This thesis describes research into the development of a Computer Aided Design (CAD) tool that uses a Genetic Algorithm (GA) to generate and evolve original design concepts through human interaction. CAD technologies are firmly established in the later stages of design, and include many applications of Evolutionary Algorithms (EAs). The use of EAs as generative and search tools for conceptual design is less evident in fields other than abstract art, architecture and styling. This research gains its originality in aiming to assist designers early in the design process, by creating and evolving aesthetically interesting forms (objects). The integration of GA software with a solid modelling system has enabled the development of a prototype `Evolutionary Form Design' (EFD) system. Objects are defined using a genetic data structure and constructed from various geometric primitives and combinations of Boolean operators. The primitives interact in ways that are not easily predicted, often creating novel shapes that are unlikely to have been discovered through conventional means. Edge blending further adds to objects' complexity and visual appeal. Populations of objects are subjected to a `selective breeding' programme, directed through the user's allocation of scores, and may also be guided by simple geometric targets. These factors determine which objects are `fittest' and most likely to parent a new, hopefully improved generation of objects. The challenge has been to turn the concept into a genuinely useful tool, ensuring that desirable features are reproduced in subsequent populations. The key to achieving this is the way objects are recombined during reproduction. Work has included developing 4 novel routine for grouping the individual primitives that form objects using a Teamforming algorithm. Innovative, aesthetically interesting forms can be evolved intuitively and efficiently, providing inspiration and the initial models for original design concepts. Examples are given where the system'is used by undergraduates to generate seating designs, and by the author, to create virtual sculptures and a range of consumer product concepts.

This thesis is concerned with lifecycle data support for the design of large made to order products. These products have so many complex functions to perforrn that no one designer will have all of the relevant skills such as in structural design Z): or electrical engineering to produce a comprehensive design. This therefore leads to the utilisation of a team of designers who will not only fulfil logically different design roles but often work at different physical locations. In such a design environment there may be several local models, represented in local Z-- databasesT. hese databases may or may not support versioning either of the data or of the schema which evolves as the product design grows. The interfaces to these databases ID will be varied as they are intended to suit the local needs of the design aIgDe nt. This thesis proposes a model for version control in a desig4nn environment which does not alter the designers existing view. Cý tý A system of distributed co-operatinZg:, aZgD ents is presented whose goal is to manatDg e change and orgCaIDni se version sets in an enrgDin eering environment. The agents are designed for full lifecycle support and inter-operation across heterogeneous networks. The agent communication is based on CORBA but an extra messaging layer is developed which utilises a language built in VDM-SL (Vienna Development Method - Specification Language). A version model is presented in two ways informally based on the assumptions on a general design process and formally in VDM-SL. ZP tP In order to demonstrate the effectiveness of the version model, two industrial case studies are presented. The first of these is a study of offshore process engineering. The second is a study of conceptual ship design.

A growing demand for services and expensive innovative technologies is threatening the sustainability of healthcare systems worldwide. Decision makers in this environment struggle to set priorities appropriately, particularly because they lack consensus about which values should guide their decisions; this is because there is no agreement on best practices in priority setting. Decision makers (or ‘leaders’) who want to evaluate priority setting have little guidance to let them know if their efforts were successful t. While approaches exist that are grounded in different disciplines, there is no way to know whether these approaches lead to successful priority setting. The purpose of this thesis is to present a conceptual framework and an evaluation tool for successful priority setting. The conceptual framework is the result of the synthesis of three empirical studies into a framework of ten separate but interconnected elements germane to successful priority setting: stakeholder understanding, shifted priorities/reallocation of resources, decision making quality, stakeholder acceptance and satisfaction, positive externalities, stakeholder engagement, use of explicit process, information management, consideration of values and context, and revision or appeals mechanism. The elements specify both quantitative and qualitative dimensions of priority setting and relate to both process and outcome aspects. The evaluation tool is made up of three parts: a survey, interviews, and document analysis, and specifies both quantitative and qualitative dimensions and relates to both procedural and substantive dimensions of priority setting. The framework and the tool were piloted in a meso-level urban hospital. The pilot test confirmed the usability of the tool as well as face and content validity (i.e., the tool measured relevant features of success identified in the conceptual framework). The tool can be used by leaders to evaluate and improve priority setting.

国立情報学研究所で電子化したコンテンツを使用している。

Durante los últimos veinticinco años, tanto desde la psicología cognitiva, la psicología del desarrollo y la educación, como desde el estudio de la enseñanza de las ciencias, se han realizado diversos trabajos centrados en la descripción y análisis del cambio conceptual. Este campo de investigación aborda el estudio de las concepciones acerca de diversos fenómenos, las cuales se articulan y organizan en forma de teorías personales que pueden experimentar un proceso de cambio por enriquecimiento, reelaboración, revisión o reestructuración. Así, el estudio del cambio conceptual analiza las concepciones previas de los sujetos sobre algún concepto de interés, para después examinar el proceso a través del cual éstas varían o se reestructuran. Sobre la base de los estudios relativos al cambio conceptual, el presente trabajo se planteó con la finalidad de determinar las concepciones de aprendizaje que poseen estudiantes universitarios de los niveles inicial, intermedio y final en la licenciatura de Psicología, y analizar la influencia de dos factores centrales: a) la pericia en el dominio específico de la psicología, y b) el nivel de uso de las estrategias metacognitivas. Del mismo modo, nos planteamos el análisis de la posible contribución de otras variables relacionadas con el cambio conceptual de la concepción de aprendizaje: la edad, el sexo, la motivación, la percepción de la instrucción y la actuación del alumno.Los estudios acerca de las concepciones de aprendizaje se inician en la década de los setenta con base en una tradición fenomenográfica, pero tan sólo a partir de los noventa comienza a desarrollarse la investigación acerca de cómo evoluciona o cambia la concepción de aprendizaje. En este sentido, la conjunción de estas dos áreas de interés -los estudios cognitivos acerca del cambio conceptual y la tradición fenomenográfica en el estudio de las concepciones de aprendizaje- se presenta como una línea de investigación relativamente reciente.El trabajo consiste en un diseño transversal con sujetos pertenecientes a los niveles inicial, intermedio y final de la licenciatura de Psicología. El estudio se respaldó en: a) cuestionarios de autoreporte, b) entrevista semiestructurada y c) encuesta de datos personales. Así, nuestro principal objetivo es describir y analizar las concepciones de aprendizaje que poseen los estudiantes universitarios de psicología, analizar los posibles cambios que éstas experimentan y, si es ése el caso, indagar sobre el papel de la pericia y de las estrategias metacognitivas como factores que pueden estar asociados a dicho cambio.Los resultados apuntan a que las variables principales -las estrategias metacognitivas y la pericia- muestran relación significativa con la concepción de aprendizaje. Sin embargo, un análisis detallado para cada tipo de concepción señala que la concepción constructiva es la más asociada a dichas variables, mientras que el comportamiento de la concepción directa es prácticamente imposible de explicar a partir de la interdependencia entre las mismas variables. Los resultados obtenidos nos permiten afirmar que existe un proceso de cambio conceptual en las concepciones de aprendizaje, pero este proceso de cambio no parece ser homogéneo y debe explicarse desde el análisis particular de cada una de las categorías definidas. En este sentido, los estudiantes de psicología de nuestro estudio inician la carrera con una visión bastante heterogénea y poco coherente de las concepciones de aprendizaje y ello queda de manifiesto en la alta proporción de sujetos que presentan un nivel de co-dominancia al inicio de la carrera. Sin embargo, al considerar a los estudiantes que poseen un mayor nivel de pericia, se observa una disminución de la proporción de sujetos con co-dominancia; este último dato podría estar explicando que la concepción acerca del aprendizaje se hace más coherente y explícita, quizás como producto de la toma de conciencia y de la reflexión metacognitiva por parte de los sujetos.ENGLISH / When a student of Psychology begins his university studies, he has some knowledge about learning that can be categorized in different conceptions: direct, interpretative or constructive. At the beginning of the first course at the university the prevalent conception is the reproductive one (direct or interpretative), and this should be changed -by a process of enrichment, elaboration or restructuration- to the constructive conception. The conceptual change field deals with the description and explanation of this change process. There is a especial emphasis in finding out which factor is the most relevant for the change to take place. In the case of the undergraduates, some authors explain the conceptual change as a consequence of time involved in university courses (expert knowledge), whereas other authors hypothesis are that the change takes place thanks to an increase of metacognitive strategies (maturation processes). The aim of this study was to test three hypotheses about the change of undergraduate's conception of learning. The hypotheses are that the change of learning conception depends on: a) the knowledge level and knowledge structure, b) domain specific metacognitive strategies, c) the interaction of both factors (metacognition and expert knowledge on psychology). To test these hypothesis, we examined the conception of learning and metacognitive strategies in initial, intermediate and final psychology faculty students'. This was a transversal study, based on a phenomenographic approach, using quantitative and qualitative methods. Results show that the most prominent factor that explains the conceptual change from reproductive conception of learning to constructive conception is the level of metacognitive strategies. Nevertheless, the knowledge of psychology is important too, although less than metacognitive strategies. Therefore, there is an interaction of both factors but their contribution to conceptual change explanation is not equivalent.

Schema integration is the task of integrating several local schemata into one global database schema. It is a complex, error-prone and time consuming task. Problems arise in recognizing and resolving problems, such as differences and similarities, between two schemata. Problems also arise in integrating static and dynamic schemata. In this thesis, three research topics are addressed: Maintaining Vocabulary in Schema Integration, Integration of Static Schemata and Integration of Static and Dynamic Schemata, while applying the notation in the Enterprise Modeling approach. In Maintaining Vocabulary in Schema Integration an analysis of what semantic loss is and why it occurs in schema integration is conducted. Semantic loss is a problem that should be avoided because both concepts and dependencies might be lost. In the thesis, it is argued that concepts and dependencies should be retained as long as possible in the schemata. This should facilitate user involvement since the users’ vocabulary is retained even after resolving similarities and differences between two schemata. In Integration of Static Schemata two methods are developed. These methods facilitate recognition and resolution of similarities and differences between two conceptual database schemata.  By applying the first method, problems between two schemata can be recognized that otherwise could pass unnoticed; by applying the second method, problems can be resolved without causing semantic loss by retaining concepts and dependencies in the schemata. In Integration of Static and Dynamic Schemata a method on how to integrate static and dynamic schemata is developed. In the method, focus is put on pre- and post-conditions and how to map these to states and state changes in the database. By applying the method, states that are important for the database can be designed and integrated into the conceptual database schema. Also, by applying the method, active database rules can be designed and integrated into the conceptual database schema.

This dissertation deals with the development of mathematical concepts from a historical and didactical perspective. In particular, the development of concepts in mathematical analysis during the 19th century is considered. The thesis consists of a summary and three papers. In the first paper we investigate the Swedish mathematician E.G. Björling's contribution to uniform convergence in connection with Cauchy's sum theorem from 1821. In connection to Björling's convergence theory we discuss some modern interpretations of Cauchy's expression x=1/n. We also consider Björling's convergence conditions in view of Grattan-Guinness distinction between history and heritage. In the second paper we study visualizations in mathematics from historical and didactical perspectives. We consider some historical debates regarding the role of intuition and visual thinking in mathematics. We also consider the problem of what a visualization in mathematics can achieve in learning situations. In an empirical study we investigate what mathematical conclusions university students made on the basis of a visualization. In the third paper we consider Cauchy's theorem on power series expansions of complex valued functions on the basis of a paper written by E.G. Björling in 1852. We discuss Björling's, Lamarle's and Cauchy's different conditions for expanding a complex valued function in a power seris. In the third paper we also discuss the problem of the ambiguites of fundamental concpets that existed during the mid-19th century. We argue that Cauchy's and Lamarle's proofs of Cauchy's theorem on power series expansions of complex valued functions are correct on the basis of their own definitions of the fundamental concepts involved.

Despite being one of the largest aquifers of its type in Australia, the Burdekin River Delta (BRD) is an area that has received comparatively little research on its groundwater resources. This study conceptualises the hydrogeology of the BRD and characterises the relationships between the stratigraphic elements and the physical and chemical components of the groundwater system that influence the major governing processes. Importantly, a large amount of spatial and temporal groundwater information exists in database form, which enables an integrated conceptual model of the BRD aquifer to be developed from the key hydrogeological and hydrogeochemical relationships. Conceptualisation of the BRD aquifer is achieved by categorising four main aspects of the groundwater resource: 1. Surface characterisation; 2. Geologic characterisation; 3. Hydrogeologic characterisation; and 4. Groundwater System characterisation. The BRD is a large cuspate delta comprising a complex stratigraphy of Pleistocene to Holocene sediments of fluvial, deltaic and marine origin to a maximum depth of about 150 metres. The lower Pleistocene sediments lie predominantly below sea level and are typified by laterally discontinuous sands, silts and clays that have formed in response to fluctuating sea levels. The upper Pleistocene boundary is differentiated from the overlying Holocene sediments by a formerly exposed surface of semiconsolidated oxidised sandy clays and gravel. By contrast, the Holocene sediments comprise loose, uncompacted sequences of fluvial channel sands, interdistributary floodplain silts and marine incursions of estuarine clays and mangrove muds. The anastomosing array of fluvial sand bodies of former Burdekin River channels and levees is the setting for the main shallow aquifer units. Aquifer units of the lower Pleistocene sediments are in hydraulic connection with the Holocene units, effectively categorising the whole BRD as a single unconfined aquifer. Hydraulic gradients from both sides of the river divide the BRD into two broad flow regimes. Interpreted flow zones based on hydrograph patterns further subdivide the flow system based on seasonal recharge response to elevated river heights and flooding, and response to long-term rainfall patterns associated with La Niña episodes of the Southern Oscillation. Stable isotope data (2H and 18O) indicate that the dominant isotopic signature of groundwater throughout the BRD corresponds with intense rainfall activity, however high deuterium-excess values indicate that significant evaporation occurs prior to recharge. This infers dominant recharge by the Burdekin River that drains a massive catchment extending hundreds of kilometres inland. Direct recharge via rainfall infiltration is largely dependant on soil texture. More conductive soils are associated with the major levee systems that comprise the main shallow aquifers. Two evolutionary hydrogeochemical paths exist for the north and south sides of the river, and are constrained by the interpreted flow zones. In the south side, groundwater enters the main aquifer from river recharge and leakage out of weathered granite outcrops (exposed bedrock). Mineral hydrolysis and evaporative concentration of salts initially evolve groundwater in the weathered granite to a combination of Na-Cl and Na-HCO3 type. Leakage through clay-rich hillwash and marginal sediments causes reverse cation-exchange reactions where excess Na replaces Ca and Mg on ion-exchange surfaces. This leads to the formation of Mg,Ca-Cl type groundwaters into the southern parts of the main aquifer (supersaturated with respect to calcite and dolomite). Discharge towards the coast is characterised by seawater mixing where salinity increases with corresponding evolution to Na-Cl type waters. Recharge waters from the Burdekin River are fresh (<250mS/cm) Ca-HCO3 type, undersaturated with respect to calcite, and are easily distinguishable from the ion-exchange groundwater. In the north, only one smaller outcrop of bedrock exists, which hosts similar mineral hydrolysis reactions and base-exchange reactions. An absence of associated Na-Cl type waters means that reverse-cation exchange reactions are negligible, and so water types are predominantly Na-HCO3 type. Aquifer sands in the north are more widespread than in the south, so the fresh Ca-HCO3 recharge waters tend to dominate the overall groundwater composition, with Na-HCO3 types limited to the exposed bedrock areas. Towards the coastline, groundwater mixes with seawater towards Na-Cl type waters, similar to that observed in the south. The mangrove mud sequences that flank the coastline of the BRD are associated with high-Fe and low-pH groundwater formed by the oxidation of Fe-sulphides such as pyrite). SO4 is a product of this reaction, but does not achieve abnormally high concentrations, possibly due to the presence of sulphate-reducing bacteria. Carbonate dissolution is a possible side effect of acid sulphate generation, with possible gypsum dissolution as a secondary source of SO4. This study tested an alternative method to characterising groundwater to determine if the spatial extent of hydrogeochemical processes could be defined and comparable results achieved. This method involved discriminating discrete statistical groups of ionic ratios based on their cumulative frequency distribution. The statistical groups are bounded by critical values that distinguish different chemical processes, referred to as hydrogeochemical indicators. Various tested ionic ratios produced analogous indicators, proving their reliability as a valid method for the characterisation of groundwater chemistry. The significance of this research underlies the importance of groundwater use in the BRD as a primary source of irrigation supplies. Land use expansion and unregulated pumping pose a risk to future groundwater quality and sustainable volumes. The understanding of the relationship between the main geologic elements and the subsequent hydrogeochemical processes provides a scientific basis for conceptualising the groundwater resource. This establishes a framework for initiating future groundwater management options.

The study focuses on the adjacent Tenthill and Ma Ma catchments which converge onto the heavily cultivated alluvial plain of Lockyer Creek. Groundwater extracted from the alluvial aquifers is the primary source of water for intensive irrigation. Within the study the hydrogeology is investigated, a conceptual groundwater model produced and a numerical groundwater flow model is developed from this. The hydrochemistry and stable isotope character of groundwater are also investigated to determine processes such as recharge and evaporation. Examination of bore logs confirms the Quaternary alluvium comprises a laterally continuous gravel aquifer with an average thickness of 4.5 m, overlain by mixed sands and clays which form a semi-confining layer with an average thickness of 22 m. Variations in long term groundwater hydrographs indicate the aquifer changes from confined to unconfined in some locations as water levels drop, while bores adjacent to creek banks display a rapid response to a flood event. Pump testing of bores screened in the gravel produces estimates of hydraulic conductivity ranging from 50-80 m/day and storativity of 0.00166 which are both within realistic bounds for this aquifer material. Major ion chemistry of surface water collected during a flood is Mgdominated, similar to alluvial groundwater in the Tenthill catchment and the Lockyer plain, suggesting a strong connection between surface and groundwater in these locations. Alluvial groundwater salinity in Tenthill catchment is typically less than 3500 ìS/cm but may approach 6000ìS/cm on the Lockyer plain. By contrast Ma Ma catchment alluvial groundwater is Na-dominated with conductivity up to 12000 ìS/cm and more associated with groundwater from the underlying sandstone bedrock. Stable isotope analyses of alluvial groundwater from throughout both catchments and the Lockyer plain are compared with basalt and sandstone groundwater. A range of processes have been identified including recharge to alluvium from basalt groundwater and evaporated surface water; and alluvial-bedrock groundwater mixing at some locations. Integration of the components of the study enabled the production of a conceptual hydrogeological model of the Lockyer alluvial plain, proposing two hydrostratigraphic units; the gravel aquifer and the overlying mixed sand and clay which acts as a semi confining unit. Hydrochemical and stable isotopic evidence suggests seepage from creek channels as the dominant recharge process. A single layer groundwater flow model using MODFLOW was developed, based on groundwater extraction data, to represent flow in the gravel aquifer. The model was calibrated to transient conditions with groundwater fluctuations, incorporating both drought and flood conditions. A sensitivity analysis for each of the aquifer properties demonstrates the model is insensitive to variations within realistic bounds for the gravel aquifer material, however, the model is highly sensitive to changes in the chosen boundary conditions. Predictive simulations with several annual extraction scenarios ranging from 1.75 to 0.5 ML/ha indicate the resulting minimum saturated aquifer thickness ranges from 0.03 to 1.4 m.

Thesis (MSc(Comp & InfoSc))--University of South Australia, 2001