A dissertation submitted in fulfilment of the requirements for the degree of Masters of Science in Engineering (Electrical) to the Faculty of Engineering and the Built Environment, University of the Witwatersrand, Johannesburg, 2017 / The Travelling Salesman Problem (TSP) is an important NP-hard combinatorial optimisation problem that forms the foundation of many modern-day, practical problems such as logistics or network route planning. It is often used to benchmark discrete optimisation algorithms since it is a fundamental problem that has been widely researched. The Flower Pollination Algorithm (FPA) is a continuous optimisation algorithm that demonstrates promising results in comparison to other well-known algorithms. This research proposes the design, implementation and testing of two new algorithms based on the FPA for solving discrete optimisation problems, more specifically the TSP, namely the Discrete Flower Pollination Algorithm (DFPA) and the iterative Discrete Flower Pollination Algorithm (iDFPA). The iDFPA uses two proposed update methods, namely the Best Tour Update (BTU) and the Rejection Update (RU), to perform the iterative update process. The two algorithms are compared to the Ant Colony Optimisation’s (ACO) MAX−MIN Ant System (MMAS) as well as the Genetic Algorithm (GA) since they are well studied and developed. The DFPA and iDFPA results are significantly better than the GA and the iDFPA is able to outperform the ACO in all tested instances. The iDFPA with 300 iterations was able to achieve the optimal solution in the Berlin52 benchmark TSP problem as well as have improvements of up to 4.56% and 41.87% compared to the ACO and GA respectively. An analysis of how the RU and the annealing schedule used in the RU impacts on the overall results of the iDFPA is given. The RU analysis demonstrates how the annealing schedule can be manipulated to achieve certain results from the iDFPA such as faster convergence or better overall results. A parameter analysis is performed on both the DFPA and iDFPA for different TSP problem sizes and the suggested initial parameters for these algorithms are outlined. / XL2018

Shear banding is a commonly observed yet complex form of instability in granular media by which the deformation is localized in a narrow zone along a certain path. The aim of this study is to investigate the micromechanics of shear banding using the discrete element method (DEM). For this purpose, a model was developed and calibrated to simulate the macroscale behavior of sand under plane strain conditions. Upon validation against laboratory experiments, two types of confining boundaries, displacement- and force-controlled, were examined to study the kinematics of shear bands. A constant volume test was then used to investigate the evolution of antisymmetric stresses before, during, and after shear band formation. The results indicate that the antisymmetric stresses significantly increase within the shear band throughout the loading history, but may not describe the precursory shear band conditions. The DEM model is shown to properly capture the micromechanics of shear bands.

Optimization techniques are widely applied in the power system planning and operation to achieve more efficient and reliable power supply. With the introduction of new technologies, the complexity of today’s power system increased significantly. Intelligent optimization techniques, such as Particle Swarm Optimization (PSO), can efficiently deal with the new challenges compared to conventional optimization techniques. This thesis presents applications of discrete PSO in two specific environments. The first one is for day-ahead optimal scheduling of the reconfigurable gird with distributed energy resources. The second one is a two-step method for rapid reconfiguration of shipboard power system. Effective techniques, such as graph theory, optimal power flow and heuristic mutation, are employed to make the PSO algorithm more suitable to application environments and achieve better performance.

No description available.

Hazy images are often subject to color distortion, blurring and other visible quality degradation. Some existing CNN-based methods have shown great performance on removing the homogeneous haze, but they are not robust in the non-homogeneous case. The reason is twofold. Firstly, due to the complicated haze distribution, texture details are easy to get lost during the dehazing process. Secondly, since the training pairs are hard to be collected, training on limited data can easily lead to the over-fitting problem. To tackle these two issues, we introduce a novel dehazing network using the 2D discrete wavelet transform, namely DW-GAN. Specifically, we propose a two-branch network to deal with the aforementioned problems. By utilizing the wavelet transform in the DWT branch, our proposed method can retain more high-frequency information in feature maps. To prevent over-fitting, ImageNet pre-trained Res2Net is adopted in the knowledge adaptation branch. Owing to the robust feature representations of ImageNet pre-training, the generalization ability of our network is improved dramatically. Finally, a patch-based discriminator is used to reduce artifacts of the restored images. Extensive experimental results demonstrate that the proposed method outperforms the state-of-the-art quantitatively and qualitatively. / Thesis / Master of Applied Science (MASc)

Igneous and metamorphic rock units have long been considered marginal aquifers yet they are a significant source for potable drinking water in many areas worldwide. Additionally, use of these systems is on the rise due to many factors including, contamination and overuse of surficial systems, as well as expanding population and drought. The Nashoba Terrane is a fault-bounded block of high-grade, steeply dipping metavolcanic and metasedimentary rock located in eastern Massachusetts, U.S.A. The Nashoba is northeast trending, extending from Oxford, MA to the Gulf of Maine south of Newburyport, MA. Seventeen previously drilled wells throughout the Nashoba were selected for use in this study. The goal of this study was to characterize the hydrogeologic system of the Nashoba Terrane. Wells studied were in three bedrock types: granite, schist and amphibolite. Three fracture types were identified: FPF, subhorizontal unloading joints and tectonic joints. Several major fracture orientation sets were also identified including northeast trending FPF, east-west trending and north-south trending tectonic joints as well as northwest trending tectonic joints. Dominant sets varied in the three rock types and the frequency of fractures was found to decrease with depth. Only four percent of all fractures measured in this study were flowing. Approximately 32% of the flowing fractures were northeast trending, 17% of subhorizontal fractures were flowing and the remaining 51% were of variable orientation and dip. In general, the orientation of fractures was not found to determine whether a fracture flows, nor was rock type a significant determinant of flow. There was no flow identified below 170 meters and the majority of flow in the Nashoba Terrane is constrained to the upper 100 meters. This is most likely due to decreased fracture frequency and permeability with depth. This study is significant to the search for a sustainable groundwater source in bedrock because results show that the few fractures are actually contributing to flow and that flow is primarily occurring near the surface.

The implementation of a fast track section is a commonly used strategy to improve patient flow in emergency departments (EDs). A fast track section reserves resources (beds, doctors and nurses) for lower acuity patients, and is thus aimed to reduce the wait time and length of stay of these patients. We use a discrete event simulation to investigate the impact of adding a fast track section to an emergency department. We quantify the effect of introducing a fast track on length of stay and waiting time to bed for low and high acuity patients in a crowded ED and compare it to an ED without fast track (Combined ED). We simulate a crowded ED by increasing the patient arrival rate, changing the acuity mix and increasing the time taken for admitted patients in the ED to obtain an inpatient bed (boarding time). We demonstrate that, when compared to a Combined ED with the same number of resources, the introduction of a fast track reduces the wait time to bed for lower acuity patients. However, this comes at the cost of increased waiting time for some higher acuity patients, which is unacceptable in practice. In investigating the solutions to this problem, we find that changing patient prioritization is the most effective way of reducing wait times under crowding. This change in priority does not require the addition of beds, doctors and nurses, and is therefore a cost-effective approach. Finally, we discuss the implications of our results for emergency departments.

Cementation is known to significantly influence the mechanical and hydrologic properties of granular porous media by increasing the stiffness of the elastic response to stress and reducing permeability. The relationship between the changes in cementation and changes in permeability are well documented in literature. However, limited quantitative data exists on the relationship between changes in the amount of cementation and changes in the mechanical response of granular media. The goal of this research is to quantify the effects of cementation on the mechanical properties of granular porous media at the meso-scale and investigate the influence of the competing roles of mechanical and hydrologic properties on fluid flow and deformation at the macro-scale. To accomplish this goal, we developed a multiple scale approach that utilizes the parameterization control of meso-scale Discrete Element Method (DEM) models and the ability to couple fluid flow and solid deformation physics with macro-scale poro-elasticity models. At the meso-scale, a series of DEM models are designed to simulate biaxial tests of variably cemented sandstone in order to investigate the effects of cementation on the elastic and inelastic response of the porous media. The amount of cementation in the DEM model is quantified using a bond to grain ratio (BGR). The BGR is the number of bonds (the bonds represent the cement) divided by the number of grains in each model. The BGRs of the DEM models correlate to BGRs of natural samples and allow constraint of the percent cementation in the DEM models. A decrease in BGR from 2.25 to 1.00 results in a two fold decrease in shear modulus. The resulting shear moduli from the DEM models are used as input properties into two dimensional, axial symmetric poroelastic models of an isotropic confined aquifer. The poroelastic models address the implications of changes in mechanical properties and hydrologic properties on large scale fluid removal and deformation as well as address the importance of the competing roles of hydrologic and mechanical properties.

No description available.

No description available.