backprojection for synthetic aperture radar (SAR)
    (2024) Dass, Reevelen; Paine, Stephen
    The Council of Scientific and Industrial Research (CSIR) has evolving synthetic aperture radar (SAR) capabilities in the C-band and the L-band. Currently, these capabilities are used to generate aerial landscape images; however, to explore the feasibility of using this technology in different environments, an experimental SAR system has been developed. This is referred to as the wall scanner. The purpose of the wall scanner is to image the interior of a wall, revealing details of the substructures inside the wall such as conduits and piping. This is done by moving the antenna system across the wall surface to create SAR images using backprojection. The radar used two different types of antennas, a log periodic dipole array (LPDA) antenna and horn antenna. The horn antenna performed well in the experiments, producing images with minimal artefacts. On the contrary, the LPDA antenna did not perform as well in the experiments and as such the characteristics of the antenna were investigated. The investigation revealed that the antenna did not function throughout the frequency range specified by its manufacturer. This produced artefacts in the image; however, some of the effects of these artefacts were minimised by a series of preprocessing techniques. A variety of preprocessing techniques were used to improve image quality. In addition to compensating for the properties of the LPDA antenna, windowing and different methods of background subtraction were used. It was difficult to compensate for the antenna issues in preprocessing; however, windowing and background subtraction had a significant effect on the images that were produced. Two postprocessing techniques were used, gradient descent optimisation based on image contrast and polarimetry. The developed gradient descent optimiser was able to automatically adjust for the system group delay based on the contrast of the image. Polarimetry post-processing revealed that the horizontally transmitted horizontally received polarisation (HH) and vertically transmitted vertically received polarisation (VV) were effective in creating images in this environment; however, cross-polarisation in the form of horizontally transmitted vertically received polarisation (HV) was not effective. The wall scanning environment that was measured consisted of scanning both drywall and brick wall. This was split into three experiments. The experiments used different materials that were placed in front of a wall, behind the wall at a distance, and directly behind the wall. The wall scanner was able to successfully create images of the three different experiments for the drywall; however, the desired results for the brick wall were not achieved. For drywall, the substructures placed directly behind the wall were more difficult to see because they were masked by the wall and its sidelobes. The materials scanned were a copper pipe, a PVC pipe, a wooden beam, and a highly reflective calibration target. The calibration target and the copper target performed well in the three experiments. The wooden beam did not perform as well; especially when placed directly behind the wall; however, it was still visible in all experiments. The PVC performed the worst and was only faintly visible in the experiments and was not visible when placed directly behind the wall.
  • Thumbnail Image
    Item
    Open Access
    3D model reconstruction using photoconsistency
    (2007) Joubert, Kirk Michael; Nicolls, Fred; De Jager, Gerhard
    Model reconstruction using photoconsistency refers to a method that creates a photohull, an approximate computer model, using multiple calibrated camera views of an object. The term photoconsistency refers to the concept that is used to calculate the photohull from the camera views. A computer model surface is considered photoconsistent if the appearance of that surface agrees with the appearance of the surface of the real world object from all camera viewpoints. This thesis presents the work done in implementing some concepts and approaches described in the literature.
  • Thumbnail Image
    Item
    Open Access
    3D reconstruction and camera calibration from 2D images
    (2000) Henrichsen, Arne; De Jager, Gerhard
    A 3D reconstruction technique from stereo images is presented that needs minimal intervention from the user. The reconstruction problem consists of three steps, each of which is equivalent to the estimation of a specific geometry group. The first step is the estimation of the epipolar geometry that exists between the stereo image pair, a process involving feature matching in both images. The second step estimates the affine geometry, a process of finding a special plane in projective space by means of vanishing points. Camera calibration forms part of the third step in obtaining the metric geometry, from which it is possible to obtain a 3D model of the scene. The advantage of this system is that the stereo images do not need to be calibrated in order to obtain a reconstruction. Results for both the camera calibration and reconstruction are presented to verify that it is possible to obtain a 3D model directly from features in the images.
  • Thumbnail Image
    Item
    Open Access
    A 500kHz-5MHz CW stepped frequency borehole tomographic imaging system
    (2001) Isaacson, Adam Rhett; Inggs, Michael
    This dissertation involves a study of Cross-Borehole Tomography. The mathematical physical models of the Radon Transform are reviewed. The entire Cross-Borehole Tomographic process is simulated, based on these physical models of the Radon Transform. The system specifications for the final design are based on the results from the simulation. Finally, the final design is built, and tested. The phase yields a better quality of image reconstruction when compared to amplitude, and hence a coherent system is a good choice. The system is frequency to frequency coherent for the entire transmit frequency range, which satisfies the main aim of this dissertation.
  • No Thumbnail Available
    Item
    Open Access
    A machine vision-based approach to measuring the size distribution of rocks on a conveyor belt
    (2004) Mkwelo, Simphiwe; de Jager, Gerhard
    This work involves the development of a vision-based system for measuring the size distribution of rocks on a conveyor belt. The system has applications in automatic control and optimization of milling machines, and the selection of optimal blasting methods in the mining industry. Rock size is initially assumed to be the projected rock surface area due to the constraint imposed by the 2D nature of images. This measurement is facilitated by locating connected rock-edge pixels. Rock edge detection is achieved using a watershed-based segmentation process. This process involves image pre-filtering with edge preserving filters at various degrees of filtering. The output of each filtering stage is retained and marker-driven watersheds are applied on each output resulting to traces of detected rock boundaries. Watershed boundary selection is then applied to select boundaries which are most likely to be rock edges based on rock features. Finally, rock recognition using feature classification is applied to remove non-rock watershed boundaries. The projected rock area distribution of a test-set is measured and compared to corresponding projected areas of manually segmented images. The obtained distributions are found to be similar with an RMS error of 2.37% on the test-set. Finally, sieve data is collected in the form of actual rock size distributions and a quantitative comparison between the actual and machine measured distributions is performed. The overall quantitative result is that the two rock size distributions are significantly different. However, after incorporating a stereology-based correction, hypothesis tests on a 3m belt-cut test-set show that the obtained distributions are similar.
  • Thumbnail Image
    Item
    Open Access
    AAL2 signalling framework to support a Gigabit AAL2 switching node
    (2003) Van Zyl, André; Ventura, Neco
    Although Asynchronous Transfer Mode (ATM) Adaptation Layer type 2 (AAL2) has been in existence for a few short years, it is widely being adopted as the technology of choice in VoDSL. VOTA trunking, as well as in UMTS wireless networks. The two most important concerns when transporting voice in a packet based network, are end-to-end delay and the efficient use of available bandwidth.
  • Thumbnail Image
    Item
    Open Access
    AAL2 switching node to support voice services in 3rd and 4th generation networks
    (2002) Shepstone, Sven E; Ventura, Neco
    The research community and industry alike have, over the past decade, been showing considerable interest in packet-switching networks to support voice services as well as data services. A technology that was standardised to accommodate these delay-sensitive requirements is Asynchronous Transfer Mode (ATM), which deals particularly well at transporting uncompressed voice and data. However, due to the exponential increase in wireless applications and their supporting access technologies, a need has arisen for an infrastructure in the wide area network to support and maintain the QoS requirements of low-bit rate, compressed voice. An adaptation layer known as AAL2 was re-standardised to support these specialised voice services. However, a severe side-effect of using AAL2 with traditional ATM switches results in inefficient routing and waste-age of resources. In this study, a design for an AAL2 switching node will be proposed to address the above-mentioned issues. The design is comprised of modules that perform the following functions: Buffering, payload interrogation, protocol translations, packet classification, packet re- routing, timing, scheduling and support for signalling and management interfacing. The supporting architecture is targeted towards an embedded >286-based computing system, which itself is overlaid upon one or several ports of a high-speed, research-oriented ATM switch, known as the Washington University Gigabit Switch (WUGS). In order to evaluate the operation and performance of the AAL2 switch architecture, a testbed is proposed and implemented, comprising the AAL2 switch at the core, with a supporting infrastructure to emulate the generation and analysis of low bit-rate voice traffic over an AAL2 connection. By conducting a set of experiments, a series of operational and performance results will be presented. Particular focus will be placed on the performance and efficiency of the AAL2 layer over ATM, as well as the ability of the switch to route packets from multiple sources to a set of output connections in the correct manner.
  • Thumbnail Image
    Item
    Open Access
    Accelerating Gauss-Newton filters on FPGA's
    (2010) Da Conceicao, Jean-Paul Costa; Inggs, Michael
    Radar tracking filters are generally computationally expensive, involving the manipulation of large matrices and deeply nested loops. In addition, they must generally work in real-time to be of any use. The now-common Kalman Filter was developed in the 1960's specifically for the purposes of lowering its computational burden, so that it could be implemented using the limited computational resources of the time. However, with the exponential increases in computing power since then, it is now possible to reconsider more heavy-weight, robust algorithms such as the original nonrecursive Gauss-Newton filter on which the Kalman filter is based. This dissertation investigates the acceleration of such a filter using FPGA technology, making use of custom, reduced-precision number formats.
  • Thumbnail Image
    Item
    Open Access
    Accelerating software radio astronomy FX correlation with GPU and FPGA co-processors
    (2010) Woods, Andrew; Inggs, Michael
    This thesis attempts to accelerate compute intensive sections of a frequency domain radio astronomy correlator using dedicated co-processors. Two co-processor implementations were made independently with one using reconfigurable hardware (Xilinx Virtex 4LXlOO) and the other uses a graphics processor (Nvidia 9800GT). The objective of a radio astronomy correlator is to compute the complex valued correlation products for each baseline which can be used to reconstruct the sky's radio brightness distribution. Radio astronomy correlators have huge computation demands and this dissertation focuses on the computational aspects of correlation, concentrating on the X-engine stage of the correlator.
  • Thumbnail Image
    Item
    Open Access
    Acceleration of parasitic multistatic radar system using GPGPU
    (2011) John, Mathew; Inggs, Michael
    This dissertation details the implementation of PMR [Parasitic Multistatic Radar] signal processing chain in the GPGPU [General Purpose Graphic Processing Units] platform. The primary objective of the project is to accelerate the signal processing chain without compromising the algorithm efficiency and to prove that GPGPUs are a promising platform for parasitic radar signal processing.
  • Thumbnail Image
    Item
    Open Access
    Access network selection schemes for multiple calls in next generation wireless networks
    (2017) Orimolade, Joseph Folorunsho; Ventura, Neco; Falowo, Olabisi E
    There is an increasing demand for internet services by mobile subscribers over the wireless access networks, with limited radio resources and capacity constraints. A viable solution to this capacity crunch is the deployment of heterogeneous networks. However, in this wireless environment, the choice of the most appropriate Radio Access Technology (RAT) that can Tsustain or meet the quality of service (QoS) requirements of users' applications require careful planning and cost efficient radio resource management methods. Previous research works on access network selection have focused on selecting a suitable RAT for a user's single call request. With the present request for multiple calls over wireless access networks, where each call has different QoS requirements and the available networks exhibit dynamic channel conditions, the choice of a suitable RAT capable of providing the "Always Best Connected" (ABC) experience for the user becomes a challenge. In this thesis, the problem of selecting the suitable RAT that is capable of meeting the QoS requirements for multiple call requests by mobile users in access networks is investigated. In addressing this problem, we proposed the use of Complex PRoprtional ASsesment (COPRAS) and Consensus-based Multi-Attribute Group Decision Making (MAGDM) techniques as novel and viable RAT selection methods for a grouped-multiple call. The performance of the proposed COPRAS multi-attribute decision making approach to RAT selection for a grouped-call has been evaluated through simulations in different network scenarios. The results show that the COPRAS method, which is simple and flexible, is more efficient in the selection of appropriate RAT for group multiple calls. The COPRAS method reduces handoff frequency and is computationally inexpensive when compared with other methods such as the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), Simple Additive Weighting (SAW) and Multiplicative Exponent Weighting (MEW). The application of the proposed consensus-based algorithm in the selection of a suitable RAT for group-multiple calls, comprising of voice, video-streaming, and file-downloading has been intensively investigated. This algorithm aggregates the QoS requirement of the individual application into a collective QoS for the group calls. This new and novel approach to RAT selection for a grouped-call measures and compares the consensus degree of the collective solution and individual solution against a predefined threshold value. Using the methods of coincidence among preferences and coincidence among solutions with a predefined consensus threshold of 0.9, we evaluated the performance of the consensus-based RAT selection scheme through simulations under different network scenarios. The obtained results show that both methods of coincidences have the capability to select the most suitable RAT for a group of multiple calls. However, the method of coincidence among solutions achieves better results in terms of accuracy, it is less complex and the number of iteration before achieving the predefined consensus threshold is reduced. A utility-based RAT selection method for parallel traffic-streaming in an overlapped heterogeneous wireless network has also been developed. The RAT selection method was modeled with constraints on terminal battery power, service cost and network congestion to select a specified number of RATs that optimizes the terminal interface utility. The results obtained show an optimum RAT selection strategy that maximizes the terminal utility and selects the best RAT combinations for user's parallel-streaming for voice, video and file-download.
  • Thumbnail Image
    Item
    Open Access
    Active filter current compensation for transmission optimisation
    (2015) Carpenter, Paul Andrew; Gaunt, C Trevor; Malengret, Michel
    This dissertation is based on the fact that any m-wire electrical system can be modelled as m-equivalent Thevenin voltages and impedances when viewed from any node. The dissertation describes how to calculate the optimal distribution of currents, so a specific amount of power can flow through and reach the network equivalent Thevenin voltages with minimal losses. The optimal current distribution method uses a recently patented method which calculates the optimal currents for each of the wires which are shown to be obtained from the Thevenin parameters and power flow at any instant in time at any node. Once the ideal currents are found, these can be obtained by active and passive devices to inject a specific amount of power (positive and negative) as to compensate existing currents. The focus is particularly on the proof of concept by simulations and physical experiments with work not specifically described in the patent with more emphasis on the optimisation to active compensation. It is explained and shown how this can be implemented using the Malengret and Gaunt method. This method reduces the cost in application where not all the currents need to be processed through a converter (e.g. inverter) but only the difference between the existing and desired optimal currents. A smaller shunt parallel converter can result with ideal current flow without the need for interrupting the currents as described in the present patent. The methodology is explained and demonstrated by simulation.
  • Thumbnail Image
    Item
    Open Access
    Active object recognition for 2D and 3D applications
    (2015) Govender, Natasha; Nicolls, Frederick
    Active object recognition provides a mechanism for selecting informative viewpoints to complete recognition tasks as quickly and accurately as possible. One can manipulate the position of the camera or the object of interest to obtain more useful information. This approach can improve the computational efficiency of the recognition task by only processing viewpoints selected based on the amount of relevant information they contain. Active object recognition methods are based around how to select the next best viewpoint and the integration of the extracted information. Most active recognition methods do not use local interest points which have been shown to work well in other recognition tasks and are tested on images containing a single object with no occlusions or clutter. In this thesis we investigate using local interest points (SIFT) in probabilistic and non-probabilistic settings for active single and multiple object and viewpoint/pose recognition. Test images used contain objects that are occluded and occur in significant clutter. Visually similar objects are also included in our dataset. Initially we introduce a non-probabilistic 3D active object recognition system which consists of a mechanism for selecting the next best viewpoint and an integration strategy to provide feedback to the system. A novel approach to weighting the uniqueness of features extracted is presented, using a vocabulary tree data structure. This process is then used to determine the next best viewpoint by selecting the one with the highest number of unique features. A Bayesian framework uses the modified statistics from the vocabulary structure to update the system's confidence in the identity of the object. New test images are only captured when the belief hypothesis is below a predefined threshold. This vocabulary tree method is tested against randomly selecting the next viewpoint and a state-of-the-art active object recognition method by Kootstra et al.. Our approach outperforms both methods by correctly recognizing more objects with less computational expense. This vocabulary tree method is extended for use in a probabilistic setting to improve the object recognition accuracy. We introduce Bayesian approaches for object recognition and object and pose recognition. Three likelihood models are introduced which incorporate various parameters and levels of complexity. The occlusion model, which includes geometric information and variables that cater for the background distribution and occlusion, correctly recognizes all objects on our challenging database. This probabilistic approach is further extended for recognizing multiple objects and poses in a test images. We show through experiments that this model can recognize multiple objects which occur in close proximity to distractor objects. Our viewpoint selection strategy is also extended to the multiple object application and performs well when compared to randomly selecting the next viewpoint, the activation model and mutual information. We also study the impact of using active vision for shape recognition. Fourier descriptors are used as input to our shape recognition system with mutual information as the active vision component. We build multinomial and Gaussian distributions using this information, which correctly recognizes a sequence of objects. We demonstrate the effectiveness of active vision in object recognition systems. We show that even in different recognition applications using different low level inputs, incorporating active vision improves the overall accuracy and decreases the computational expense of object recognition systems.
  • Thumbnail Image
    Item
    Open Access
    Adaptive and autonomous protocol for spectrum identification and coordination in ad hoc cognitive radio network
    (2017) Ohize, Henry Ohiani; Dlodlo, Mqhele E
    The decentralised structure of wireless Ad hoc networks makes them most appropriate for quick and easy deployment in military and emergency situations. Consequently, in this thesis, special interest is given to this form of network. Cognitive Radio (CR) is defined as a radio, capable of identifying its spectral environment and able to optimally adjust its transmission parameters to achieve interference free communication channel. In a CR system, Dynamic Spectrum Access (DSA) is made feasible. CR has been proposed as a candidate solution to the challenge of spectrum scarcity. CR works to solve this challenge by providing DSA to unlicensed (secondary) users. The introduction of this new and efficient spectrum management technique, the DSA, has however, opened up some challenges in this wireless Ad hoc Network of interest; the Cognitive Radio Ad Hoc Network (CRAHN). These challenges, which form the specific focus of this thesis are as follows: First, the poor performance of the existing spectrum sensing techniques in low Signal to Noise Ratio (SNR) conditions. Secondly the lack of a central coordination entity for spectrum allocation and information exchange in the CRAHN. Lastly, the existing Medium Access Control (MAC) Protocol such as the 802.11 was designed for both homogeneous spectrum usage and static spectrum allocation technique. Consequently, this thesis addresses these challenges by first developing an algorithm comprising of the Wavelet-based Scale Space Filtering (WSSF) algorithm and the Otsu's multi-threshold algorithm to form an Adaptive and Autonomous WaveletBased Scale Space Filter (AWSSF) for Primary User (PU) sensing in CR. These combined algorithms produced an enhanced algorithm that improves detection in low SNR conditions when compared to the performance of EDs and other spectrum sensing techniques in the literature. Therefore, the AWSSF met the performance requirement of the IEEE 802.22 standard as compared to other approaches and thus considered viable for application in CR. Next, a new approach for the selection of control channel in CRAHN environment using the Ant Colony System (ACS) was proposed. The algorithm reduces the complex objective of selecting control channel from an overtly large spectrum space,to a path finding problem in a graph. We use pheromone trails, proportional to channel reward, which are computed based on received signal strength and channel availability, to guide the construction of selection scheme. Simulation results revealed ACS as a feasible solution for optimal dynamic control channel selection. Finally, a new channel hopping algorithm for the selection of a control channel in CRAHN was presented. This adopted the use of the bio-mimicry concept to develop a swarm intelligence based mechanism. This mechanism guides nodes to select a common control channel within a bounded time for the purpose of establishing communication. Closed form expressions for the upper bound of the time to rendezvous (TTR) and Expected TTR (ETTR) on a common control channel were derived for various network scenarios. The algorithm further provides improved performance in comparison to the Jump-Stay and Enhanced Jump-Stay Rendezvous Algorithms. We also provided simulation results to validate our claim of improved TTR. Based on the results obtained, it was concluded that the proposed system contributes positively to the ongoing research in CRAHN.
  • Thumbnail Image
    Item
    Open Access
    Adaptive Bandwidth Management and Joint Call Admission Control to Enhance System Utilization and QoS in Heterogeneous Wireless Networks
    (Springer, 2007) Falowo, Olabisi E; Chan, HAnthony
    :The coexistence of different cellular networks in the same area necessitates joint radio resource management for enhanced QoS provisioning and efficient radio resource utilization. We propose adaptive bandwidth management and joint call admission control (JCAC) scheme for heterogeneous cellular networks. The objectives of the proposed adaptive JCAC scheme are to enhance average system utilization, guarantee QoS requirements of all accepted calls, and reduce new call blocking probability and handoff call dropping probability in heterogeneous wireless networks. We develop a Markov chain model for the adaptive JCAC scheme and derive new call blocking probability, handoff call dropping probability, and average system utilization. Performance of the proposed adaptive JCAC scheme is compared with that of nonadaptive JCAC scheme in the same heterogeneous wireless network. Results show an improvement in average system utilization of up to 20%. Results also show that connection-level QoS can be significantly improved by using the proposed adaptive JCAC scheme.
  • Thumbnail Image
    Item
    Open Access
    Adaptive radio resource management for mobile satellite systems
    (2007) Abel, Ajibesin Adeyemi; Gaunt, C T
    In this thesis, a set of unique strategies and enhanced schemes for adaptive CDMA modulation are devised. A graded resource system is proposed for better radio resource management. Subsequently, a successful adaptive CDMA algorithm is designed and a prioritised processing gain for adaptive CDMA algorithm in satellite system is introduced. The idea of the critical section in the downlink system when a user controller scheme has to be activated to improve the performance is initiated. The diversity technique and rate compatible punctured turbo-code (RCPT), which has been found to give improved throughput performance in a direct sequence (DS) CDMA, are exploited.
  • Thumbnail Image
    Item
    Open Access
    Adaptive single to three-phase converter using variable passive elements
    (2004) Ashrafi-Monfared, ShI; Malengret, Michel
    The main objective of this thesis is to investigate the possible methods of converting a single-phase supply to a three-phase balanced voltage, using passive elements. Three distinct methodologies have been developed, each with its own merits and constructed prototypes. The seven chapters of this thesis contain large amound of non-real and real-time simulated results. These results were then compared with the experimental results.
  • Thumbnail Image
    Item
    Open Access
    An adaptive SVC-video streaming scheme for mobile WiMAX networks
    (2011) Keta, Khotso Hamilton; Murgu, Alexandru
    The problem that is being addressed in this study is bandwidth fluctuation that occurs during a video stream delivery over the Mobile WiMAX network, which results in video quality degradation. To address this problem, H.264AVC/SVC coded video is used.