Browse
Recent Submissions
Item Open Access A theory of spatial granular computing(Faculty of Graduate Studies and Research, University of Regina, 2023-04) Zhao, Liquan; Yao, Yiyu; Hepting, Daryl; Yao, JingTao; Jia, Na; Kinsner, WitoldThis thesis proposes a theory of spatial granular computing. A granule is composed of one or more atomic granules, and the rationality of this definition is expounded from four aspects: simplicity, applicability, measurability, and visualization. A projection is introduced, which is a one-to-one mapping from granules to points in an nD unit hypercube. Based on this, some basic questions of spatial granular computing are discussed. When discussing the rough set models from the perspective of micro knowledge space, all the models can be easily applied to incomplete information systems and continuous information systems as well. For any equivalence granule in the macro knowledge space, its spatial rough granule model is introduced and the induced three regions are obtained from each of the three planes. Expansions of the spatial rough granule model are examined in three aspects of binary relation, domain and approximation space. Not only are (fuzzy) tolerance relation, (fuzzy) equivalence relation, and coarse-fine relation but also all operations and models defined by relation matrices, and the properties can be easily proved. By focusing on relation matrices and coarse-fine relation, the reduction of knowledge is strictly related to the general concept of independence, so we can use the bases of micro and macro knowledge spaces for granular analysis, which can successfully solve the bottleneck of computing reducts. The measures are divided into granularity and fineness. All measure functions that satisfy nonnegativity, monotonicity and norm invarant can be generalized to measure granules. Subsethood holds twelve axioms, and it can be generalized and extended to conditional granularity (entropies) and conditional fineness (entropies) which are used to measure the coarse-fine relation between two granules and satisfy twelve corresponding axioms. Different definitions of similarity and difference may have different axioms which can be divided into three aspects: symmetry, boundary conditions and monotonicity. Nonnegativity can be replaced by boundary conditions, and any one of the kinds of triangle inequalities can be used as monotonic condition. Structural problems have always been a major challenge in computational science. This thesis explores the multidimensional structures from the perspective of granular computing. Any nD object can be viewed from different dimensions, and it corresponds to an adjacency relation matrix in each view. By discussing the relation between geometry and topology, granular topology algorithm is proposed and it is the general process of granular analysis that all algorithms should follow. By projecting a hypergraph to a hyperedge project graph, all adjacency relations between vertices of the hypergraph are naturally revealed by the meet of the relation vectors corresponding the vertices of the project graph. In this way, it can not only largely reduce the dimension of tensor but also be used to represent any (directed) hypergraph. Two kinds of graph/digraph information systems are defined, each graph/digraph information system can produce one micro and one macro knowledge spaces in which micro and macro rough graph/digraph models are built on its vertices and its edges respectively. This thesis introduces how to obtain the vertex and edge adjacency relation matrices of a quotient graph/digraph, performs structural analysis for fuzzy granules based on their corresponding fuzzy relation matrices respectively as well as for fuzzy graphs/digraphs based on their corresponding vertex and edge adjacency relation matrices respectively, and generalize all measure methods and operations for (equivalence) granules to (equivalence) graphs/digraphs.Item Open Access Ensemble-based framework and its applications to characterizing growth and propagation dynamics of wormhole networks and steam chambers in a heavy oil reservoir(Faculty of Graduate Studies and Research, University of Regina, 2023-01) Yang, Shikai; Yang, Daoyong; Torabi, Farshid; Jin, Yee-chung; Qing, Hairuo; Yang, ChaodongAs for the enormous heavy oil resources all over the world, cold heavy oil production with sand (CHOPS) technique has been successfully utilized to exploit them as the primary recovery due to the induced high-permeability channels (i.e., wormholes) together with foamy oil flow, while the steam-assisted gravity drainage (SAGD) process has been widely used as an efficient enhanced oil recovery (EOR) method because of the steam chamber propagation together with viscosity reduction and gravity drainage. Numerous efforts have been made to characterize the growth and propagation dynamics of wormhole networks and steam chambers; however, it is still a challenging task to reproduce the experimental measurements and field observations due to the physical constraints and strong nonlinearity. Therefore, it is of fundamental and practical significance to develop robust and versatile techniques for characterizing the growth and propagation dynamics of wormhole networks in a CHOPS process and a steam chamber in a SAGD process, respectively. A modified pressure-gradient-based (PGB) sand failure criterion has been developed for quantifying the dynamical and preferential wormhole propagation during a CHOPS process. By upscaling it from pore-scale to grid-scale, such a sand failure criterion is successfully validated with a laboratory experiment and then extended to a field case. Subsequently, a damped iterative-ensemble-Kalman-filter (IEnKF) technique has been developed for dynamically and implicitly determining the model uncertainties during a CHOPS simulation, while an improved upscaling process has been proposed to improve the versatility and accuracy of the PGB sand failure criterion. Not only does such an ensemble-based data assimilating algorithm demonstrate a satisfactory convergence, but also the simulated wormhole network shows its great similarity to the experimental observation. Moreover, the impacts of foamy oil flow, sand failure, and slurry flow are comprehensively evaluated by employing various sets of multiphase relative permeability for the simulation. By integrating the modified PGB sand failure criterion, foamy oil model, multiphase relative permeability model, and damped-IEnKF algorithm into a reservoir simulator (i.e., CMG STARS), a unified, efficient, and consistent ensemble-based framework has been established to characterize the field-scale wormhole growth and propagation dynamics with good accuracy and efficiency, demonstrating its practicality and robustness in simulating the CHOPS processes in heavy oil reservoirs. A heat-penetration criterion has then been newly developed and integrated into a reservoir simulator (i.e., CMG STARS) for characterizing the steam chamber growth and propagation dynamics during a SAGD process, while thermally sensitive co/countercurrent relative permeabilities are incorporated in the simulation. Temperature distribution ahead of a steam chamber interface (SCI) is dynamically calibrated by such a heatpenetration criterion, providing a simple and convenient way to pragmatically, dynamically, and accurately quantify heat transfer during a SAGD simulation.Item Open Access Improvement of weakly compressible moving particle semi-implicit method and its application(Faculty of Graduate Studies and Research, University of Regina, 2023-01) Xiao, Huiwen; Jin, Yee-chung; Wu, Peng; Zhao, Gang; Yao, Yiyu; Li, SamuelMoving Particle semi-implicit method (MPS) is one of the well-known Lagrange particle methods that is advantageous in addressing complex hydraulic engineering problems. Although the accuracy and reliability can be addressed in particle methods, instability arises due to pressure fluctuations from particle clustering and improvements are beneficial for method development. Particle distribution is an area that requires improvement. In this study, a particle smoothing algorithm was developed and incorporated to the weakly compressible MPS, namely sWCMPS. From the definition and derivation of basic MPS operators, uniform particle distribution is critically important to numerical accuracy. Within the framework of sWCMPS, numerical operators were modified by implementing coordinate transfer and smoothing algorithm. Modification of the numerical operators showed significant improvements resulting in the reduction of particle clustering, smoother pressure distributions and pressure oscillations. To validate the numerical feasibility, several numerical cases were simulated to compare sWCMPS to the weakly compressible MPS (WCMPS), including a pre-defined two-dimensional (2-D) analytical function, Poiseuille’s flow, Taylor Green vortex and a dam break. Simulation results showed a reduction of the error caused by irregular particle distribution with lower particle clustering and smaller pressure oscillation. In addition, the newly developed sWCMPS algorithm can achieve numerical accuracy even by applying a larger courant number, indicating a better computational efficiency. To improve computational efficiency with a simple algorithm, explicit algorithms have been proposed such as the WCMPS method. Although accuracy can be fulfilled in WCMPS, instability arises due to pressure fluctuations of the moving particles. In this study, an explicit Poisson Pressure Equation (PPE) was introduced to address particle instability by improving the pressure calculation, referred to as the Weakly Compressible Moving Particle Explicit (WCMPE) method. By predicting pressure through the equation of state for WCMPS, divergence-free condition is conserved without the implementation of small time-steps. Several numerical cases were investigated by the WCMPE method, including 2D jet impinging on a solid plate, oscillating drop, lid driven cavity, rolling tank experiment, sloshing tank simulation and jet impinging from the bottom of a tank. In general, simulation results showed smooth pressure distribution and improved particle stability were obtained with the incorporation of WCMPE. WCMPE method improved the pressure calculation and stability of WCMPS while preserving its coding system with high computational efficiency. The WCMPE method is then extended to multiphase flow simulation. In this study, the multiphase WCMPE method was developed with a simple density smooth technique and interface detection procedure. Several numerical cases were investigated by the multiphase WCMPE method, including Rayleigh-Taylor (R-T) instability cases with various density and viscosity ratio conditions, oil spill and oil-water sloshing experiments. Numerical simulation results showed good agreements with analytical solutions. Experimental data obtained in multiphase WCMPE results also exhibited negligible unphysical pressure oscillation and pressure noise points.Item Open Access Experimental and numerical investigations on wind characteristics and wind induced vibrations of bridge structures(Faculty of Graduate Studies and Research, University of Regina, 2022-12) Xia, Dandan; Dai, Liming; Henni, Amr; Mehrandezh, Menrand; Morgan; Mobed, Nader; Li, XianguoAs more and more flexible structures such as long span bridges and suspension bridges are built in the world, their increased flexibility can cause serious concerns for researchers and engineers. For such structures, wind load has become one of the most important dominant loads under consideration in the analysis and design of structures. Thus, accurate and reliable evaluations of the wind characteristics are critical, as the evaluations may provide a solid foundation for which the design of wind-resistant structures can be relied upon. Therefore, according to the importance and sensitivity of this subject in some special applications, the current research has been presented both analytically and experimentally, which may be utilized as guidance for researchers and engineers around the world in this field. The wind field measurement system used for this research acquires the high frequency wind speed data for the process of tropical cyclone and monsoon wind. Wind characteristics are studied systematically with wind data collected. Furthermore, the comparative study of stationary and non-stationary models, which is established with a self-adaptive procedure, is conducted. The established non-stationary model presents advantages for measuring the time related wind speed variations. Analytical results obtained in the research expose the difference between the traditional stationary and non-stationary models cannot be ignored, especially over a large time horizon. Moreover, in order to predict the wind speed with higher accuracy, a more accurate prediction method based on wavelet decomposition and chaotic diagnosis is also proposed. The wind speed prediction method considers the frequency domain characteristics of wind speed series for improving prediction accuracy. Wind speed data collected from the long term experiments and downloaded historical data are applied to verify the accuracy and reliability of the proposed prediction method. Wind induced vibration is another aspect which will affect the safety of the structures. As a typical and commonly seen type of wind induced vibration, vortex induced vibration (VIV) of bridges and the aerodynamic influences of the bridge attachments on the VIV are studied experimentally with a wind tunnel test. The effect of the attachments installed on the bridge deck such as crash barriers, wind barriers and traffic flows on VIV behaviors including lock-in wind speeds and vibration characteristics are systematically investigated. A new method by combing the unscented Kalman Filter with unknown input and simplification of Tylor series expression for studying the loads generated by vortex and VIV is proposed in the research. The governing equation developed is simplified by Taylor expression which can be applied for general bridge decks, unscented Kalman Filter with unknown input (UKF-UI) method is utilized to identify the self-excited and aerodynamic forces in VIV lock in regions. Instead of using fitted mathematical model, such a method can be directly applied on wind tunnel test data of general bridge deck types. The research findings are valuable for better analyzing and comprehending wind characteristics as well as interactions between wind and structures, and are expected to provide a practically sound guidance for designing wind-resistant structures especially flexible structures subject to wind exertion.Item Open Access Development of high-resolution climate projections over Canada in the 21st century(Faculty of Graduate Studies and Research, University of Regina, 2023-03) Wu, Yinghui; Huang, Gordon; Young, Stephanie; Zhu, Hua; Deng, Dianliang; Nasiri, FuzhanIn this research, high-resolution climate projections over Canada have been developed through the WRF model. The spatial and temporal variations of Canada’s temperature, precipitation, and precipitation extremes in the 21st century have been comprehensively analyzed. Potential mechanisms in terms of temperature, precipitation, and precipitation extremes over Canada have been investigated, and effects of anthropogenic warming on these variables have been revealed. In addition, permafrost degradation under climate change is extended to a global scale, and a dynamic longterm classification scheme for discontinuous permafrost is proposed. The annual mean temperature over Canada is projected to increase by [1.53, 1.98], [2.51, 3.86], and [2.94, 6.19]°C in the 2030s, the 2050s, and the 2080s under RCP4.5 and RCP8.5 respectively, with largest increase in winter. The annual total precipitation is projected to increase by [16.33, 68.96], [64.80, 121.62], and [123.62, 184.33] mm in three future periods. It is found that the westerlies have significant impacts on Canada’s precipitation variations, and extreme precipitation frequency are related to ENSO phases. A dynamic SOM-based permafrost classification scheme (DSOMPCS) is developed to provide robust, objective, and detailed classifications for discontinuous permafrost. Linear relationship is found between global temperature increase and permafrost loss. When global temperature increases by 1°C, there will be a 2.5 million km2 of permafrost loss. This research will help provide insights into effects of anthropogenic warming on multiple aspects of climatology. The results can provide valuable information for mitigation and adaptation of climatic changes in a Canadian and global context.Item Open Access Walking and thinking: Critical reflections on “Walking the Bypass: A Meditation on Place”(Faculty of Graduate Studies and Research, University of Regina, 2022-09) Wilson, Kenneth Clayton; Farrell-Racette, Sherry; Dashchuk, James; Bracht, Kathryn; Irwin, Kathleen; Archibald-Barber, Jesse; Donald, Dwayneôma kihci-kiskinwahamasinahikan âpacihtâmakan êsa kwayask itahkamikisiwina taitwêstamâkêmakahk ôma pimohtêwin-êkwa-masinahikêwin kiskinwahamawi-osîhcikâsowin kihci-kiskinwahamasinahikan “ê-pimohtêhk âsokanihk: ê-mâmitonêyihcikêhk ôma itâpatakêyimisowin,” ê-itwêmakahk kwayask âcimowin ôma pêyakwan wîhowin ohci. Nîkân masinahikanêkinohk atamiskawikow ôma atoskêwin, tânisi ê-pimakotêk, êkwa mîna kakwêcihkêmowina: “pimohtêwin cî ka-kî-mêskwacipayin pakwâcaskiy isi ôma kihci-askiy?” êkwa, misawâc ê-kihcêyihtakwâhk, “môniyâwak cî ka-kî-nisitohtamwak ôma ôta askîhk ê isipimohtêcik?”. nîso masinahikanêkinohk pîkiskwêmakan tânisi kâ-pimohtêhk naspasinahikân ôma ê-isi-âniskôstêk isi ôma atoskêwinihk; nôkohcikêmakan nanâtohk âcimowina kâmiyonâkwahk êkwa kâ-isîhcikêhk isi-pimohtêwin êsa ê-kikinikâtêk ôma atoskêwin. Nisto masinahikanêkinohk pîkiskwêmakan tânisi kâ-itahkamikisihk kwayask, tâpiskôc kwayask itahkamikisiwina êwako askiy êkwa kâ-nisitohtamâhk; iyiniwi-kiskêyihtamowin ôma ê-miyoitôtâmiyit askiy êkwa atâhko-kiskinwahamâtowin; êkwa tânisi kâ-nisitohtâsoyahk êsa askiy ohci, ê-mêskwacipayihk itahkamikisiwina, êkwa tânisi kâ-itâpisihkik iyiniwak êwako anima kiskêyihtamowina. nêwo masinahikanêkinohk pîkiskwêmakan kâ-âpacihtâhk êsa atoskêwinihk: masinahikêwina ê-masinahikêhk Willam Least Heat-Moon, Nick Papadimitriou, êkwa Iain Sinclair ohci; tânisi kâ-wîkiyâhk ê-ispayihikoyâhk; pêyako-pimohtêwin; kwayask âcimowina êmiywâsik; êkwa masinâpiskahikêwin. piyisk, iskwêyânihk pîkiskwêmakan tânisi kâ-isiitôtamâhk ê-atoskêyâhk ôma kihci-kiskinwahamasinahikanihk—êkwa nika-kî-itôtênân, kîspin êsa kihci-âhkosiwin êkâ ê-kî-âyimahk ta-atoskêmitoyâhk ôma nîso askîwin aspin ohci. kiskêyihtâkwan kahkiyaw ê-nêstosiyâhk êwako kihci-âhkosiwin, kistêyihtâkwan ta- nisitohtamâhk ôma misawâc ê-miywâsik êtikwê êwako atoskêwinihk, kâ-kî-mêskwacipayihk ayisk kihci-âhkosiwin COVID-19 ê-astêk ôma ôta. This thesis presents a critical, theoretical, and methodological exegesis of the walking-and-writing research-creation doctoral project “Walking the Bypass: A Meditation on Place,” as represented by the creative-nonfiction manuscript of the same title. The first chapter introduces the project, its aims and scope, and its research questions: “can walking turn non-places into places?” and, more importantly, “can settlers come into a relationship with the land through walking?” The second chapter discusses the context of walking art as it applies to this project; it presents accounts of a range of aesthetic and political walking practices that influenced this project. The third chapter presents the project’s theoretical context, including theories of place and space; Indigenous theories of land and cosmology; and object-oriented ontology, affect theory, and Indigenous critiques of those ways of thinking. The fourth chapter explores the project’s methodological touchstones: writing by William Least Heat-Moon, Nick Papadimitriou, and Iain Sinclair; psychogeography; solo walking practices; creative nonfiction; and photography. Finally, the conclusion suggests future directions which this research could take—and would have taken, had the pandemic not made working with others so difficult during the past two years. As much as everyone is tired of the pandemic, it’s important to acknowledge that it may have been the greatest influence on this project, which changed significantly in response to the challenges presented by Covid-19.Item Open Access Determination of evaporative flux from water and soil surfaces(Faculty of Graduate Studies and Research, University of Regina, 2023-02) Suchan, Jared Joseph; Azam, Shahid; Xue, Jinkai; Azadbakht, Saman; Chi, Guoxiang; Gurrapu, SunilThe Canadian Prairies has the highest water demand-to-availability ratio in Canada. The region is characterized by a scarcity of water resources because evaporation generally exceed precipitation. Evaporation is a complex phenomenon based on interactions between meteorological and physiological factors. The purpose of this research was to accurately determine evaporative flux from water and soil surfaces. The main research contributions of this research are summarized as follows: (i) a Bench-scale Atmospheric Simulator (BAS/BAS2) was developed to control artificially generated parameters for short time periods; (ii) a Controlled Photogrammetry System (CPS) using Structure-from-Motion (SfM) technology was developed for accurate and non-destructive measurement of dimensional changes during evaporation; (iii) high-quality evaporation datasets were developed for calibration of devices, validation of prediction equations and evaluation of theoretical frameworks; (iv) predictive models for evaporation from water were evaluated to select appropriate empirical equations for predicting potential evaporative flux from water; and (v) a theoretical framework for evaporation from soils was developed to correlate with soil behavior (water retention and soil shrinkage) for both fresh water and saline water. The findings of this research are useful for developing methods to minimize evaporation and for helping with devising water use policy.Item Open Access MPS modeling of cross-sectional averaged shallow-water equations: Development and application(Faculty of Graduate Studies and Research, University of Regina, 2022-12) Sarkhosh, Payam; Jin, Yee-chung; Xue, Jinkai; Yang, Daoyong; Fan, Lisa; Zhang, WenmingIn the last two decades, mesh-free or particle methods have been widely applied to solve shallow water equations. This research presents a moving particle simulation (MPS) scheme to spatially integrate the cross-sectional average shallow water equations, called 1D MPS-SWE solver. First, a first-order accurate numerical scheme in space and time is developed, applicable to prismatic open channels with closed boundaries. A density-ratio formula is derived for computing the cross-sectional area using the Newton-Raphson iteration and based on the particle number density concept. The newly derived formula does not miscompute the cross-sectional area if an incorrect searching radius parameter is selected, unlike the typical volume-summation equation in mesh-free shallow water flows. Dynamic Stabilization is utilized to capture shockwave problems, a case-independent procedure based on the inelastic collision with unequal masses. The solver is then improved to a second-order accuracy in space and time applicable for non-prismatic open channels with open boundaries. For this purpose, the well-known MacCormack predictor-corrector scheme in the mesh-based schemes is extended to the Lagrangian framework. Furthermore, upwind and downwind MPS gradient operators are introduced. Second-order schemes usually suffer from unphysical oscillations. To solve this numerical issue, the total variation diminishing (TVD) flux-limiter scheme is developed for the MPS formulation to preserve monotonicity, which does not require characteristic speeds of the system. In mesh-based methods, the non-conservative form of the momentum equation is problematic in scenarios with discontinuities, particularly in higher-order schemes. However, the non-conservative convective flux term does not exist under the Lagrangian framework in preserving momentum conservation. Since the algebraic density-ratio formula computes the cross-sectional area in the present work, the open boundary treatment requires less computational effort than mesh-based methods because the partial differential continuity equation is omitted as a boundary value problem. The 1D MPS-SWE solver satisfies positivity-preserving and well-balancing properties without the need for any threshold value for treating the dry bed and friction term. Additionally, it can model free-surface flows in prismatic and non-prismatic open channels with various open and/or closed boundary conditions. Later, a Picard iteration-based framework to solve the density-ratio formula is presented. Compared to the conventional Newton-Raphson iteration, the new strategy is advantageous as the iterative convergence is strongly accomplished with accuracy up to double-machine precision. Notably, the results of the MPS-SWE solver are inconsistent with the experimental data at the early dam-break stages subject to the non-hydrostatic pressure effect. To enhance the accuracy of the solver to mimic frictional dam-break flows, a hybrid solver of 1D MPS-SWE and a 2D MPS large-eddy simulation accompanied by a sub-particle scale is introduced. The configuration of sub-domains is established so that the 1D solves the regions with dominantly hydrostatic pressure, while the 2D solver is accounted for the non-hydrostaticity. A moving coupling system is proposed based on two-layer shallow water equations with identical densities that construct two sub-domains. Then, the region with highly vertical acceleration is substituted by the 2D model. Also, two different section and layer coupling systems are developed in which the overlapping zone and boundary interface move with time. As regards the second, a new method is presented to deal with irregular boundary interface while satisfying the momentum transfer. The model validation and application show that the present hybrid solver performs well and considerably decreases the computational time from 2D models.Item Open Access A novel scaling approach for micro-optical analysis of nanoparticle-stabilized CO2 foam flooding for enhanced heavy oil recovery(Faculty of Graduate Studies and Research, University of Regina, 2023-01) Rahman, Arifur; Shirif, Ezeddin; Torabi, Farshid; Muthu, Jacob; Mobed, Nader; Henni, Amr; Azaiez, JalelHeavy oil reservoirs are considered one of the prime hydrocarbon resources for the upcoming decades in Canada. However, more technological and effective methods are needed to overcome the economic and environmental limitations of existing sole surfactants, polymers, nanoparticles, and CO2 flooding recovery methods. Laboratory experiments of these combined methods yields promising outcomes, but cost-effective manners and optimal operating conditions still need to be determined. In this dissertation, mathematical and experimental investigations of nanoparticle-stabilized CO2 foam flooding were performed to enhance heavy oil recovery. CO2 injection is one of the most promising techniques that can enhance heavy oil recovery by reducing oil viscosity and interfacial tension. However, it poses a variety of problems, including low volumetric sweep efficiency due to gravity overriding, and gas fingering. Several methods have been attempted to solve this issue. Foam flooding is one of the most promising strategies for overcoming this issue. However, several challenges are also associated with foam generation and stabilization in reservoir conditions. The addition of nanoparticles with surfactants can overcome this weakness and generate stronger and more stable CO2 foams. In this research, a series of bulk and dynamic foam generation and stability tests were performed with various concentrations of surfactants with nanoparticles. The foam formation and stability improved by 23% and 17%, respectively, when nanoparticles were used in foamability tests in the bulk and dynamic phases. The interaction mechanisms between rock-fluid and fluid-fluid and their effect on oil recovery were also investigated by FTIR spectroscopy, and SARA analysis along with a series of direct visualized micromodel tests. Nanoparticle-stabilized CO2 foam flooding is an effective technique that can enhance heavy oil recovery. However, due to their physical constraints, replicating the genuine behavior of a reservoir on a field scale is unfeasible. As a result, developing scaling criteria for representing actual fluid behavior in heavy oil reservoirs is crucial. In this dissertation, a series of connections were developed between two configurations and the outcome of the lab-scale model was used to foresee the behavior of a field-scale process. In this study, a mathematical scaled model was developed using governing equations with their initial and boundary conditions by incorporating a novel concept. The effects of these developed dimensionless numbers on oil recovery were studied. The results indicate that capillary numbers, gravity numbers, mobility ratios, gravity to fluid movement ratios and Peclet numbers can significantly affect oil recovery. In addition, the effects of silicon dioxide NPs and the SDS surfactant on the rheological characteristics and composition of extra-heavy crude oil were investigated in this study. A series of sandpacked/micromodel tests have been conducted to visualize and estimate the process controlling parameter’s effects on oil recovery. To evaluate the lab scale results to the field scale process, a synthetic reservoir study was compared with a small-scale model to investigate different process controlling parameters that affect oil recovery. Finally, a correlation was developed between lab scale results to field scale process. Mathematical and experimental studies were combined to find efficient and effective nanoparticle stabilized CO2 foam flooding for enhanced heavy oil recovery (EHOR). These research findings will have practical benefits in selecting EHOR schemes and will lead to new opportunities for improving current practices and implementing techniques that have not been previously considered.Item Open Access Effects of high permeability channel parameters on VAPEX performance: A numerical and experimental approach utilizing the reservoir-on-the-chip(Faculty of Graduate Studies and Research, University of Regina, 2023-01) Rahimbakhsh, Aria; Torabi, Farshid; Muthu, Jacob; Zeng, Fanhua; Mobed, Nader; Nouri, AlirezaNumerous heavy oil and bitumen reservoirs exist in Canada. Regarding the ever-growing energy demands and the economic aspects and the rising environmental concerns, these underground resources need to be produced from through advanced, environmentally friendly, and economically viable methods. The implementation of such techniques can secure the future of the heavy oil industry. Despite resulting in lower production rates, the solvent-based methods hold major advantages over the thermal recovery methods in many other areas such as those mentioned. The present research carries out an in-detail study of the VAPEX process, as one of the most recent solvent-based heavy oil recovery techniques, in the reservoirs with high permeability channels to evaluate the effect of high permeability channel parameters on the process performance. Heavy oil recovery processes, in this case the VAPEX, have been rarely studied, which signifies the reason why this process has not been fully commercialized in the industry. To achieve this purpose, several patterns with high permeability channels with distinct features were designed and engraved on glass pieces to manufacture state-of-the-art microfluidic models mimicking a typical Canadian heavy oil reservoir. Two heavy oil samples were utilized (around 1500 and 3000 cP) during the conducted experiments as well as the three injected solvents of pure propane, pure carbon dioxide, and the propane-carbon dioxide mixture. Separate sets of experiments were conducted to measure the pure and mixture solvents solubilities in both heavy oil samples at various pressures below their corresponding vapour pressures, and room temperature. This allowed for tuning the PVT data in the simulator and then, obtaining a history-matched model capable of simulating the injection- production scenarios that could not be experimentally tested due to the high number of cases. A thorough image analysis operation was carried out over the experimental models to determine the heavy oil produced, residual oil saturation, ultimate recovery factors, and monitor solvent chamber expansion. Together with the simulation outcomes, a comprehensive data bank was obtained from the 90 scenarios designed and 54 VAPEX experiments conducted. The effects of high permeability channels orientation, length, width, intensity, and position on the process performance were identified and numerically evaluated. Moreover, in terms of the solvents used, propane proved the most efficient owing primarily to its and high solubility in both heavy oil samples. It was observed that all high permeability channels, regardless of their properties, enhanced the heavy oil recovery in comparison to the base case (no high permeability channels) scenario. The highest recovery factors were obtained by positioning two long horizontal high permeability channels between the injection-production well pair and injecting propane. Almost equal to that was the process efficiency when implementing 2 wide vertical high permeability channels on either side of the well pair. The obtained results from the present work, combined with data from the literature will serve to develop a predictive model for the VAPEX recovery factor evaluation using an artificial neural network technique.Item Open Access Efficient estimation for time series models(Faculty of Graduate Studies and Research, University of Regina, 2023-01) Pandher, Sharandeep Singh; Hossain, Shakhawat; Volodin, Andrei; Deng, Dianliang; Sardali, Arzu; Yao, Yiyu; Triacca, UmbertoIn this dissertation, we discuss the pretest, improved pretest and shrinkage estimators for some time series regression models, such as ARIMA, GARMA, and GARFIMA. We study the asymptotic biases and risks of these estimators and compare their relative performance with respect to the maximum likelihood estimator (MLE) analytically and numerically using Monte Carlo experiments and real data examples. In Chapter 2, we consider the James-Stein shrinkage estimation procedure for the ARIMA regression model when some prior/auxilliary information is available about insignificant covariates in the model. The asymptotic properties of the shrinkage estimators, under local alternatives, are implemented including the derivations of the asymptotic distributional bias and the asymptotic quadratic risk. These results showed the effectiveness of the suggested estimation technique. Monte Carlo experiments were conducted to demonstrate the superiority of the shrinkage estimators over the MLE. In Chapter 3, we propose the pretest and shrinkage approaches in estimating the regression parameters of the generalized autoregressive moving average (GARMA) model, which are pervasive for modelling binary and count time series data. We want to estimate regression and ARMA parameters when some of the regression parameters may belong to a subspace. We establish the asymptotic distributional biases and risks of the proposed estimators and evaluate their relative performance with respect to the unrestricted maximum partial likelihood estimator. The performance of the proposed estimators is investigated using simulation studies and real-life applications. Chapter 4 focuses on pretest and shrinkage estimation strategies for generalized autoregressive fractionally integrated moving average (GARFIMA) models when some of the regression parameters can be restricted to a subspace. These estimation strategies are constructed on the assumption that some covariates are not statistically significant for the response. We enlighten the statistical properties of the shrinkage and pretest estimators in terms of asymptotic bias and risk. We show that the shrinkage estimators have a lower relative mean squared error as compared to the UMPLE, when the number of significant covariates exceeds two. Monte Carlo simulations are conducted to examine the relative performance of the proposed estimators to the UMPLE. An empirical application is used for the usefulness of our proposed estimation strategies Finally, we summarize the findings of the thesis in Chapter 5. Also, some problems for future research are outlined in Chapter 5.Item Open Access Cross product ratio under different sampling schemes and zero truncated negative binomial weighted Weibull distribution with applications(Faculty of Graduate Studies and Research, University of Regina, 2022-11) Nadeem, Hira; Volodin, Andrei; Ahmed, Syed Ejaz; Deng, Dianliang; Sardali, Arzu; Feng, CindyA successful clinical trial requires efficient strategies for enrolling and retaining the study participants. In fact, Global data research suggests that participant enrollment issues are one of the major problems for the clinical trial termination. This thesis aims to compute the point estimators for the cross-product ratio ρ from two independent Bernoulli samples and determine the best sampling scheme. The idea of using Bernoulli samples to conduct clinical trials has been very successful since it is cost effective. Bias and Mean Squared Errors are used as the principal criteria. According to the simulation results, the special case of Direct-Inverse Binomial sampling scheme works the best, where the number of successes in the Direct sampling scheme is used in the second sampling scheme of the Inverse Binomial scheme. Results of Bias and Mean Squared Errors are compiled in the tables for different levels of samples and probabilities. We focus on the idea of enrolling participants using the special case of the Direct- Inverse sampling scheme. Asymptotic confidence intervals for the cross-product ratio ρ are constructed. Closeness of the confidence coefficient to the nominal confidence level is our main evaluation criterion, and we use the Monte-Carlo method to investigate the key probability characteristics of intervals. We present estimations of the coverage probability and interval width in tables. As a real world practical application of the technique, CYP-GUIDES case study is discussed where the standard and genetically guided therapy are compared to determine whether one therapy is more effective than the other. We compute point estimators for the cross-product ratio under the different sample schemes and determine the confidence interval for the special case of Direct-Inverse sampling scheme. In this thesis, we propose two new discrete distributions, the negative binomialweighted Weibull and the zero truncated negative binomial-weighted Weibull distributions. Some statistical properties of the proposed distributions are presented. The parameter estimates for the proposed distributions have been derived by the maximum likelihood estimation. The applications to real data sets are presented in order to compare the performance with other distributions.Item Open Access Spatiotemporal heterogeneity in precipitation and moisture transport over China and their connections with anthropogenic emissions and natural variability(Faculty of Graduate Studies and Research, University of Regina, 2022-09) Lu, Chen; Huang, Gordon; Zhu, Hua; Ng, Kelvin Tsun Wai; Deng, Dianliang; Chen, ZhiIn this dissertation research, the following scientific questions are explored: (i) Has the probability distribution of precipitation over China undergone variations since the midtwentieth century? (ii) How are the above changes related to the modes of climate variability? (iii) Can these changes be attributed to anthropogenic behaviors? (iv) What are the mechanisms for these changes in terms of moisture transport and recycling? Specifically, through quantile regression, the quantile trends in monthly precipitation anomalies over China, as well as the individual and combined quantile effects of teleconnection patterns, are examined. The results show that the quantile trends exhibit apparent seasonal variations, with a greater number of stations showing trends in winter, and larger average magnitudes of trends at nearly all quantile levels in summer. The effects of El Niño–Southern Oscillation (ENSO), North Atlantic Oscillation (NAO), and Pacific Decadal Oscillation (PDO) exhibit evident variations with respect to the quantile level. Spatial clusters are subsequently identified based on the quantile trends, and the individual and combined effects from the teleconnection patterns are further investigated from the perspective of moisture budget. Seven spatial clusters with distinct seasonal quantile trends can be identified; three of them are located in southeastern China and are characterized by increasing trends in summer and winter precipitation. Summer precipitation over this region is positively influenced by ENSO and negatively influenced by NAO, with the former affecting both the dynamic and thermodynamic components of vertically integrated moisture divergence and the latter affecting only the dynamic component. The interaction effect of ENSO and NAO on summer precipitation anomalies in extremely-wetter-than-normal months is statistically significant. The influences of anthropogenic greenhouse gas and aerosol emissions on the probability distribution of daily precipitation over China are explored through a formal detection and attribution analysis. It is found that the increasing trends in winter precipitation at high and extremely high quantile levels, as well as that in spring precipitation at all quantile levels, can be attributed to the effects of historical (ALL) or anthropogenic (ANT) forcing. The effect of anthropogenic greenhouse gas forcing (GHG) is evident over the domain, to which the increasing precipitation trends at all quantile levels in all seasons can be attributed; this effect can be separated from that of anthropogenic aerosol forcing (AER) for winter precipitation trends at high and extremely high quantile levels, and for spring, summer, and autumn trends at low quantile levels. Through integrating detection and attribution analysis and moisture tracking into one framework, the anthropogenic influence on the moisture source-receptor relationship over China is investigated. The subdomains of China can be grouped into 3 categories according to the major moisture sources, which are regions mainly dependent on (a) oceanic sources, (b) external terrestrial sources, and (c) local recycling. It is found that the GHG forcing, in general, favor reduced moisture contributions from oceanic sources to the west of the domain in winter and enhanced those from the ones to the east of the domain in both winter and summer for regions in the first category. Under the GHG scenario, moisture contributions from external terrestrial sources are reduced in summer for subdomains in the first category and show a shift from the north to the south for other regions; under the AER scenario, the opposite case can be observed. Local recycling is generally enhanced in summer under both GHG and AER scenarios, with an exception for the regions in the first category, which exhibit reduced local recycling under the GHG forcing.Item Open Access Statistical models in ecological and health economic data(Faculty of Graduate Studies and Research, University of Regina, 2023-04) Liu, Sichen; Volodin, Andrei; Hall, Britt; Deng, Dianliang; Sardali, Arzu; Feng, CindyDifferent statistical models can be used interdisciplinary areas such as health studies, epidemiology, biology, etc., Traditional models can allow for analysis in different ways. Data on methylmercury (MeHg) concentrations and other environment parameters were collected between 2006 to 2012 from wetlands in the prairie pothole region in Saskatchewan by members of the Department of Biology, University of Regina. Data for health studies were collected from previous researchers (such as Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke from Albers, G.W., et. al.). My research goal was to use statistical models to provide researchers in other fields with some ideas and statistical results, as well as an analysis of uncertainty and variation. With MeHg data, I provide the concentration trend and with health economic data, I provide possible of models for different real cases, in particular some new mixed models and methods to simplify the pure statistical models economist may use. In this thesis, Chapter 2, I continued my MSc’s research and used the part of my MSc’s results to predict. Chapter 3 compared probabilistic analysis and deterministic analysis in three different cases, and Chapter 4 provide analytical proof and propose simplification methods and introduce a conditional probability approach to deal with more than two state transitions in a single cycle of a Markov model. My analysis shows the MeHg concentration in water in Saskatchewan will slowly increase in next few years based on time series analysis in Chapter 2, and suggesting that probabilistic analysis may be associated with greater bias in model inputs and that model output compares to deterministic analysis and present a small extension of the regular usage with a few examples in Chapter 3 and 4.Item Open Access Characterization of fracture networks using tracer tests in a hydrocarbon reservoir(Faculty of Graduate Studies and Research, University of Regina, 2023-03) Liu, JinJu; Yang, Daoyong; Shirif, Ezeddin; Jin, Yee-chung; Qing, Hairuo; Clarkson, Christopher R.In a hydrocarbon reservoir, natural and/or hydraulic fractures can not only provide main paths for fluid flow and increase formation permeability, but also complicate flow behaviour and production performance. Tracer tests have been widely used for reservoir characterization to determine residual oil saturation, reservoir wettability, and fluid flow paths; however, few attempts have been made to characterize complex fracture networks by quantifying tracer transport behaviour due to the associated technical challenges including the composite effects of fluid flow and geomechanical dynamics, fracture geometries, and tracer transport mechanisms. It is, therefore, of fundamental and practical significance to quantitatively identify such fracture networks and accurately evaluate the well/reservoir performance. A pragmatic technique has been developed to describe tracer flowback behaviour for a vertically fractured well by coupling fluid flow and geomechanical dynamics. TheBarton-Bandis model is employed to describe the relationship between effective stress and fracture properties, while tracer flowback profiles are quantified by taking tracer advection, tracer dispersion, and tracer adsorption into account. Effects of the fracture propagation dynamics on the tracer flowback behaviour are then examined and discussed. The tracer recovery factor with considering geomechanics is nearly 20% higher than that without considering geomechanics. An efficient and effective numerical model based on the embedded discrete fracture model (EDFM) is then developed and applied to deal with a fractured reservoir by independently constructing matrix and fracture grids. Subsequently, non-neighbouring connections (NNCs) are employed to couple the flow of fluid and tracer between the non-neighbouring grids. The tracer flowback profiles have been investigated for a fractured horizontal well with four different fracture network patterns including bi-wing fracture network (BWFN), opening-fissure fracture network (OFFN), fractal-like fracture network (FLFN), and mutually orthogonal fracture network (MOFN). It is found that tracer flowback concentration varies greatly with different fracture network patterns. To improve the simulation accuracy and reduce the grid orientation effects associated with fractal-like fracture networks, the EDFM based on the perpendicular bisection (PEBI) grids is then employed so that the generated grids flexibly conform to the complex fractal-like fractures. Furthermore, the tracer flowback profiles, which can reveal the complexity of fractal-like discrete fracture networks, are quantified by considering complex tracer transport mechanisms. The structured grid-based EDFM is further used to characterize the fracture distributions in a naturally fractured reservoir conditioned to interwell tracer transport behaviour. The stochastic fracture modelling approach is implemented to generate the randomly-distributed natural fractures. The response of such an interwell tracer model is found sensitive to the fracture parameters rather than tracer properties. All of the numerical models have been validated and then extended to field cases to characterize the fracture networks. Tracer production concentration profiles are generated and beneficial to examine the effects of different parameters on the tracer transport behaviour in a hydrocarbon reservoir. Through comparison with other reservoir characterization methods (e.g., microseismic monitoring), the accuracy and efficiency of the newly developed models are further confirmed.Item Open Access Genomic and experimental insights into the ecology and evolution of transferable genetic elements(Faculty of Graduate Studies and Research, University of Regina, 2022-08) Lerminiaux, Nicole Anne; Cameron, Andrew; Finlay, Kerri; Yost, Christopher; Zilles, Sandra; Hynes, AlexanderBacteria have dynamic genomes that allow them to adapt and survive almost anywhere on Earth. This genetic flexibility is facilitated by mobile DNA elements, which can transfer within or between genomes independently of cell lineages and drive bacterial evolution. Mobile genetic elements influence many aspects of bacterial life by encoding and transferring antimicrobial resistance genes, pathogenicity factors, and toxin-antitoxin modules. Plasmids, integrons, and genomic islands are several types of mobile DNA elements that can carry beneficial genes which help host cells adapt to new environments or provide new ecological functions. However, much is still unknown about how mobile genetic elements persist in the environment, how they mediate cell survival, and how they impact interactions in bacterial communities. In this thesis, genomic and experimental techniques were used to investigate the ecology and evolution of mobile genetic elements at three different scales: genes, species, and ecosystems. At the gene level, I performed a literature review to determine how the main processes of horizontal gene transfer (conjugation, transduction, and natural transformation) impact antimicrobial resistance gene transmission in clinical environments. Conjugation of antimicrobial resistance gene-encoding plasmids does occur within clinics and patients, but there is less evidence for transduction and transformation. This may be due to low transfer rates or difficulty detecting the transfer of core genes with key mutations that provide resistance. In another project, I used comparative genomics to assess the evolutionary history of the Salmonella pathogenicity island SPI-1, which encodes a type three secretion system for invasion of mammalian intestinal cells and is remarkably conserved across the Salmonella genus. I defined the multiple genomic islands that comprise the mosaic structure of SPI-1, with some islands having arrived in the Salmonella clade earlier than others. Related pathogenicity islands possess homologs to SPI- 1 transcriptional regulator hilA, but are missing homologs to hilD and hilC; the high nucleotide identity between hilD and hilC suggests they may be paralogs. To assess how mobile genetic elements impact communities at the species level, I designed a synthetic multi-species community wherein only Escherichia coli can access a carbon source and Salmonella enterica must rely on cross-feeding to survive while simultaneously killing the E. coli with a plasmid-encoded colicin toxin. Despite relying on cross-feeding to survive, S. enterica consistently emerged as the dominant community member. Experimental results and mathematical modeling confirmed that the colicin liberates nutrients through cell lysis and benefits the colicin-producing population, which had not been previously described for this type of toxin. At the ecosystem level, I attempted to isolate all of the plasmids (“plasmidome”) from surface water of agricultural ponds (dugouts) to determine how the gene content changes over time using long-read DNA sequencing. However, ongoing technical difficulties caused by environmental contaminants resulted in limited DNA sequencing data and interference in enzymatic reactions during library preparation. Nevertheless, additional insights into these ecosystems were possible through 16S rRNA bacterial community profiling and qPCR of select antimicrobial resistance genes and mobile elements. Antimicrobial resistance genes are detectable and persist over time in dugouts, and regression modeling demonstrated that their abundance was explained by the abundance of the mobile integrin integrase intI1 and cattle presence. Altogether, these diverse projects have contributed new knowledge on the transmission, persistence, and ecological impact of mobile genetic elements at multiple scales (gene, species, ecosystem) by using experimental techniques and high-throughput sequencing technologies.Item Open Access Fusions of association schemes(Faculty of Graduate Studies and Research, University of Regina, 2023-01) Joshi, Neha; Herman, Allen; Meagher, Karen; Fallat, Shaun; Floricel, Remus; Fan, Lisa; Sankey, AlyssaSince their introduction as symmetric coherent configurations by Bose and Mesner in 1959, association schemes have gained significant importance in algebraic combinatorics. An important breakthrough was achieved by Delsarte’s PhD thesis where he proved that many problems from coding theory, combinatorial design theory and statistics can be treated using the concept of association schemes [12]. Since its initial introduction, many algebraists and graph theorists have been studying the existence, construction and generalizations of various association schemes [1, 3, 7, 8, 17, 19, 28, 29, 30]. Because of their impressive construction, association schemes are useful to these subjects and there is always a search for new association schemes. One easy way to construct a new association scheme is by taking either direct or wreath products of two existing association schemes. One such example that we studied was given by Sankey in [32]. Another way to construct a new association scheme is by fusing specific relations of an existing association scheme. The resulting association scheme is known as a fusion (previously referred to as “subscheme” by mathematicians) [4]. The focus of this thesis is to examine a few important association schemes and classify them based on their fusions. It can be observed from literature that the nonexistence of nontrivial fusions is a rare phenomenon and this is the motivation behind this thesis. An association scheme, A, is said to be fusion-primitive if there does not exist any nontrivial fusions of A. In 1992, Muzychuk proved that there does not exist any nontrivial fusions of the Johnson scheme, J (n, k) for all n > 3k + 4 with k ≥ 4 [29]. In 1994, this result was further refined for all n > 3k + 1 by Uchida in [36]. In this thesis, we prove that the Johnson scheme does not have any nontrivial fusions for all n, n > 2k + 1 with k ≤ 20. In addition, we classify almost all of the multiplicity-free subgroups of the symmetric group based on their nontrivial fusions. The Hamming scheme, H(n, q) is an important example in coding theory [10]. The fusionprimitivity of the Hamming scheme has been discussed before (see [28]). In this thesis, we study the generalized Hamming scheme, H(n,A) and prove that it is fusion-imprimitive (that is, it always has a nontrivial fusion). We also classify all fusions of the generalized Hamming scheme, H(2,A) where A is the association scheme corresponding to a strongly-regular graph.Item Open Access Evaluating and optimizing the performance of single and blended amines based on their chemical structures for carbon dioxide capture from industrial gas streams(Faculty of Graduate Studies and Research, University of Regina, 2023-03) Fraij, Heba; Idem, Raphael; Ibrahim, Hussameldin; Supap, Teeradet; Mobed, Nader; Suriyapraphadilok, UthaipornWith the well-known fact on negatively impact of the human activities and the tremendously growth of industrial sectors and the increase in the global energy demand, the increase of the Green House Gas (GHG) including CO2 is very central because almost all fossil fuel activities lead to generation of this environmentally harmful GHG which found to cause increasing on the average global temperature which causing several major issues such as extreme weather conditions, heat waves, sea level rise, wild fires, health problems and many more. Actions are required immediately to reduce the emission such as using alternative energy source with less GHG emission and use Carbon Capture and Sequestration (CCS). Post combustion capture by using a liquid absorbent solution is successful method specially the ability to regenerate the solution which makes this method cost sufficient for industrial applications. The scientists are still looking for perfect fit solution in both performance and management’s levels. This research is focused on finding a good solution that has high absorption-desorption performance and has lower corrosion rate, foaming and degradation. The relation between the performance and the chemical structure, finding the optimum condition of the single amine solutions and the blend ratio for best performance were also studied. However, two sets of amines studied mono amines and diamines. In the diamine set, the ethanol was added to the nitrogen atom in the structure while in the monoamine, the alkyl group was added to the structure. The desorption and absorption parameters criteria were used for selecting components of amine blend. The concentration and the ratio of the blend components varied in order to find the optimum ratio and concentration. The optimum blend and its single components were then studied for corrosion, foaming and degradation. The results of screening of 4AB and 4A2MB showed that adding methyl group to the straight chain enhance the absorption and the desorption performance while reduce the heat duty. The study of the diamines 22AEE and BMEM showed adding 3 methyl group sto the nitrogen atoms in the structure reduced the absorption and desorption performance and increased the heat duty, while in 22AEE and EDA; adding the -OH group to EDA as seen in 22AEE increased the rich loading, desorption rate and cyclic capacity while absorption rate, pKa, and heat duty reduced by adding the ethanol group to the structure. Considering –OH group in the structure increases the solubility of the amine and makes it less volatile which is preferred. However, the mass transfer limitations on all amines in this research had no impact on the performance at the concentrations used in the research. On the other hand, increasing the number of amine groups from mono to diamine caused to generate larger amounts of bicarbonate ions which lead to higher CO2 desorption rates and cyclic capacity, but lower heat duty. Also, the higher alkyl group found to have high viscosity. Adding ethanol groups to the diamine increased the viscosity in general but it had no impact on the performance. The developed criteria of blend selection of the diamine in terms of absorption parameter was based on taking the average pKa1 and pKa2 which resulted at the end in selection of the best performance fit of 22AEE:EDA (3:1) of overall 1M blend after screening several ratios and concentration. This had an outstanding desorption characteristics/heat duty as well as very good absorption characteristics. 22AEE to EDA 3:1 blend, implying that it is a good potential solvent for post combustion CO2 capture. Then, carried 3:1 blend for further management testing like (corrosion, foaming and degradation/emission). 22AEE found to have lower foaming than EDA and thus due to the existing three hydrophilic groups and higher surface tension compared to two groups in EDA structure. Corrosion found to be higher in EDA than 22AEE. The degradation rate found higher in 22AEE while the accumulated emission found higher in EDA.Item Open Access Institutional betrayal and trauma in child welfare workers in Western Canada(Faculty of Graduate Studies and Research, University of Regina, 2023-02) Boughner, Emily Nicole; Klest, Bridget; Loutzenhiser, Lynn; Asmundson, Gordon; Kikulwe, Daniel; Ricciardelli, RosemaryChild welfare workers report a high prevalence of mental health concerns, such as posttraumatic stress and depressive symptoms, after directly or indirectly experiencing workplace traumatic events (Regehr et al., 2000, 2004). Factors within an organization, such as supportive administrators, supervisors, and coworkers (Boyas et al., 2012; Dagan et al., 2016) can impact reported mental health symptoms. Institutional betrayal research, as an extension of betrayal trauma theory, suggests that institutions play a role in impacting the development and severity of mental health symptoms in individuals who are dependent on them. Institutional betrayal, the actions or inactions of an institution when responding to a traumatic event, has been used to explain how organizational factors impact mental health symptoms in different settings (Smith et al., 2016; Smith & Freyd, 2013). To date, there has been no specific measurement of institutional betrayal linking child welfare organizations to the mental health and professional well-being of child welfare workers, after a workplace traumatic event occurs. The purpose of the current study was to explore this relationship in Canadian child welfare workers. Using a mixed-method approach was used, 70 child welfare workers in Saskatchewan and Alberta completed surveys about workplace traumatic events, mental health and professional well-being, and institutional betrayal and support. Thirteen child welfare workers in frontline and supervisory positions participated in semi-structured interviews to gather further information about mental health and professional well-being after workplace traumatic events. The survey data showed how institutional support was significantly related to professional well-being, such that workers who reported higher levels of institutional support also reported higher levels of professional well-being. ii Experiences of workplace traumatic events were significantly related to institutional betrayal, such that workers who reported direct exposure to workplace traumatic events also reported higher levels of institutional betrayal. Additionally, child welfare workers reported experiences of traumatic events at a prevalence similar to first responder professions (Regehr, Chau, et al., 2002; Regehr, Goldberg, et al., 2002; Van Ameringen et al., 2008). Using thematic analysis, three major themes were identified in the interview data: traumatic and stressful workplace experiences, mental health and professional well-being, and organizational responses. The experiences child welfare workers reported through survey and interview data fit within an institutional betrayal lens. Workers reported feeling distrustful and/or let down by their organizations after workplace traumatic events occurred, which was identified as impacting their mental health and professional wellness. Implications and workplace considerations are discussed, such as organizations increasing worker awareness of mental health and professional well-being supports and providing long-term and ongoing supports. The current research is the first to show that the workplace traumatic events and mental health and professional well-being of child welfare workers can be understood through the theoretical orientation of institutional betrayal.Item Open Access Clustering and dimensionality reduction for time-series service monitoring data(Faculty of Graduate Studies and Research, University of Regina, 2022-09) Anowar, Farzana; Sadaoui, Samira; Fan, Lisa; Sharhiar, Nashid; Shafiq, M. OmairService monitoring applications allow customers to measure the performance, availability, and resolve application issues before they affect users. Since service monitoring applications continuously produce data to monitor their availability, therefore, high dimensionality, unlabeled data and changing data distribution are all prevalent. In this thesis, we efficiently address these three issues using the constructed service monitoring dataset. Higher dimensionality means higher computational cost to perform training and often leads to over-fitting while l earning a model. Furthermore, in the presence of high dimensionality, data are highly correlated resulting in insignificant and irrelevant features. These features have less impact on the prediction. To this end, the first part of the thesis conceptually and empirically explores the most representative dimensionality reduction (DR) methods from different categories. Next, we construct a new and challenging End-to-End (E2E) service monitoring dataset by extracting heterogeneous sub-datasets from multiple subservers, tackling data incompleteness in each sub-dataset using several imputation techniques, and fusing all the optimally imputed sub-datasets. This target dataset is highly dimensional, temporal, unlabeled, and non-linear. As the dataset is new, based on robust clustering approaches, we thoroughly assess the quality of the initial dataset and the reconstructed datasets (same dimensionality as the initial dataset) produced with Deep and Convolutional AutoEncoders. The experiments disclose that the reconstructed dataset with Deep AutoEncoder is the most performing. Later, we propose an ensemble-based DR approach to effectively handle the high-dimensionality of the E2E dataset. The approach combines Deep AutoEncoder with Kernel Principal Component Analysis to produce better data, and then reduce the feature space respectively. Due to the massive size of the dataset, we divide it into six weekly sub-datasets. We show that no vital information is lost for the reduced sub-datasets using the reconstruction error and total explained variance ratio. Based on time-series data clustering methods, and metrics, we thoroughly evaluate the efficacy of the ensemble approach. As the initial dataset is unlabeled (so are the reduced sub-datasets), we improve the previously developed ensemble-based DR approach by further combining it with incremental DR to improve clusters’ performance and increase the cluster labels’ confidence. We consider the weekly datasets as chunks for the experimental purpose. The experiments reveal that clustering performances increase significantly after utilizing the improved ensemble-based DR. Hence, the clusters’ labels are considered as the target class labels. Finally, to process the incoming data for any service monitoring application, it is critical to classify data accurately in real-time. Hence, we consider the labeled data chunks as incoming data, and propose an adaptive classification framework using Learn++ that also handles evolving data distributions. This approach sequentially predicts and updates the monitoring model with new data, and gradually forgets past knowledge. We employ consecutive data chunks to evaluate the performance of the predictors incrementally. The experimental results demonstrate that the proposed method provides high detection rates and low misclassification rate for most of the adaptive chunks.