The Unified Modeling Language (UML) has become the standard language for the design of object-oriented software systems over the past decade. Even though there exist various tools which claim to support design with UML, their functionality is usually focused on drawing UML diagrams and generating code from the UML model. The task of choosing a suitable design which fulfils the requirements still has to be accomplished by the human designer alone. The aim of this thesis is to develop concepts for UML design tools which assist the modeller in improving the system design and requirements incrementally. For this approach a variant of formal games called exploration games is introduced as underlying technique. Exploration games can be defined on the basis of incomplete and imprecise UML models as they occur frequently in practice. The designer repeatedly plays an exploration game to detect flaws or incompleteness in the design and its specification, which are both incorporated in the game definition. At any time the game definition can be incremented by the modeller which allows him to react to the discoveries made during a play and experiment with new design solutions. Exploration games can be applied to UML in different variants. For each variant must be specified how the UML diagrams are used to set up the game and how the semantic variation points of UML should be interpreted. Furthermore some parts of the game definition may not be contained in the UML model and have to be provided separately. The emphasis of this thesis is on game variants which make use of UML diagrams for modelling system behaviour, especially state machines and activity diagrams. A prototypical implementation demonstrates how the concepts developed in this thesis can be put into practice. The tool supports the user in defining, playing and incrementing a game. Moreover it can compute winning strategies for the players and may act as opponent of the modeller. As example a game variant based on UML state machines has been implemented. The architecture that has been chosen for the tool leaves room for extension by additional game variants and alternative algorithms.

Switched reluctance motors with fully pitched windings are a relatively recent advancementin motor technology having only been in existences ince the early 1990's. They have been shown previously to offer greater torque per unit copper loss, and hence higher torque density, than conventional switched reluctance machines with short pitched windings. Early work by Mecrow and Barrass has demonstrated operation of prototype machines, developed and assessedv arious methods of control strategy, and made some comparisons of machine efficiency and inverter rating. The results presented here build on this early work by, in essence, examining the aspects of machine design, control strategy and inverter topology that affect drive performance and cost. Detailed comparisons of inverter rating and machine efficiency are made under equal conditions with the various methods of excitation that are possible. This is achieved with results from a test rig, including temperature rise tests, and the use of accurate dynamic simulation. The latter is developed to accurately model the motor with its strong mutual coupling between phases, various inverter topologies and the details of the controller such as digital PWM. As a result comparisons between simulated and measured results are shown to be very good. The fundamentals of machine design are examined with a view to optimising the machine for fully pitched windings. Previous work has indicated that good results are achieved when a conventional machine is simply rewound, however it is shown that further improvements can be made. Proposals are made to improve the drive in terms of both machine performance and power electronic rating. A search method is proposed that optimises current waveshape for either maximum torque per unit copper loss, or smooth torque for lowest loss. The method works over the entire speed range, as the rate of change of flux linkage is taken into account. Three alternative power electronic converters are developed, one of which is also particularly suitable for the short pitched winding machine. Aspects of silicon rating, current controllability, and current sensor requirements are discussed.

Classifying pipe cracks by size from their pulse-echo ultrasonic signal is difficult but highly significant for the defect evaluation required in pipe testing and maintenance decision making. For this thesis, a binary Support Vector Machine (SVM) classifier, which divides pipe cracks into two categories: large and small, was developed using collected ultrasonic signals. To improve the performance of this SVM classifier in terms of reducing test errors, we first combined the Sequential Backward Selection and Sequential Forward Selection schemes for input feature reduction. Secondly, we used the data dependent kernel instead of the Gaussian kernel as the kernel function in the SVM classifier. Thirdly, as it is time-consuming to use the classic grid-search method for parameter selection of SVM, this work proposes a Kernel Fisher Discriminant Ratio (KFD Ratio) which makes it possible to more quickly select parameters for the SVM classifier.

Thesis (PhD in Metallurgical Engineering)--University of South Australia, 1996

Production of micro-scale components is an important emergent field. One underdeveloped area is the production of micro-scale 3D surfaces, which has important applications in micro-optics and fibre optic sensors. One particular application is the production of micro-lenses. With scales of less than 200 ??m these lenses can improve light coupling efficiencies in micro-optic systems. However, current lens production techniques have limitations in accuracy and versatility. Creating these surfaces through mechanical micro-grinding has the potential to improve the precision and variety of profiles that can be produced, thus improving transmission efficiencies and leading to new applications. This work presents a novel micro-grinding method for the production of microscale asymmetric, symmetric and axisymmetric curved components from brittle materials such as glasses. A specialised micro-grinding machine and machining system has been designed, constructed and successfully tested and is presented here. This system is capable of producing complex profiles directly on the tips of optical fibre workpieces. A five degree of freedom centring system is presented that can align and rotate these workpieces about a precision axis, enabling axisymmetric grinding. A machine vision system, utilising a microscope lens system and sub-pixel localisation techniques, is used to provide feedback for the process, image processing techniques are presented which are shown to have a sensing resolution of 300 nm. Using these systems, workpieces are centred to within 500 nm. Tools are mounted on nanometre precise motion stages and motion and infeed are controlled. Tooling configurations with flat and tangential grinding surfaces are presented along with control and path generation algorithms. The capabilities and shortcomings of each are presented along with methods to predict appropriate feed rates based on experimental data. Both asymmetric and axisymmetric flat and curved micro-profiles have been produced on the tips of optical fibres using this system. These are presented and analysed and show that the system, as described, is capable of producing high quality micro-scale components with submicron dimensional accuracy and nanometric surface quality. The advantages of this technique are compared with other processes and discussed. Further development of the system and technique are also considered.

Device drivers constitute a significant portion of an operating system's source code. The effort to develop a new driver set is a sobering hurdle to the pursuit of novel operating system ventures. A practical solution is to reuse drivers, but this can contradict design goals in a new operating system. We offer a new approach to device-driver reuse, with a focus on promoting novel operating-system construction, which insulates the new operating system from the invariants of the reused drivers, while also addressing development effort. Our solution runs the drivers along with their original operating systems inside virtual machines, with some minor reuse infrastructure added to the driver's operating system to interface with the rest of the system. This approach turns the drivers into de-privileged applications of the new operating system, which separates their architectures and reduces cross-influences, and improves system dependability. Virtual machines help reuse drivers, but they also penalize performance. The known solution for improving virtual machine performance, para-virtualization, modifies the operating system to run on a hypervisor, which has an enormous cost: substantial development effort, and abandonment of many of virtualization's benefits such as modularity. These costs contradict our goals for driver reuse: to reduce development effort, and to easily reuse from a variety of operating systems. Thus we introduce a new approach to constructing virtual machines: pre-virtualization. Our solution combines the performance of para-virtualization with the modularity of traditional virtual machines. We still modify the operating system, but according to a set of principles called soft layering that preserves modularity, and via automation which reduces implementation costs. With pre-virtualization we can easily reuse device drivers. We describe our driver-reuse approach applied to a real system: we run virtual machines on the L4Ka::Pistachio microkernel, with reused Linux drivers. We include an evaluation and demonstrate that we achieve throughput comparable to the native Linux drivers, but with moderately higher CPU and memory utilization. Additionally, we describe how to apply pre-virtualization to multiple hypervisor environments. We include an evaluation of pre-virtualization, and demonstrate that it achieves comparable performance to para-virtualization for both the L4Ka::Pistachio and Xen hypervisors, with modularity.

Thesis (PhD in Metallurgical Engineering)--University of South Australia, 1996

No description available.

No description available.

No description available.