Public transport network design deals with finding the most efficient network solution among a set of alternatives, that best satisfies the often-conflicting objectives of different network stakeholders like passengers and operators. Simulation-based Optimisation (SBO) is a discipline that solves optimisation problems by combining simulation and optimisation models. The former is used to evaluate the alternative solutions, while the latter searches for the optimal solution among them. A SBO model for designing public transport networks is developed in this dissertation. The context of the research is the MyCiTi Bus Rapid Transit (BRT) network in the City of Cape Town, South Africa. A multi-objective optimisation algorithm known as the Non-dominated Sorting Genetic Algorithm (NSGA-II) is integrated with Activity-based Travel Demand Model (ABTDM) known as the Multi-Agent Transport Simulation (MATSim). The steps taken to achieve the research objectives are first to generate a set of feasible network alternatives. This is achieved by manipulating the existing routes of the MyCiTi BRT with a computer based heuristic algorithm. The process is guided by feasibility conditions which guarantee that each network has routes that are acceptable for public transport operations. MATSim is then used to evaluate the generated alternatives, by simulating the daily plans of travellers on each network. A typical daily plan is a sequential ordering of all the trips made by a commuter within a day. Automated Fare Collection (AFC) data from the MyCiTi BRT was used to create this plan. Lastly, the NSGA-II is used to search for an efficient set of network solutions, also known as a Pareto set or a non-dominated set in the context of Multi-objective Optimisation (MOO). In each generation of the optimisation process, MATSim is used to evaluate the current solution. Hence a suitable encoding scheme is defined to enable a smooth iv translation of the solution between the NSGA-II and MATSim. Since the solution of multi-objective optimisation problems is a set of network solutions, further analysis is done to identify the best compromise solution in the Pareto set. Extensive computational testing of the SBO model has been carried out. The tests involve evaluating the computational performance of the model. The first test measures the repeatability of the model's result. The second computational test considers its performance relative to indicators like the hypervolume and spacing indicators as well as an analysis of the model's Pareto front. Lastly, a benchmarking of the model's performance when compared with other optimisation algorithms is carried out. After testing the so-called Simulation-based Transit Network Design Model (SBTNDM), it is then used to design pubic transport networks for the MyCiTi BRT. Two applications are considered for the model. The first application deals with the public transport performance of the network solutions in the Pareto front obtained from the SBTNDM. In this case study, different transport network indicators are used to measure how each solution performs. In the second scenario, network design is done for the 85th percentile of travel demand on the MyCiTi network over 12 months. The results show that the model can design robust transit networks. The use of simulation as the agency of optimisation of public transport networks represents the main innovation of the work. The approach has not been used for public transport network design to date. The specific contribution of this work is in the improved modelling of public transport user behaviour with Agent-based Simulation (ABS) within a Transit Network Design (TND) framework. This is different from the conventional approaches used in the literature, where static trip-based travel demand models like the four-step model have mostly been used. Another contribution of the work is the development of a robust technique that facilitates the simultaneous optimisation of network routes and their operational frequencies. Future endeavours will focus on extending the network design model to a multi-modal context.

The present thesis is devoted to the development of models and algorithms to improve metabolic simulations of cyanobacterial metabolism. Cyanobacteria are photosynthetic bacteria of great biotechnological interest to the development of sustainable bio-based manufacturing processes. For this purpose, it is fundamental to understand metabolic behaviour of these organisms, and constraint-based metabolic modelling techniques offer a platform for analysis and assessment of cell's metabolic functionality. Reliable simulations are needed to enhance the applicability of the results, and this is the main goal of this thesis. This dissertation has been structured in three parts. The first part is devoted to introduce needed fundamentals of the disciplines that are combined in this work: metabolic modelling, cyanobacterial metabolism and multi-objective optimisation. In the second part the reconstruction and update of metabolic models of two cyanobacterial strains is addressed. These models are then used to perform metabolic simulations with the application of the classic Flux Balance Analysis (FBA) methodology. The studies conducted in this part are useful to illustrate the uses and applications of metabolic simulations for the analysis of living organisms. And at the same time they serve to identify important limitations of classic simulation techniques based on mono-objective linear optimisation that motivate the search of new strategies. Finally, in the third part a novel approach is defined based on the application of multi-objective optimisation procedures to metabolic modelling. Main steps in the definition of multi-objective problem and the description of an optimisation algorithm that ensure the applicability of the obtained results, as well as the multi-criteria analysis of the solutions are covered. The resulting tool allows the definition of non-linear objective functions and constraints, as well as the analysis of multiple Pareto-optimal solutions. It avoids some of the main drawbacks of classic methodologies, leading to more flexible simulations and more realistic results. Overall this thesis contributes to the advance in the study of cyanobacterial metabolism by means of definition of models and strategies that improve plasticity and predictive capacities of metabolic simulations. / La presente tesis está dedicada al desarrollo de modelos y algoritmos para mejorar las simulaciones metabólicas de cianobacterias. Las cianobacterias son bacterias fotosintéticas de gran interés biotecnológico para el desarrollo de bioprocesos productivos sostenibles. Para este propósito, es fundamental entender el comportamiento metabólico de estos organismos, y el modelado metabólico basado en restricciones ofrece una plataforma para el análisis y la evaluación de las funcionalidades metabólicas de las células. Se necesitan simulaciones fidedignas para aumentar la aplicabilidad de los resultados, y este es el objetivo principal de esta tesis. Esta disertación se ha estructurado en tres partes. La primera parte está dedicada a introducir los fundamentos necesarios de las disciplinas que se combinan en este trabajo: el modelado metabólico, el metabolismo de cianobacterias, y la optimización multiobjetivo. En la segunda parte, se encara la reconstrucción y la actualización de los modelos metabólicos de dos cepas de cianobacterias. Estos modelos se usan después para llevar a cabo simulaciones metabólicas con la aplicación de la metodología clásica Flux Balance Analysis (FBA). Los estudios realizados en esta parte son útiles para ilustrar los usos y aplicaciones de las simulaciones metabólicas para el análisis de los organismos vivos. Y al mismo tiempo sirven para identificar importantes limitaciones de las técnicas clásicas de simulación basadas en optimización lineal mono-objetivo que motivan la búsqueda de nuevas estrategias. Finalmente, en la tercera parte, se define una nueva aproximación basada en la aplicación al modelado metabólico de procedimientos de optimización multiobjetivo. Se cubren los principales pasos en la definición de un problema multiobjetivo y la descripción de un algoritmo de optimización que aseguren la aplicabilidad de los resultados obtenidos, así como el análisis multi-criterio de las soluciones. La herramienta resultante permite la definición de funciones objetivo y restricciones no lineales, así como el análisis de múltiples soluciones en el sentido de Pareto. Esta herramienta evita algunos de los principales inconvenientes de las metodologías clásicas, lo que lleva a obtener simulaciones más flexibles y resultados más realistas. En conjunto, esta tesis contribuye al avance en el estudio del metabolismo de cianobacterias por medio de la definición de modelos y estrategias que mejoran la plasticidad y las capacidades predictivas de las simulaciones metabólicas. / La present tesi està dedicada al desenvolupament de models i algorismes per a millorar les simulacions metabòliques de cianobacteris. Els cianobacteris són bacteris fotosintètics de gran interés biotecnològic per al desenvolupament de bioprocessos productius sostenibles. Per a aquest propòsit, és fonamental entendre el comportament metabòlic d'aquests organismes, i el modelatge metabòlic basat en restriccions ofereix una plataforma per a l'anàlisi i l'avaluació de les funcionalitats metabòliques de les cèl·lules. Es necessiten simulacions fidedignes per a augmentar l'aplicabilitat dels resultats, i aquest és l'objectiu principal d'aquesta tesi. Aquesta dissertació s'ha estructurat en tres parts. La primera part està dedicada a introduir els fonaments necessaris de les disciplines que es combinen en aquest treball: el modelatge metabòlic, el metabolisme de cianobacteris i l'optimització multiobjectiu. En la segona part, s'adreça la reconstrucció i l'actualització dels models metabòlics de dos soques de cianobacteris. Aquests models s'empren després per a portar a terme simulacions metabòliques amb l'aplicació de la metodologia clàssica Flux Balance Analysis (FBA). Els estudis realitzats en aquesta part són útils per a il·lustrar els usos i aplicacions de les simulacions metabòliques per a l'anàlisi dels organismes vius. I al mateix temps serveixen per a identificar importants limitacions de les tècniques clàssiques de simulació basades en optimització lineal mono-objectiu que motiven la cerca de noves estratègies. Finalment, en la tercera part, es defineix una nova aproximació basada en l'aplicació al modelatge metabòlic de procediments d'optimització multiobjectiu. Es cobreixen els principals passos en la definició d'un problema multiobjectiu i la descripció d'un algorisme d'optimització que asseguren l'aplicabilitat dels resultats obtinguts, així com l'anàlisi multi-criteri de les solucions. La ferramenta resultant permet la definició de funcions objectiu i restriccions no lineals, així com l'anàlisi de múltiples solucions òptimes en el sentit de Pareto. Aquesta ferramenta evita alguns dels principals inconvenients de les metodologies clàssiques, el que porta a obtenir simulacions més flexibles i resultats més realistes. En conjunt, aquesta tesi contribueix a l'avanç en l'estudi del metabolisme de cianobacteris per mitjà de la definició de models i estratègies que milloren la plasticitat i les capacitats predictives de les simulacions metabòliques. / Siurana Paula, M. (2017). Modelling and multiobjective optimization for simulation of cyanobacterial metabolism [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/90578

Yes / The integration of two or more processes in a hybrid system is one of the most desirable options to provide flexibility, interoperability and data sharing between the connected processes. Various examples of hybrid systems have been developed with coherent seawater desalination systems such as the combination of thermal and membrane technologies. This paper focuses on the simulation and optimisation of an integrated (hybrid) system of multi effect distillation and double Reverse Osmosis (RO) processes to produce different grades of water needed in a smart city from seawater resources. The optimisation-based model investigates five scenarios to obtain the highest productivity of drinking water, irrigation water, water for livestock and power plant water, whilst constraining the product water salinity to be within the required standards and with lowest specific energy consumption. For this purpose, multi objective optimisation problem was formulated using the gPROMS (general Process Modelling System) software. The results confirm the superiority of the developed hybrid system to sustain different grades of water in a smart city.

No / Clustering is an essential research problem which has received considerable attention in the research community for decades. It is a challenge because there is no unique solution that fits all problems and satisfies all applications. We target to get the most appropriate clustering solution for a given application domain. In other words, clustering algorithms in general need prior specification of the number of clus- ters, and this is hard even for domain experts to estimate especially in a dynamic environment where the data changes and/or become available incrementally. In this paper, we described and analyze the effec- tiveness of a robust clustering algorithm which integrates multi-objective genetic algorithm into a frame- work capable of producing alternative clustering solutions; it is called Multi-objective K-Means Genetic Algorithm (MOKGA). We investigate its application for clustering a variety of datasets, including micro- array gene expression data. The reported results are promising. Though we concentrate on gene expres- sion and mostly cancer data, the proposed approach is general enough and works equally to cluster other datasets as demonstrated by the two datasets Iris and Ruspini. After running MOKGA, a pareto-optimal front is obtained, and gives the optimal number of clusters as a solution set. The achieved clustering results are then analyzed and validated under several cluster validity techniques proposed in the litera- ture. As a result, the optimal clusters are ranked for each validity index. We apply majority voting to decide on the most appropriate set of validity indexes applicable to every tested dataset. The proposed clustering approach is tested by conducting experiments using seven well cited benchmark data sets. The obtained results are compared with those reported in the literature to demonstrate the applicability and effectiveness of the proposed approach.

Thesis (MScEng (Civil Engineering))--University of Stellenbosch, 2011. / ENGLISH ABSTRACT: Waterborne sewage system design and expansion objectives are often focused on minimising initial investment while increasing system capacity and meeting hydraulic requirements. Although these objectives make good sense in the short term, the solutions obtained might not represent the optimal cost-effective solution to the complete useful life of the system. Maintenance and operation of any system can have a significant impact on the life-cycle cost. The costing process needs to be better understood, which include maintenance and operation criteria in the design of a sewer system. Together with increasing public awareness regarding global warming and environmental degradation, environmental impact, or carbon cost, is also an important factor in decisionmaking for municipal authorities. This results in a multiplicity of different objectives, which can complicate the decisions faced by waterborne sewage utilities. Human settlement and migration is seen as the starting point of expansion problems. An investigation was conducted into the current growth prediction models for municipal areas in order to determine their impact on future planning and to assess similarities between the models available. This information was used as a platform to develop a new method incorporating indirect costs into models for planning waterborne sewage systems. The need to balance competing objectives such as minimum cost, optimal reliability, and minimum environmental impact was identified. Different models were developed to define the necessary criteria, thus minimising initial investment, operating cost and environmental impact, while meeting hydraulic constraints. A non-dominated sorting genetic algorithm (NSGA-II) was applied to certain waterborne sewage system (WSS) scenarios that simulated the evolutionary processes of genetic selection, crossover, and mutation to find a number of suitable solutions that balance all of the given objectives. Stakeholders could in future apply optimisation results derived in this thesis in the decision making process to find a solution that best fits their concerns and priorities. Different models for each of the above-mentioned objectives were installed into a multi-objective NSGA and applied to a hypothetical baseline sewer system problem. The results show that the triple-objective optimisation approach supplies the best solution to the problem. This approach is currently not applied in practice due to its inherent complexities. However, in the future this approach may become the norm. / AFRIKAANSE OPSOMMING: Spoelafvoering rioolstelsel ontwerp en uitbreiding doelwitte is dikwels gefokus op die vermindering van aanvanklike belegging, terwyl dit die verhoging van stelsel kapasiteit insluit en ook voldoen aan hidrouliese vereistes. Alhoewel hierdie doelwitte goeie sin maak in die kort termyn, sal die oplossings verkry dikwels nie die optimale koste-effektiewe oplossing van die volledige nuttige lewensduur van die stelsel verteenwoordig nie. Bedryf en instandhouding van 'n stelsel kan 'n beduidende impak op die lewensiklus-koste hê, en die kostebepalings proses moet beter verstaan word en die nodige kriteria ingesluit word in die ontwerp van 'n rioolstelsel. Saam met 'n toenemende openbare bewustheid oor aardverwarming en die agteruitgang van die omgewing, is omgewingsimpak, of koolstof koste, 'n belangrike faktor in besluitneming vir munisipale owerhede. As gevolg hiervan, kan die diversiteit van die verskillende doelwitte die besluite wat munisipale besluitnemers in die gesig staar verder bemoeilik. Menslike vestiging en migrasie is gesien as die beginpunt van die uitbreiding probleem. 'n Ondersoek na die huidige groeivoorspelling modelle vir munisipale gebiede is van stapel gestuur om hul impak op die toekomstige beplanning te bepaal, en ook om die ooreenkomstes tussen die modelle wat beskikbaar is te asesseer. Hierdie inligting is gebruik as 'n platform om ‘n nuwe metode te ontwikkel wat indirekte kostes inkorporeer in die modelle vir die beplanning van spoelafvoer rioolstelsels. Die behoefte is geïdentifiseer om meedingende doelwitte soos minimale aanvanklike koste, optimale betroubaarheid en minimum invloed op die omgewing te balanseer. Verskillende modelle is ontwikkel om die bogenoemde kriteria te definiëer, in die strewe na die minimaliseering van aanvanklike belegging, bedryfskoste en omgewingsimpak, terwyl onderhewig aan hidrouliese beperkinge. ‘n Nie-gedomineerde sorteering genetiese algoritme (NSGA-II), istoegepas op sekere spoelafvoering rioolstelsel moontlikhede wat gesimuleerde evolusionêre prosesse van genetiese seleksie, oorplasing, en mutasie gebruik om 'n aantal gepaste oplossings te balanseer met inagname van al die gegewe doelwitte. Belanghebbendes kan in die toekoms gebruik maak van die resultate afgelei in hierdie tesis in besluitnemings prosesse om die bes-passende oplossing vir hul bekommernisse en prioriteite te vind. Verskillende modelle vir elk van die bogenoemde doelwitte is geïnstalleer in die nie-gedomineerde sorteering genetiese algoritme en toegepas op 'n hipotetiese basislyn rioolstelsel probleem. Die resultate toon dat die drie-objektief optimalisering benadering die beste oplossing vir die probleem lewer. Hierdie benadering word tans nie in die praktyk toegepas nie, as gevolg van sy inherente kompleksiteite. Desnieteenstaande, kan hierdie benadering in die toekoms die norm word.

Thesis (PhD (Logistics))--University of Stellenbosch, 2011. / ENGLISH ABSTRACT: The design of a water distribution system (WDS) involves finding an acceptable trade-off between cost minimisation and the maximisation of numerous system benefits, such as hydraulic reliability and surplus capacity. The primary design problem involves cost-effective specifica- tion of a pipe network layout and pipe sizes (which are typically available in a discrete set of commercial diameters) in order to satisfy expected consumer water demands within required pressure limits. The problem may be extended to consider the design of additional WDS com- ponents, such as reservoirs, tanks, pumps and valves. Practical designs must also cater for the uncertainty of demand, the requirement of surplus capacity for future growth, and the hydraulic reliability of the system under different demand and potential failure conditions. A detailed literature review of exact and approximate approaches towards single-objective (minimum cost) WDS design optimisation is provided. Essential topics which have to be included in any modern WDS design paradigm (such as demand estimation, reliability quantification, tank design and pipe layout) are discussed. A number of formative concepts in multi-objective evo- lutionary optimisation are also reviewed (including a generic problem formulation, performance evaluation measures, comparative testing strategies, and suitable classes of metaheuristics). The two central themes of this dissertation are conducting multi-objective WDS design optimi- sation using metaheuristics, and a critical examination of surrogate measures used to quantify WDS reliability. The aim in the first theme is to compare numerous modern metaheuristics, in- cluding several multi-objective evolutionary algorithms, an estimation of distribution algorithm and a recent hyperheuristic named AMALGAM (an evolutionary framework for the simulta- neous incorporation of multiple metaheuristics applied here for the first time to a real-world problem), in order to determine which approach is most capable with respect to WDS design optimisation. Several novel metaheuristics are developed, as well as a number of new variants of existing algorithms, so that a total of twenty-three algorithms were compared. Testing with respect to eight small-to-large-sized WDS benchmarks from the literature reveals that the four top-performing algorithms are mutually non-dominated with respect to the vari- ous performance metrics. These algorithms are NSGA-II, TAMALGAMJndu, TAMALGAMndu and AMALGAMSndp (the last three being novel variants of AMALGAM). However, when these four algorithms are applied to the design of a very large real-world benchmark, the AMALGAM paradigm outperforms NSGA-II convincingly, with AMALGAMSndp exhibiting the best perfor- mance overall. As part of this study, a novel multi-objective greedy algorithm is developed by combining several heuristic design methods from the literature in order to mimic the design strategy of a human engineer. This algorithm functions as a powerful local search. However, it is shown that such an algorithm cannot compete with modern metaheuristics, which employ advanced strategies in order to uncover better solutions with less computational effort. The second central theme involves the comparison of several popular WDS reliability surro- gate measures (namely the Resilience Index, Network Resilience, Flow Entropy, and a novel mixed surrogate measure) in terms of their ability to produce designs that are robust against pipe failure and water demand variation. This is the first systematic study on a number of WDS benchmarks in which regression analysis is used to compare reliability surrogate measures with probabilistic reliability typically derived via simulation, and failure reliability calculated by considering all single-pipe failure events, with both reliability types quantified by means of average demand satisfaction. Although no single measure consistently outperforms the others, it is shown that using the Resilience Index and Network Resilience yields designs that achieve a better positive correlation with both probabilistic and failure reliability, and while the Mixed Surrogate measure shows some promise, using Flow Entropy on its own as a quantifier of re- liability should be avoided. Network Resilience is identified as being a superior predictor of failure reliability, and also having the desirable property of supplying designs with fewer and less severe size discontinuities between adjacent pipes. For this reason, it is recommended as the surrogate measure of choice for practical application towards design in the WDS industry. AMALGAMSndp is also applied to the design of a real South African WDS design case study in Gauteng Province, achieving savings of millions of Rands as well as significant reliability improvements on a preliminary engineered design by a consulting engineering firm. / AFRIKAANSE OPSOMMING: Die ontwerp van waterverspreidingsnetwerke (WVNe) behels die soeke na ’n aanvaarbare afruiling tussen koste-minimering en die maksimering van ’n aantal netwerkvoordele, soos hidroliese betroubaarheid en surpluskapasiteit. Die primere ontwerpsprobleem behels ’n koste-doeltreffende spesifikasie van ’n netwerkuitleg en pypgroottes (wat tipies in ’n diskrete aantal kommersiele deursnedes beskikbaar is) wat aan gebruikersaanvraag binne sekere drukspesifikasies voldoen. Die probleem kan uitgebrei word om die ontwerp van verdere WVN-komponente, soos op- gaardamme, opgaartenks, pompe en kleppe in ag te neem. Praktiese WVN-ontwerpe moet ook voorsiening maak vir onsekerheid van aanvraag, genoegsame surpluskapsiteit vir toekom- stige netwerkuitbreidings en die hidroliese betroubaarheid van die netwerk onder verskillende aanvraag- en potensiele falingsvoorwaardes. ’n Omvattende literatuurstudie word oor eksakte en benaderde oplossingsbenaderings tot enkel- doelwit (minimum koste) WVN-ontwerpsoptimering gedoen. Sentrale temas wat by heden- daagse WVN-ontwerpsparadigmas ingesluit behoort te word (soos aanvraagvooruitskatting, die kwantifisering van betroubaarheid, tenkontwerp en netwerkuitleg), word uitgelig. ’n Aantal basiese konsepte in meerdoelige evolusionˆere optimering (soos ’n generiese probleemformulering, werkverrigtingsmaatstawwe, vergelykende toetsingstrategie¨e, en sinvolle klasse metaheuristieke vir WVN-ontwerp) word ook aangeraak. Die twee sentrale temas in hierdie proefskrif is meerdoelige WVN-ontwerpsoptimering deur mid- del van metaheuristieke, en ’n kritiese evaluering van verskeie surrogaatmaatstawwe vir die kwantifisering van netwerkbetroubaarheid. Die doel in die eerste tema is om ’n aantal moderne metaheuristieke, insluitend verskeie meerdoelige evolusionere algoritmes en die onlangse hiper- heuristiek AMALGAM (’n evolusionere raamwerk vir die gelyktydige insluiting van ’n aantal metaheuristieke wat hier vir die eerste keer op ’n praktiese probleem toegepas word), met mekaar te vergelyk om sodoende ’n ideale benadering tot WVN-ontwerpoptimering te identi- fiseer. Verskeie nuwe metaheuristieke sowel as ’n aantal nuwe variasies op bestaande algoritmes word ontwikkel, sodat drie en twintig algoritmes in totaal met mekaar vergelyk word. Toetse aan die hand van agt klein- tot mediumgrootteWVN-toetsprobleme uit die literatuur dui daarop dat die vier top algoritmes mekaar onderling ten opsigte van verskeie werkverrigtings- maatstawwe domineer. Hierdie algoritmes is NSGA-II, TAMALGAMJndu, TAMALGAMndu en AMALGAMSndp, waarvan laasgenoemde drie nuwe variasies op AMALGAM is. Wanneer hierdie vier algoritmes egter vir die ontwerp van ’n groot WVN-toetsprobleem ingespan word, oortref die AMALGAM-paradigma die NSGA-II oortui-gend, en lewer AMALGAMSndp die beste resultate. As deel van hierdie studie is ’n nuwe meerdoelige gulsige algoritme ontwerp wat verskeie heuristiese ontwerpsmetodologiee uit die literatuur kombineer om sodoende die on- twerpstrategie van ’n ingenieur na te boots. Hierdie algoritme funksioneer as ’n kragtige lokale soekprosedure, maar daar word aangetoon dat die algoritme nie met moderne metaheuristieke, wat gevorderde soekstrategie¨e inspan om beter oplossings met minder berekeningsmoeite daar te stel, kan meeding nie. Die tweede sentrale tema behels die vergelyking van ’n aantal gewilde surrogaatmaatstawwe vir die kwantifisering van WVN-betroubaarheid (naamlik die elastisiteitsindeks, netwerkelastisiteit, vloei-entropie en ’n gemengde surrogaatmaatstaf ) in terme van die mate waartoe hul gebruik kan word om WVNe te identifiseer wat robuust is ten opsigte van pypfaling en variasie in aanvraag. Hierdie proefskrif bevat die eerste sistematiese vergelyking deur middel van regressie-analise van ’n aantal surrogaatmaatstawwe vir die kwantifisering van WVN-betroubaarheid en stogastiese betroubaarheid (wat tipies via simulasie bepaal word) in terme van ’n aantal toetsprobleme in die literatuur. Alhoewel geen enkele maatstaf as die beste na vore tree nie, word daar getoon dat gebruik van die elastisiteitsindeks en netwerkelastisiteit lei na WNV-ontwerpe met ’n groter positiewe korrelasie ten opsigte van beide stogastiese betroubaarheid en falingsbetroubaarheid. Verder toon die gebruik van die gemengde surrogaatmaatstaf potensiaal, maar die gebruik van vloei-entropie op sy eie as kwantifiseerder van betroubaarheid behoort vermy te word. Netwerkelastisiteit word as ’n hoe-gehalte indikator van falingsbetroubaarheid geidentifiseer en het ook die eienskap dat dit daartoe instaat is om ontwerpe met ’n kleiner aantal diskontinuiteite sowel as van ’n minder ekstreme graad van diskontinuiteite tussen deursnedes van aangrensende pype daar te stel. Om hierdie rede word netwerkelastisiteit as die surogaatmaatstaf van voorkeur aanbeveel vir toepassings van WVN-ontwerpe in die praktyk. AMALGAM word ook ten opsigte van ’n werklike Suid-Afrikaanse WVN-ontwerp gevallestudie in Gauteng toegepas. Hierdie toepassing lei na die besparing van miljoene rande asook noe- menswaardige verbeterings in terme van netwerkbetroubaarheid in vergeleke met ’n aanvanklike ingenieursontwerp deur ’n konsultasiefirma.

Receiver operating characteristic (ROC) curves are widely used for evaluating classifier performance, having been applied to e.g. signal detection, medical diagnostics and safety critical systems. They allow examination of the trade-offs between true and false positive rates as misclassification costs are varied. Examination of the resulting graphs and calcu- lation of the area under the ROC curve (AUC) allows assessment of how well a classifier is able to separate two classes and allows selection of an operating point with full knowledge of the available trade-offs. In this thesis a multi-objective evolutionary algorithm (MOEA) is used to find clas- sifiers whose ROC graph locations are Pareto optimal. The Relevance Vector Machine (RVM) is a state-of-the-art classifier that produces sparse Bayesian models, but is unfor- tunately prone to overfitting. Using the MOEA, hyper-parameters for RVM classifiers are set, optimising them not only in terms of true and false positive rates but also a novel measure of RVM complexity, thus encouraging sparseness, and producing approximations to the Pareto front. Several methods for regularising the RVM during the MOEA train- ing process are examined and their performance evaluated on a number of benchmark datasets demonstrating they possess the capability to avoid overfitting whilst producing performance equivalent to that of the maximum likelihood trained RVM. A common task in bioinformatics is to identify genes associated with various genetic conditions by finding those genes useful for classifying a condition against a baseline. Typ- ically, datasets contain large numbers of gene expressions measured in relatively few sub- jects. As a result of the high dimensionality and sparsity of examples, it can be very easy to find classifiers with near perfect training accuracies but which have poor generalisation capability. Additionally, depending on the condition and treatment involved, evaluation over a range of costs will often be desirable. An MOEA is used to identify genes for clas- sification by simultaneously maximising the area under the ROC curve whilst minimising model complexity. This method is illustrated on a number of well-studied datasets and ap- plied to a recent bioinformatics database resulting from the current InChianti population study. Many classifiers produce “hard”, non-probabilistic classifications and are trained to find a single set of parameters, whose values are inevitably uncertain due to limited available training data. In a Bayesian framework it is possible to ameliorate the effects of this parameter uncertainty by averaging over classifiers weighted by their posterior probabil- ity. Unfortunately, the required posterior probability is not readily computed for hard classifiers. In this thesis an Approximate Bayesian Computation Markov Chain Monte Carlo algorithm is used to sample model parameters for a hard classifier using the AUC as a measure of performance. The ability to produce ROC curves close to the Bayes op- timal ROC curve is demonstrated on a synthetic dataset. Due to the large numbers of sampled parametrisations, averaging over them when rapid classification is needed may be impractical and thus methods for producing sparse weightings are investigated.

The shareholders of E&P companies evaluate the future performance of these companies in terms of multiple performance attributes. Hence, E&P decision makers have the task of allocating limited resources to available project proposals to deliver the best performance on these various attributes. Additionally, the performance of these proposals on these attributes is uncertain and the attributes of the various proposals are usually correlated. As a result of the above, the E&P portfolio optimisation decision setting is characterised by multiple attributes with uncertain future performance. Most recent contributions in the E&P portfolio optimisation arena seek to adapt modern financial portfolio theory concepts to the E&P project portfolio selection problem. These contributions generally focus on understanding the tradeoffs between risk and return for the attribute NPV while acknowledging the presence of correlation among the assets of the portfolio. The result is usually an efficient frontier where one objective is set over the expected value of the NPV and the other is set over a risk metric calculated from the same attribute where, typically, the risk metric has a closed form solution (e.g., variance, standard deviation, semi-standard deviation). However, this methodology fails to acknowledge the presence of multiple attributes in the E&P decision setting. To fill this gap, this thesis proposes a decision support model to optimise risk and return objectives extracted from the NPV attribute and from other financial and/or operational attributes simultaneously. The result of this approach is an approximate Pareto front that explicitly shows the tradeoffs among these objectives whilst honouring intra-project and inter-project correlations. Intra-project correlations are incorporated into the optimisation by integrating the single project models to the portfolio model to be optimised. Inter-project correlation is included by modelling of the oil price a global variable. Additionally, the model uses a multi-objective simulation-optimisation approach and hence it overcomes the need of using risk metrics with closed form solutions. The model is applied to a set of realistic hypothetical offshore E&P projects. The results show the presence of complex relationships among the objectives in the approximate Pareto set. The ability of the method to unveil these relationships hopes to bring more insight to the decision makers and hence promote better investment decisions in the E&P industry. / http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1320463 / Thesis (M.Eng.Sc.) -- University of Adelaide, Australian School of Petroleum, 2008

Short Term Conflict Alert (STCA) is an automated warning system designed to alert air traffic controllers to possible loss of separation between aircraft. STCA systems are complex, with many parameters that must be adjusted to achieve best performance. Current procedure is to manually ‘tune’ the governing parameters in order to finely balance the trade-off between wanted alerts and nuisance alerts. We present an incremental approach to automatically optimising STCA systems, using a simple evolutionary algorithm. By dividing the parameter space into regional subsets, we investigate methods of reducing the number of evaluations required to generate the Pareto optimal Receiver Operating Characteristic (ROC) curve. Multi-archive techniques are devised and are shown to cut the necessary number of iterations by half. A method of estimating the fitness of recombined regional parameter subsets without actual evaluation on the STCA system is presented, however, convergence is shown to be severely stunted when relatively weak sources of noise are present. We describe a method of aggressively perturbing parameters outside of their known ‘safe’ ranges when complex inhibitory interactions are present that prevent an exhaustive search of permitted values. The scheme prevents the optimiser from repeating ‘mistakes’ and unnecessarily wasting evaluations. Results show that a more complete picture of the Pareto-optimal ROC curve may be obtained without increasing the number of necessary iterations. Efficacy of the new methods is discussed, with suggestions for improving efficiency. Sources of parameter interdependence and noise are explored and where possible mitigating techniques and procedures suggested. Classifier performance on training and test data is investigated and potential solutions for reducing overfitting are evaluated on a toy problem. We comment on potential uses of the ROC in characterising STCA performance, for comparison to other systems and airspaces. Many industrial systems are structured in a similar way to STCA, we hope that techniques presented will be applicable to other highly parametrised, expensive problem domains.

Changes in the electric power industry have brought great challenges and uncertainties in transmission planning area. More effective planning of transmission grids with the appropriate development of advanced planning technologies is badly-needed. The aim of this research is to develop an advanced probabilistic transmission expansion planning (TEP) methodology in a continually changing market environment. The methodology should be able to strengthen and increase the robustness of existing transmission network. By using the proposed probabilistic TEP methodology, it can reduce the risks of major outages and identify weak buses in the system. The significance of this research is shown by its comprehensiveness and powerful practicability. Results from this research are able to improve the planning efficiency and reliability with consideration of financial risks in an electricity market. In order to achieve the target, this research methodologies focused on two main important issues, (1) probability based technical assessment and (2) financial investment evaluation. During the first stage study, probabilistic congestion management, probabilistic reliability evaluation and probabilistic load flow for TEP under uncertainties have been investigated and improved. The developed methodologies and indices, which truly represent the composite impact from both critical state and probability, have linked with financial terms. At financial investment evaluation part, Monte Carlo market simulation is performed to assist economic analysis. The overall planning process has been treated as a constrained multi-objective optimisation task. Comprehensive investigations are conducted on several test systems and testified by real power systems using the available reliability data and economic information from the Australian National Electricity Market (NEM). Overall, this research developed probabilistic transmission planning methodologies that can reflect modern market structures more accurately and it enable a greater utilization of current generation and transmission resources to increase potential operation efficiencies.