Find Theses

Theses
  • All
  • Scholar Profiles
  • Research Units
  • Research Output
  • Theses

61.Intelligent Global Maximum Power Point Tracking Strategies Based on Shading Perception for Photovoltaic Systems

Author:Ziqiang Bi 2021
Abstract:When a Photovoltaic (PV) system is partially shaded in the environment, the current-voltage (I-V) and power-voltage (P-V) curves exhibit multiple stairs/peaks and the locus of Maximum Power Point (MPP) varies over a wide range. Such Partial Shading Conditions (PSC) bring challenges to the Maximum Power Point Tracking (MPPT) systems. This thesis presents some novel shading information to characterize the complex PSC and MPPT techniques based on the shading perception. Shading information is the mathematical indicator to express the shading patterns. The existing shading information, such as shading rate and shading strength, has the limitations that they can only characterize the PSC with two irradiation levels. To improve the application range of the shading information, the shading matrix and shading vector are proposed in this thesis. The identification and detection methods for the proposed shading information are also included. Results from simulations and experiments have shown the effectiveness and accuracy of the proposed shading detection methods. Under PSC, the power characteristics of the PV systems are too complicated that there exist multiple MPPs. The traditional MPPT techniques may be trapped in the Local MPPs (LMPPs) instead of the Global MPP (GMPP). In this thesis, some novel methods are proposed to estimate the GMPP location from the detected shading information. The proposed MPPT techniques based on the shading perception are capable of tracking the GMPP fast and accurately. Simulations and experiments are conducted to validate the performance of the proposed MPPT methods with the comparison with some well-known MPPT methods.

62.Fibre Distribution Characterization and Its Impact on Mechanical Properties of Ultra High Performance Fibre Reinforced Concrete

Author:Lufan Li 2019
Abstract:Ultra-high performance fibre reinforced concrete (UHPFRC) is the most innovative cement based engineering material. It is also a big leap for the performance of this engineering material. The mechanical properties of UHPFRC not only depend on the properties of concrete matrix and fibres, but also depend on the interaction between these two elements. Moreover, this reaction is highly influenced by the fibre volume content distribution and fibre orientation distribution. Previous researchers had developed different methods to test the fibre distribution. However, apart from a genral fibre effciency reduction factor, there was no quantified relationship between the different fibre distribution and its corredponding mechanical performance. This research focuses on testing the fibre distribution and investigating their influences on the mechanical properties of UHPFRC. This research adopted the C-shape ferromagnetic probe inductive test.The effective depth of the magnetic probe was determined, and then this method was applied for testing the specimens with different thicknesses to obtain fibre volume content and fibre orientation angle. Image analysis was carried out on a number of specimens to prove the accuracy of the magnetic probe inductive test. Mechanical tests including compressive tests, uniaxial tensile tests and bending tests were carried out after the fibre distribution tests. The level of material performance enhancement is dependent on the fibre volume content and orientation angle. For tensile performance, the low dosage of fibres has little enhancement on the peak tensile/bending strength. Linear relationships can be found between the peak uniaxial tensile strength and fibre distribution with higher fibre dosages. This relationship was then further proved using the OpenSees programme. From the industrial point of view, over-dosing with fibres increases the construction cost. Furthermore, it may cause non-uniform fibre distribution and early concrete cracking. In order to improve the tensile behaviour of UHPFRC, adjusting the fibre orientation angle rather than simply increasing fibre volume content can be considered.

63.Green Supply Chain Management in Manufacturing Small and Medium‐sized Enterprises: Perspective from Chang Chiang Delta

Author:XiangMeng HUANG 2013
Abstract:This research started from an interest in how small and medium-sized enterprises (SMEs) in the manufacturing industry within the geographical area of Chang Chiang Delta in China operate with respect to sustainability by developing green supply chain management (GSCM). Therefore, the aim of this study is to investigate what the pressures are for SME manufacturers to implement GSCM practices, and to examine the relationship between those practices and corresponding performance at a regional level in the context of Chang Chiang Delta in China. To accomplish this task, a range of literature is evaluated, focusing on GSCM theories and adoptions. This review reveals a research gap regarding SMEs’ implementation of GSCM, to which this study responds. The research is underpinned by an interpretive epistemology and a multi-method design. It is an exploratory and empirical study with two rounds of primary data collection gathered from SME manufacturers in the Chang Chiang Delta region of China, which contains the triangular-shaped territory of Shanghai, southern Jiangsu Province and northern Zhejiang Province, including the urban cores of five cities – Shanghai, Nanjing, Hangzhou, Suzhou and Ningbo. In addition, a qualitative case study is employed in this research to provide more detailed information of GSCM implementation in SMEs. The results derived from both the questionnaire survey and the case study provide strong evidence that Chinese manufacturing SMEs have been under pressures relating to regulatory, customer, supplier, public and internal aspects from different stakeholder parties in terms of GSCM. In response to these pressures, SMEs have tried some GSCM practices, including green purchasing, eco-design, investment recovery, cooperation with customers and internal environmental management, and these practices are specific to the industrial sector considered in this study. But these practices do contribute to improving performance economically, environmentally and operationally. From the literature review and the empirical findings, this research provides contributions to knowledge, as well as managerial implications. It contributes to knowledge by providing conceptual and empirical insights into how GSCM is viewed and developed among SME manufacturers, clarifying the conceptions relating to sustainability, and incorporating stakeholder theory and the theory of industrial ecology in examining GSCM development. This study also provides practical implications by providing suggestions and guidance to governments, the public, suppliers and customers across the chain, as well as the managers of SMEs, and proposing an optimised model for the selected case for improved GSCM performance.

64.Middleware Techinques in A Rapid Response System in Wireless Sensor Networks

Author:Yuechun Wang 2021
Abstract:Wireless Sensor Networks (WSNs) are composed of embedded computers equipped with sensors, actuators and low-power radios that self-organise to form wireless networks capable of sensing the physical world. Despite their proven efficacy, modern WSNs uptake remains limited. This is primarily due to the applications' growing demand on the valid information extracted from the timeliness data flow and the complexity of data processing on the massive dynamic sensed data on motes. Two research problems could be addressed. The first one is how to ensure the effectiveness of the information extracted from the real-time sensed data; the second one is how to address the complex computing problem on the WSN nodes with limited computing ability. In the distributed Internet architecture, edge computing technology has outstanding performances in solving the problem of rapid network service response. It provides computing, storage, and network bandwidth near data source or users. Meanwhile, as an extension of cloud computing, the application of fog computing proposed by Cisco in the WSN environment has extensively exerted the effectiveness of collaboration among the massive WSN nodes, thereby improving the computing efficiency of the node clusters. Suppose the advantages of edge computing can be brought into play in the WSN scenario and combined with the characteristics of fog computing. In that case, the core problem, which is how to balance the trade-off between the responding time of a WSN system and the computing complexity of data processing, can be solved. Inspired by the research problems and potential solutions described above, the thesis explores several technologies that can take the characteristics of WSN and the requirements of logistics applications into account, such as hierarchical edge computing and Mobile Agent (MA). The hierarchical edge computing architecture proposed in the thesis in conjunction with the innovative rapid response strategy can ensure that under the premise of 80% coverage of real-time network nodes, data anomalies can be classified autonomously and responded within 20 sampling units, thereby reducing the delay caused by waiting for cloud computing results and decision communication. Besides, applying the mobile agent cooperation mechanism, a mobile agent-based middleware framework is presented in the thesis. Based on the randomly generated network topology, the middleware is tested under a variety of network scale with the node coverage by a MA applied patrol mechanism of above 98%.

65.Discriminative and Generative Learning with Style Information

Author:Haochuan Jiang 2019
Abstract:Conventional machine learning approaches usually assume that the patterns  follow the identical and independent distribution (i.i.d.). However, in many empirical cases, such condition might  be violated when data are equipped with diverse and inconsistent style information. The effectiveness of those traditional predictors may be limited due to the violation of the i.i.d. assumption brought by the existence of the style inconsistency. In this thesis,  we investigate how the style information can be appropriately utilized for further lifting up the performance of machine learning models. It is fulfilled by not only introducing the style information into some state-of-the-art models, some new architectures, frameworks are also designed and implemented with specific purposes to make proper use of the style information. The main work is listed as the following summaries: First, the idea of the style averaging is initially introduced by an example of an image process based sunglasses recovery algorithm to perform robust one-shot facial expression recognition task. It is named as Style Elimination Transformation (SET). By recovering the pixels corrupted by the dark colors of the sunglasses brought by the proposed algorithm, the classification performance is promoted on several state-of-the-art machine learning classifiers even in a one-shot training setting.  Then the investigation of the style normalization and style neutralization is investigated with both discriminative and generative machine learning approaches respectively. In discriminative learning models with style information, the style normalization transformation (SNT) is integrated into the support vector machines (SVM) for both classification and regression, named as the field support vector classification (F-SVC) and field support vector regression (F-SVR) respectively.  The SNT can be represented with the nonlinearity by mapping the sufficiently complicated style information to the high-dimensional reproducing kernel Hilbert space. The learned SNT would normalize the inconsistent style information, producing i.i.d. examples, on which the SVM will be applied. Furthermore, a self-training based transductive framework will  be introduced to incorporate with the unseen styles during training. The transductive SNT (T-SNT) is learned by transferring the trained styles to the unknown ones.  Besides, in generative learning with style information, the style neutralization generative adversarial classifier (SN-GAC) is investigated  to incorporate with the style information when performing the classification. As a neural network based framework, the SN-GAC enables the nonlinear mapping due to the nature of the nonlinearity of the neural network transformation with the generative manner. As a generalized and novel classification framework, it is capable of synthesizing style-neutralized high-quality human-understandable patterns given any style-inconsistent ones. Being learned with the adversarial training strategy in the first step, the final classification performance will be further promoted by fine-tuning the classifier when those style-neutralized examples can be well generated. Finally, the reversed task of the upon-mentioned style neutralization in the SN-GAC model, namely, the generation of arbitrary-style patterns, is also investigated in this thesis. By introducing the W-Net, a deep architecture upgraded from the famous U-Net model for image-to-image translation tasks, the few-shot (even the one-shot) arbitrary-style Chinese character generation task will be fulfilled. Same as the SN-GAC model, the W-Net is also trained with the adversarial training strategy proposed by the generative adversarial network. Such W-Net architecture is capable of generating any Chinese characters with the similar style as those given a few, or even one single, stylized examples.  For all the proposed algorithms, frameworks, and models mentioned above for both the prediction and generation tasks, the inconsistent style information is taken into appropriate consideration. Inconsistent sunglasses information is eliminated by an image processing based sunglasses recovery algorithm in the SET, producing style-consistent patterns. The facial expression recognition is performed based on those transformed i.i.d. examples. The SNT is integrated into the SVM model, normalizing the inconsistent style information nonlinearly with the kernelized mapping.  The T-SNT further enables the field prediction on those unseen styles during training. In the SN-GAC model, the style neutralization is performed by the neural network based upgraded U-Net architecture.  Trained with separated steps with the adversarial optimization strategy included, it produces the high-quality style-neutralized i.i.d. patterns. The following classification is learned to produce superior performance with no additional computation involved. The W-Net architecture enables the free manipulation of the style data generation task with only a few, or even one single, style reference(s) available. It makes the Few-shot, or even the One-shot, Chinese Character Generation with the Arbitrary-style information task to be realized. Such appealing property is hardly seen in the literature.

66.Multi-Task Learning with Convolutional Neural Networks

Author:Yizhang Xia 2018
Abstract:The CNN have achieved excellent performance in basic computer vision issues, such as, recognition and detection. However, the CNN is still an immature method, especially on multi-output classification. In traditional machine learning, the classic solution is MTL. The MTL was proposed early and has been an active topic. But, joint research on MTL and CNN are rarely mentioned. Fortunately, there is a successful integration of MTL and NN. And CNN is a typical NN. Especially, CNN is designed for computer vision. Based on the above situation, the mainly contributions of this thesis is the following three parts. Firstly, MTL and CNN is applied to face occlusion detection. This is the first time that MTL and CNN is used for detecting occluded face. The framework adopted the coarse-to-fine strategy, which consists of two CNNs. The first net is a region-based CNN detecting the head from a person upper body image while the second net is a multi-task CNN distinguishing which facial part is occluded from a head image. The experiment results prove that CNN can be integrated with MTL well. Secondly, MTL and CNN is used to jointly recognize vehicle logos and predict their attributes.In view of improving the performance of tasks, two MTL schemes, namely the adaptive weighted task learning and the switchable task learning, are proposed. To verify the algorithm, a large and realistic vehicle logo attributes dataset is prepared, which includes fifteen brands, labeled with six visual attributes and three no-visual attributes. Extensive experiments are conducted in two scenarios, equal priority learning and unequal priority learning, with promising accuracies. Thirdly, we propose a principled approach to design a evolutional tree-like multi-task deep learning framework which can be conveniently connected behind any well-known multi-class classification network and further improve their performance. Our approach starts with a basic multi-class deep architecture and dynamically deepens it during training using a criterion that groups similar tasks together. Extensive evaluation on multi-class classification datasets (MNIST and Cifar10) and multi-label prediction datasets (Berkeley Attributes of People dataset and CelebA) suggests that the models produced by the proposed method outperforms the strong baseline.

67.Bioavailability-based approach to understand the effects of metals as toxicants and nutrients: Implications for environmental management

Author:Boling Li 2022
Abstract:Environmental Management is a framing concept for the specific research topics in this thesis, and within that the work focuses on metals in the environment. Some of the work focuses on metals as toxicants, some on metal micronutrients, and some on metals that may be either, depending upon conditions. This thesis begins with work in which I developed a “two-in-one” whole-cell bioreporter approach to assess harmful effects of cadmium and lead. With the lights-on bioreporter’s unique two-in-one ability for speciation and toxicity measurement, in conjunction with the validated biotic ligand model, the bioreporter can predict toxicity endpoints over the range of the lowest Water Quality Criterion to the 50th rank-percentile of aquatic organisms sensitivity. In the context of dramatic environmental/biogeochemical change from metal pollution, relatively little work has been done on the role of micronutrients in influencing the development and progression of harmful algal blooms. In this thesis, I report results from mesocosm experiments with Microcystisand Desmodesmus spp., in mono- and mixed-cultures, to probe questions of how copper, iron, and copper-iron amendments affect the growth, short-term assemblage progression, and production of siderophore, chalkophore, and microcystin in lake water. The findings from this study are summarized: 1) copper-iron impacts on growth and community progression do not agree with lab-based findings. 2) Interplay between chalkophore/siderophore production supports a concept model wherein Microcystis spp. varies behavior to manage copper/iron requirements in a phased manner. In being able to specifically screen for chalkophores, I observed a previously unreported link between chalkophore and microcystin production that may relate to iron-limitation. 3) the lake water itself influences mesocosm changes; differentiated effects for iron regarding growth indicators and/or reduction of iron-limitation stress were found at a harmful algal bloom-free field station, likely a consequence of low bioavailability of iron in this station. My findings that Microcystis spp. varies behavior to manage copper/iron through the interplay between chalkophore/siderophore production and the previously unreported link between chalkophore and microcystin production addressed an important gap in furthering research on the effects of micronutrients bioavailability in natural water. Follow-up research with revised copper/iron amendments and increased level of algal acclimation was achieved. Similar to the initial work, I again saw a very similar dynamical phased behavior between chalkophore/siderophore production for Microcystis spp. that exhibited significant differences in trajectories according to specific differences in copper and iron amendments. The most interesting finding was that I observed a strong microcystin-chalkophore relationship again. Based on this research, I can say that chalkophore is a predictor of this cyanobacterial toxin production. While I discuss possible reasons for this new finding, it is previously undocumented, and I outline follow-up work that I believe would be fruitful to further elucidate the biological mechanisms underlying this behavior and why Microcystisspp. produce the toxin, microcystin.

68.Salient Object Detection and Segmentation in Video Surveillance

Author:Siyue Yu 2022
Abstract:Video surveillance outputs different portrait information of scenes such as crime investigation, security system, automatic driving system, and environmental monitoring. Recently, deep learning based video surveillance is also an essential topic in computer vision. The specific tasks include object tracking, video object segmentation, salient object detection, and video salient object detection. Thus, this thesis studies salient object detection and segmentation in video surveillance, mainly on video object segmentation and salient object detection. In video object segmentation, we study the case of given the first frame's mask and try to design a network that can adapt to different object appearance variations. Therefore, this thesis proposes a framework based on the non-local attention mechanism to localize and segment the target object in the current frame, referring to both the first frame with its given mask and the previous frame with its predicted mask. Our approach can achieve 86.5$% IoU on DAVIS-2016 and 72.2% IoU on DAVIS-2017, with a speed of 0.11s per frame. Then for salient object detection, this thesis focuses on scribble annotations. However, scribbles fail to contain enough integral appearance information. To solve this problem. A local saliency coherence loss is proposed to assist partial cross-entropy loss and thereby help the network learn more complete object information. Further, A self-consist mechanism is designed to help the network not sensitive to different input scales. Our method can achieve comparable results compared with fully supervised methods. Our method achieves a new state-of-the-art performance on six benchmarks (e.g. for the ECSSD dataset: Fβ = 0.8995, Eξ = 0.9079 and MAE = 0.0489). Lastly, co-salient object detection is also studied. Recent methods explore both intra- and inter-image consistency through an attention mechanism. We find that existing attention mechanisms can only focus on limited related pixels. Thus, we propose a new framework with a self-contrastive loss to mine more related pixels to obtain comprehensive features. Our method obtains 0.598 for maximum F-measure for COCA. In this way, the tasks in this thesis are well handled and our methods can serve as new baselines for future works.

69.Towards a theory of sharing economy-based service triad

Author:Dun Li 2020
Abstract:The sharing economy is a fast-growing phenomenon that has significantly disrupted traditional businesses. Sharing-economy businesses are involved in many sectors, such as the transportation, accommodation, labour, financial and food sectors. Thus, these companies have been considered important by both industrial and academic areas. Sharing-economy platform companies in different sharing-economy industries seek to constantly improve their “sharing” business to provide better services to users. Fundamentally, the nature of the sharing economy consists of three actors, the platform, service supplier and customer, forming a triadic structure within one specific sharing economy context. Among the main streams of service operations management research, it is surprising that, with a few exceptions, the role of platform service operations management in the sharing economy context has been ignored by researchers. Little is known about how sharing-economy platforms carry out their daily operations management in different sectors. To address this gap in the literature, four papers have been developed, including one literature review paper (conceptual paper) and three empirical papers. Seven unicorn level sharing-economy platform companies from three sharing-economy industries were selected for investigation in this research. They are DiDi and Uber China (ridesharing industry), OfO, Mobike and Hellobike (bike-sharing industry) and Huochebang and Yunmanman (logistics-sharing industry). By adopting different theories, such as balance theory, social capital theory, contingency theory, social exchange theory, information processing theory and the knowledge-based view, this study investigates different aspects of operations management under the sharing-economy context accordingly, such as the role of different platform strategies in sustainability, the influence of contingent factors on platform stickiness, bike-sharing platforms’ operations management and information management of the sharing-economy platform, and thus makes a significant theoretical contribution to the service operations management literature, providing insightful practical implications for sharing-economy platforms.

70.Power Line Communications over Time-Varying Frequency-Selective Power Line Channels for Smart Home Applications

Author:Wenfei ZHU 2014
Abstract:Many countries in the world are developing the next generation power grid, the smart grid, to combat the ongoing severe environmental problems and achieve e?cient use of the electricity power grid. Smart metering is an enabling technology in the smart grid to address the energy wasting problem. It monitors and optimises the power consumption of consumers’ devices and appliances. To ensure proper operation of smart metering, a reliable communication infrastructure plays a crucial role. Power line communication (PLC) is regarded as a promising candidate that will ful?l the requirements of smart grid applications. It is also the only wired technology which has a deployment cost comparable to wireless communication. PLC is most commonly used in the low-voltage (LV) power network which includes indoor power networks and the outdoor LV distribution networks. In this thesis we consider using PLC in the indoor power network to support the communication between the smart meter and a variety of appliances that are connected to the network. Power line communication (PLC) system design in indoor power network is challeng-ing due to a variety of channel impairments, such as time-varying frequency-selective channel and complex impulsive noise scenarios. Among these impairments, the time-varying channel behaviour is an interesting topic that hasn’t been thoroughly investi-gated. Therefore, in this thesis we focus on investigating this behaviour and developing a low-cost but reliable PLC system that is able to support smart metering applications in indoor environments. To aid the study and design of such a system, the characterisation and modelling of indoor power line channel are extensively investigated in this thesis. In addition, a ?exible simulation tool that is able to generate random time-varying indoor power line channel realisations is demonstrated. Orthogonal frequency division modulation (OFDM) is commonly used in existing PLC standards. However, when it is adopted for time-varying power line channels, it may experience signi?cant intercarrier interference (ICI) due to the Doppler spreading caused by channel time variation. Our investigation on the performance of an ordinary OFDM system over time-varying power line channel reveals that if ICI is not properly compensated, the system may su?er from severe performance loss. We also investigate the performance of some linear equalisers including zero forcing (ZF), minimum mean squared error (MMSE) and banded equalisers. Among them, banded equalisers provide the best tradeo? between complexity and performance. For a better tradeo? between complexity and performance, time-domain receiver windowing is usually applied together with banded equalisers. This subject has been well investigated for wireless communication, but not for PLC. In this thesis, we in-vestigate the performance of some well-known receiver window design criteria that was developed for wireless communication for time-varying power line channels. It is found that these criteria do not work well over time-varying power line channels. There-fore, to ?ll this gap, we propose an alternative window design criterion in this thesis. Simulations have shown that our proposal outperforms the other criteria.

72.Characterization of the cross-interactions between Deformed Wing Virus (DWV), honey bee, and ectoparasitic mite, Tropilaelaps mercedesae

Author:Yunfei Wu 2020
Abstract:Recently, honey bee colony losses have been reported to be associated with both presence of the pathogen Deformed Wing Virus (DWV) and ectoparasitic mites. The DWV vectoring role of Varroa destructor is well established while the role of Tropilaelaps mercedesae in viral transmission has not been fully investigated. In this project, I examined the effects of both mite species infestation on honey bee by comparing the DWV copy number and alteration of DWV variants of individual pupae and their infesting mites. Infestation with either mite species causes increased DWV copy number in honey bee pupae, which proves the vector role of V. destructor on honey bee and as well as suggesting a similar viral vectoring role for T. mercedesae. Through artificial infestation and wound induction experiments, a biological and mechanical vector role for T. mercedesae has been established. I also identified a positive correlation between DWV copy number in pupae and copy number in infesting mites, which forms two clusters with either high or low copy number in both honey bee pupae and infesting mites. The same DWV type A variant was present in either low or high copy number in both honey bee pupae and infesting V. destructor or T. mercedesae. These data suggest a previously proposed hypothesis that DWV suppressed the honey bee immune system when DWV copy number reaches a specific threshold, promoting greater replication. Tropilaelaps mercedesae infestation induces Hymenoptaecin and Defensin-1 expression in honey bee pupae; however, they are associated with two independent events, mite feeding activity and DWV replication, respectively. DWV can be transmitted from honey bee to mite via intake of fat body or other tissues through feeding activity, which is supported by the observation of accumulated DWV in the mite's intestinal region. During feeding activity, induced Hymenoptaecin is ingested by mite as well and it has a negative role, down-regulating vitellogenin synthesis, which further influences mite's reproductive capability. Hymenoptaecin expression induced by mite feeding exerts the negative feedback on the mite reproduction, and may help establishing an equilibrium between host (honey bee) and parasite (mite). I also explore the critical factors for DWV infection/replication, including 1) the host with A/T-rich genome and a skewed codon usage; 2) an intact accessible VP1-P domain on the viral virion; and 3) certain factors critical for viral replication and at least exclusively present in honey bee rather than V. destructor, T. mercedesae nor C. sonorensis.

73.Media Management and Disruptive Technology: The Nigerian Newspaper Industry Today

Author:Nelson Omenugha 2020
Abstract:New media technologies have brought about radical changes in the contemporary mass communication landscape. An important aspect of these changes which is currently provoking much interest concerns how these technologies are redefining and disrupting the operations, ethos and tastes of the old media, thus challenging the future of the traditional media institution. The Nigerian newspaper industry, like others elsewhere, is caught up in this new reality as new media technologies and the attendant alternative news sources increasingly gain footing in the country. This study, therefore, examines how newspaper managers in Nigeria, to secure their future in the new dispensation, have been responding to these urgent challenges posed by new media technologies. The research is anchored within various theories: Technological Determinism (TD), Disruptive Technology (DT), Diffusion of Innovation and Technology Acceptance Model (TAM) and puts forward the “Techno-Human Dynamism” model, as it seeks answer to the main research question: What are the observable trends in the management of Nigerian newspapers at a time when new media technologies are posing a challenge to the survival of traditional newspapers? Adopting a mixed qualitative research approach - Key Informant Interview (KII) and Focus Group Discussion (FGD), the study focuses on four major Nigerian daily newspapers - The Sun, The Nation, The Daily Trust and The Daily Times - as well as the newspaper readers of these daily newspapers. Three managerial personnel of each of the selected newspapers were interviewed, while Focus Group Discussion (FGD) of four sessions comprising six discussants each were conducted among newspaper readers in each of four purposively selected cities - Aroma junction (Awka, Anambra), Ojota junction (Ikeja, Lagos), Sky Memorial junction (Wuse, Abuja) and Rumukoro junction (Port Harcourt, Rivers) -  across the country. Employing the thematic method of data analysis, the study found that Nigerian newspapers, like their counterparts elsewhere, are already experiencing the disruptive impact of new media technologies in all major areas of their operations including content, human resources and revenue. These disruptive impacts appear to be strengthening rather than merely weakening the newspaper organisations. The newspapers in response to them have become more creative, more ethical - volatising factual, accurate, investigative and analytical reporting. These are issues that had hitherto posed huge ethical concerns about Nigerian journalism. Moreover, the hybridization (integration) of the new and old media as one of the coping strategies seems to add further strength to the newspapers as they poach on the strengths of the new media to complement the weaknesses of the old.  However, the newspaper managers still have some latitude to secure the future of the industry given the untapped potential of the industry both in the traditional and online sense. The study recommended that Nigerian newspapers should endeavour to keep pace with the technological innovations driving today’s newspaper industry while boldly considering other response strategies that have worked elsewhere - including journalistic co-operatives, mergers and conglomeration - towards arresting the dwindling fortunes of the industry.

74.Novel Numerical and Computational Techniques for Remote Sensor-Based Monitoring of Water Quality

Author:Xiaohui Zhu 2020
Abstract:Monitoring water quality in real time is one of essential measures for water environment management. Recent advances in information technology and sensor systems have catalysed the progress in remote monitoring of water quality using wireless sensor networks (WSNs). Much research has been carried on the optimal design of water quality monitoring networks, detection of anomaly water quality data and approach innovation of water quality monitoring to save the cost of building and operating water quality monitoring network, expand the monitoring area and improve the monitoring efficiency. A large number of optimization algorithms have been proposed to optimize water quality monitoring networks. Most of these algorithms consider unidirectional water flow and try to obtain global optimization monitoring networks without the consideration of special monitoring locations. This thesis studies optimization algorithms to design optimized water quality monitoring networks for bidirectional river systems. Reserved monitoring locations are also considered to satisfy particular monitoring requirements. Four optimization objectives of minimal pollution detection time, maximal pollution detection probability, maximal centrality of monitoring locations and reserved monitoring locations are considered. With the comparison of computing performance and Pareto frontier to several optimization algorithms, we propose a Constrained Multi-Objective Discrete Particle Swarm Optimization (CMODPSO) with new approaches to initialize particles and compute particles' velocities and positions during computing iterations. Experimental results show that the CMODPSO can obtain optimized water quality monitoring networks with reserved monitoring locations as well as satisfying other optimization objectives. Influenced by external interferences such as wild environment, sensor hardware errors and communication disturbance, there is a high probability that the water quality data collected by remote water quality monitoring networks is corrupted. It is a crucial challenge to detect and filter anomaly water quality data in real time during the monitoring. We propose a novel anomaly detection algorithm based on dual time-moving windows, which can detect anomaly water quality data in real time. Compared to other anomaly detection algorithms such as anomaly detection and mitigation (ADAM) and anomaly detection (AD), it can significantly improve the anomaly detection performance. To improve monitoring approaches from fixed-point monitoring to surface monitoring and expand the monitoring area, we develop an unmanned surface vehicle (USV) for water quality monitoring. An algorithm of Improved Angle Potential Field Method (IAPFM) is proposed for autonomous navigation and obstacle avoidance. A heading control algorithm is developed based on the proportional-integral-differential (PID) control. Experimental results show that the USV can autonomously navigate in a complex river system according to a predefined navigation route. In addition, it can also detect and avoid obstacles around the USV during navigation as well as collect water quality data in real time, which significantly improve the monitoring efficiency, expand the monitoring area and save the cost for building monitoring stations.

75.Body-centred Interactive Textiles for Emotion Regulation

Author:Mengqi Jiang 2022
Abstract:Textiles have always been an important carrier for people to express their emotions. Wearable technology brings the power of computing closer to the body, which has led to the rising focus of bodily emotion-related interactive textiles design. Body motion-engaged affective design for emotion regulation also become a trend in games and interactive art, but the advantages of interactive textiles in this direction have not been thoroughly studied. This research aims to answer the overarching question: What are the key factors influencing body-centred interactive textiles design for emotion regulation? First, we conducted a comprehensive review of the literature on emotion-related interactive textiles design and combed the current state of body-centred interactive textiles and the opportunities that exist. From this, we discussed its research directions in (1) exploring the materiality of interactive textiles, (2) integrating body motion into interaction, and (3) creating sensory stimuli with body-centred interactive textiles. Then, we investigated the research question following the Research through Design approach and verified the feasibility of regulating emotion through body-centred interactive textiles. Five sub-research questions were identified to concrete the research direction. In the design research and practice, first, we identified effective gross movement-based interactive textiles with multi-sensory feedback mechanisms for emotion regulation and created the E-motionWear prototype for participants to test in the lab and real life. Then, we presented GesFabri, a collection of five interactive textile interfaces with distinct textures. They were created to investigate the intuitive interaction gestures of textile texture and the emotional effects of the fine movement-based interfaces with four sensory feedback mechanisms. Last, we developed iPillowPal, an affective movement-based interaction pillow for emotional communication and engagement between long-distance relationship (LDR) partners. We conducted a user study, lab study, and field study to achieve this goal. The results revealed that (1) wearing the gross movement-based interactive textiles positively impacted the users’ immediate emotion regulation, and users presented a more positive attitude towards their work. (2) both fine movement-based interaction on textiles and the feedback mechanism influenced user emotions, also highlighted the potential of materiality in designing interactive textiles, (3) affective movements vary on the scenario, serve a specific emotional purpose, and generate a particular emotional response. The affective movement-based interactive textiles shortened the emotional distance of LDR couples and improved their emotional state. Multi experimental results show that several emotion regulation strategies can be applied to body-centred interactive textiles design. Based on the findings, we summarized our research through the lenses of body movement, feedback mechanism, textile materiality, and emotion types, then reflected on the research approach and methods, design implication, and proposed future work directions. This study offers a new perspective on the potential of emotion regulation with interactive textiles and will contribute to a deeper understanding of body-centred affective design.

76.Electric Vehicle Energy Management Considering Stakeholders' Interest in Smart Grids

Author:Bing Han 2020
Abstract:With the electri fication in transportation systems, Electric Vehicles (EVs) have developed rapidly in recent years. At the same time, with large-scale EV integration to power grids, the charging behaviours of EVs bring both challenges and opportunities to power grids operation. This thesis focuses on the EV energy management in smart grids, and the EV energy management problem is studied considering three stakeholders' interests, i.e. EV owner, aggregator and grid, respectively. First, the economic relationship between EV owners and the aggregator is studied (EV owners' and aggregator's interest). Two multi-objective optimisation methods are applied to investigate the economic relationship between these two stakeholders and the aggregator{owner economic inconsistency issue is presented. To mediate this issue, a rebate factor is proposed in the model. The results show that a signi ficant reduction in the EV owners' charging fee from self-scheduling can be achieved while the aggregator profi t is maximised. Second, the EV aggregator bidding strategy in the electricity market is studied (aggregator's interest). By jointly considering the reserve capacity in the day-ahead market and the uncertainty of reserve deployment requirements in the real-time market, a scenario-based stochastic programming method is used to maximise the expected aggregator pro fit. The risk of the deployed reserve shortage is addressed by introducing a penalty factor in the model. In addition, an owner-aggregator contract is designed to mitigate the economic inconsistency issue between EV owners and the aggregator. The results show that the expected aggregator profi t is guaranteed by maximising reserve deployment payments and mitigating the penalties and thus the uncertainty of the reserve market is well managed. Third, the EV integration in a transmission system is studied (grid's interest) to achieve the coordination between generators and EVs. To tackle the challenge of large-scale EV integration problem, a bi-level scheduling strategy is proposed. The bi-level strategy clearly de fines the responsibility of transmission system operator and the aggregator. An EV information grouping method is designed, which could efficiently tackle the optimisation complexity problem. In addition, a detailed EV battery charging model is built. The results show that the total cost of the systems is minimised and EVs could shave the peak and fill the valley loads. This thesis discusses the EV energy management problem considering three stakeholders' interests, respectively. The proposed strategies in this thesis clearly evaluate and de fine the economic relationships and responsibility among EV owners, aggregator and the grid in managing EV charging and discharging behaviours. Based on three case studies conducted in this thesis, EV energy management could bene t the stakeholders as follows: (1) the EV owner charging fee is minimised while their driving requirements are satis ed; (2) the aggregator profi t is maximised by participation in the electricity market; (3) the cost of the system is minimised by achieving the coordination between EVs and generators.

77.Investigating the Effects of Simian Retrovirus (SRV) Infection on the Autophagic Pathway, Apoptotic Pathway and m6A RNA Methylation in Jurkat Cells

Author:Jingting Zhu 2019
Abstract:Simian type D retrovirus (SRV) is an etiological agent for the fatal simian acquired immunodeficiency syndrome (SAIDS), which mainly infects Asian macaques and leads to varying degrees of immunosuppression. Until now, little was known about the underlying pathogenic mechanisms of SRV infection. Especially, the effects of SRV infection on T lymphocytes, the major host cells of SRV, are still largely unclear. Apoptosis and autophagy are two important evolutionarily conserved host immune defense pathways against viral invasion and mediate viral pathogenesis. In addition, in the last decade, a growing number of studies has revealed the emerging roles of m6A RNA modification in regulating the viral infection and virus-host cell interactions. Therefore, the aims of this thesis are to investigate the effects of SRV infection on the autophagic pathway, apoptotic pathway and m6A RNA modification in Jurkat T lymphocytes (Jurkat cells). The capacities of SRV infection and replication in Jurkat cells were also assessed. The results showed that both SRV-4 and SRV-8, the major SRV subtypes circulated in the macaque breeding colonies in China, were able to infect and replicate in Jurkat cells. In addition, both SRV-4 and SRV-8 infection have been shown to induce autophagy and apoptosis in Jurkat cells. The results demonstrated that SRV-4/SRV-8 infection was able to enhance the formation of autophagosome as well as to increase the completed autophagic flux in Jurkat cells. Moreover, the levels of activated caspase-3 and caspase-8 and apoptosis were significantly increased in Jurkat cells by SRV-4/SRV-8 infection. In addition, the SRV-8 infection-induced autophagy was shown to inhibit SRV replication and promote apoptosis in Jurkat cells. Inhibition of autophagy by knockdown of Beclin1 in SRV-8-infected Jurkat cells was shown to significantly increase the amount of SRV genome released in the culture medium, as well as to significantly decrease the levels of caspase-3/-8 activation and inhibit apoptosis. Interestingly, further investigations on the interaction between LC3 and procaspase-8 in SRV-8-infected Jurkat cells suggested that the autophagosomes was, at least partially, involve in the process of caspase-8 activation. In addition, the results in this thesis also showed that SRV-8 viral RNAs in the infected Jurkat cells contain six distinct m6A peaks. Moreover, SRV-8 infection was shown to decrease the global m6A level in Jurkat cells, as well as to reprogramme the Jurkat cellular m6A epitranscriptome. Interestingly, depletion of ALKBH5, an m6A “eraser”, or YTHDF1, an m6A “reader”, in the infected Jurkat cells was demonstrated to significantly decrease SRV-8 replication, suggesting the regulatory roles of m6A modification and the components of the cellular m6A machinery in SRV replication. The results in this thesis have revealed for the first time the effects of SRV infection on the autophagic and apoptotic pathways as well as on the m6A RNA methylation in Jurkat cells, which have the potential to provide novel insights for the development of new antiviral therapies.

78.Automated Certification of Online Auction Services

Author:Wei BAI 2016
Abstract:Auction mechanisms are viewed as an effcient approach for resource allocation and different types of auctions have been designed to allocate spectrum, determine positions for advertisements on web pages, and sell products on the Internet, among others. Online auctions can be implemented as an intermediary for both sellers and buyers in agent-mediated e-commerce systems. This raises two concerns. Firstly, the automation of online auction trading requires buyer agents to understand the auction protocol and have the ability to communicate with the seller agents (i.e., the auctioneer). Secondly, buyer agents need to automatically check desirable properties that are central to their decision making.To address both concerns, we have proposed a certification framework to enable soft-ware agents automatically verify some desirable properties of a specific auction through a formally designed communication protocol, and then make decisions according to the result of the communication. Furthermore, we have extended the communication mechanism to the area of Semantic Web Service composition and have explored the verification of combinatorial auction mechanisms. To demonstrate our approach, we have modelled online auctions as web services and have applied the technique of Semantic Web Service to represent auction protocols. Then we rely on computer-aided verification techniques to construct and check formal proofs of desirable properties for specifc auctions. Finally, dialogue games are proposed to enable decision making and service compositions for software agents.

79.Market risk management in WTI crude oil market

Author:Zihao Gong
Abstract:This research focuses on risk management in WTI crude oil market. It starts by discovering the underlying shocks that drive crude oil dynamics and how each shock can explain the price changes. After identifying the sources of the price dynamics, the third chapter focuses on constructing appropriate value-at-risk (VaR) risk models to measure the market risk of oil. The last part of the research utilizes the findings in chapters two and three to construct highly efficient hedging strategies for risk management purposes. Therefore, the links between each topic are progressive. The second chapter explores the drivers of oil price dynamics in the futures market from an economic view using the restricted vector autoregressive (VAR) model. The VAR approach is applied to decompose the futures prices into three components: supply shocks, demand shocks, and precautionary demand shocks. The level of the impact of the exogenous shocks on oil price dynamics is found to be time-varying and different in multiple economic events. In general, we find that the real demand shock plays a dominant role in determining the oil futures prices, followed by precautionary demand shocks and supply shocks. After understanding the sources of the market risks in the oil market, the third chapter solves how to measure the crude oil market risks accurately. This chapter measures the accuracy of the VaR model among the candidates for risk measurement in the oil futures market. A more flexible parametric distribution model is proposed in combination with GARCH models. We show that the newly proposed FIGARCH-SGT model improves the accuracy of VaR estimates compared with the competitors in the literature. The semi-parametric conditional POT model is less affected by the specification of volatility models and produces substantial forecasting accuracy. The fourth chapter constructs an optimal hedging strategy to manage market risks with oil-specific features discovered in previous chapters. We estimate the optimal hedge ratio based on newly proposed static and dynamic hedging models. It is found that static hedge performs better in risk reduction, especially in economic turmoil. In contrast, the dynamic hedge performs better in acquiring risk-adjusted returns, especially when the market is stable. Moreover, the term structure of the market: contango and backwardation are found to have a significant impact on the hedging models' performance.

80.Matching and Segmentation for Multimedia Data

Author:Hui Li 2022
Abstract:With the development of society, both industry and academia draw increasing attention to multimedia systems, which handle image/video data, audio data, and text data comprehensively and simultaneously. In this thesis, we mainly focus on multi-modality data understanding, combining the two subjects of Computer Vision (CV) and Natural Language Processing (NLP). Such a task is widely used in many real-world scenarios, including criminal search with language descriptions by the witness, robotic navigation with language instruction in the smart industry, terrorist tracking, missing person identification, and so on. However, such a multi-modality system still faces many challenges, limiting its performance and ability in real-life situations, including the domain gap between the modalities of vision and language, the request for high-quality datasets, and so on. Therefore, to better analyze and handle these challenges, this thesis focuses on the two fundamental tasks, including matching and segmentation. Image-Text Matching (ITM) aims to retrieve the texts (images) that describe the most relevant contents for a given image (text) query. Due to the semantic gap between the linguistic and visual domains, aligning and comparing feature representations for languages and images are still challenging. To overcome this limitation, we propose a new framework for the image-text matching task, which uses an auxiliary captioning step to enhance the image feature, where the image feature is fused with the text feature of the captioning output. As the downstream application of ITM, the language-person search is one of the specific cases where language descriptions are provided to retrieve person images, which also suffers the domain gap between linguistic and visual data. To handle this problem, we propose a transformer-based language-person search matching framework with matching conducted between words and image regions for better image-text interaction. However, collecting a large amount of training data is neither cheap nor reliable using human annotations. We further study the one-shot person Re-ID (re-identification) task aiming to match people by offering one labeled reference image for each person, where previous methods request a large number of ground-truth labels. We propose progressive sample mining and representation learning to fit the limited labels for the one-shot Re-ID task better. Referring Expression Segmentation (RES) aims to localize and segment the target according to the given language expression. Existing methods jointly consider the localization and segmentation steps, which rely on the fused visual and linguistic features for both steps. We argue that the conflict between the purpose of finding the object and generating the mask limits the RES performance. To solve this problem, we propose a parallel position-kernel-segmentation pipeline to better isolate then interact with the localization and segmentation steps. In our pipeline, linguistic information will not directly contaminate the visual feature for segmentation. Specifically, the localization step localizes the target object in the image based on the referring expression, then the visual kernel obtained from the localization step guides the segmentation step. This pipeline also enables us to train RES in a weakly-supervised way, where the pixel-level segmentation labels are replaced by click annotations on center and corner points. The position head is fully-supervised trained with the click annotations as supervision, and the segmentation head is trained with weakly-supervised segmentation losses. This thesis focus on the key limitations of the multimedia system, where the experiments prove that the proposed frameworks are effective for specific tasks. The experiments are easy to reproduce with clear details, and source codes are provided for future works aiming at these tasks.
Total 200 results found
Copyright 2006-2020 © Xi'an Jiaotong-Liverpool University 苏ICP备07016150号-1 京公网安备 11010102002019号