This thesis is concerned with the issues of design and optimisation of digital filters and multirate filter banks. The main focus and contribution of this thesis is to apply the genetic algorithm (GA) technique and to draw some comparison with the standard gradient and non-gradient based optimisation methods. The finite word length (FWL) constraint affects the accuracy of a real-time digital filter requency response. For the case of digital filters, this study is concerned with the optimisation of FWL coefficients using genetic algorithms. Some comparative study with the simple hill climber algorithms is also included. The outcome of this part of the study demonstrates a substantial improvement of the new results when compared with the simply rounded FWL coefficient frequency response. The FWL coefficient optimisation process developed in the earlier Chapters is extended to the field of multirate filter banks. All multirate filter banks suffer from the problems of amplitude, phase and aliasing errors and, therefore, constraints for perfect reconstruction (PR) of the input signal can be extensive. The problem, in general, is reduced to relaxing constraints at the expense of errors and finding methods for minimising the errors. Optimisation techniques are thus commonly used for the design and implementation of multirate filter banks. In this part of the study, GAs have been used in two distinct stages. Firstly, for the design optimisation so that the overall errors are minimised and secondly for FWL coefficient optimisation of digital filters that form the sub-band filters of the filter bank. This process leads to an optimal realisation of the filter bank that can be applied to specific applications such as telephony speech signal coding and compression. One example of the optimised QMF bank was tested on a real-time DSP target system and the results are reported. The multiple M-channel uniform and non-uniform filter banks have also been considered in this study for design optimisation. For a comparative study of the GA optimised results of the design stage of the filter bank, other standard methods such as the gradient based quasi-Newton and the non-gradient based downhill Simplex methods were also used. In general, the outcome of this part of study demonstrates that a hybrid approach of GA and standard method was the most efficient and effective process in generating the best results.

This thesis is concerned with the modelling of electrical machines for use in variable-speed drives. Even when the structure of the model of the machine is known values have to be assigned to the parameters. In addition, it is usual for only some of the state variables to be measured, any others needed being estimated using the model. The present work is a study of methods of making on-line estimates of the model parameters, using a reduced number of measured states. To offer high level dynamic torque control the non-measured state variables must be indirectly estimated to a high degree of accuracy throughout the complete range of operating conditions. The state estimator is generally classified with respect to the degree of structural complexity. At one end of the spectrum the model is constructed with a very high level of complexity in order to describe fully the system during any operating conditions. Because of this structural accuracy, the model parameters can be fixed prior to running the machine under normal conditions. However, this scheme suffers from a high computational burden in the state estimation process, and requires sophisticated commissioning strategies in order to permit the complete identification of the relatively large parameter set. The alternative is to use a simpler model structure and update the parameters with sufficient speed on-line in order to compensate for the inherently larger structural error. In this thesis the latter method is considered and preferred, as it has a greater robustness to unforeseen system behaviour and is more compatible with existing control strategies. As a consequence of the simple estimator the identification scheme has to compensate for the considerable structural errors. To this end the strategy of full parameter set identification is described. Work has also been done, and is presented, concerning on-line parameter identification using genetic optimisation techniques, which are shown to be well suited to this type of problem. The first studies of modelling and parameter extraction were concerned with dc machines, for these were thought to be simpler to model and to understand. DC machines are also different, in that the simple model includes mechanical parameters, and so represents a more complete system than the models of the ac machine studied later. A number of standard, enhanced and novel parameter identification methods are analysed and implemented on a practical machine and drive test bench. Also included were state estimators, intended to permit speed-sensorless control; however, the limitations on the experimental rig, based as it was around a commercially available drive, meant that some of the testing had to be done by running the estimator off-line, using data recorded from actual runs as the input. The thesis is primarily concerned with the induction machine and specifically the parameters required to permit field and speed sensorless rotor field orientated vector control. Sufficient work was done to allow a preliminary experimental comparison of a number of algorithms. At this stage it appears that several of these could be developed into successful drives, the precise choice depending on the specific application.

No description available.

This thesis documents the research that has led to the development of a number of methodologies for combining existing artificial intelligence and statistical techniques into a form appropriate for the development of an intelligent appraisal system for use in the residential property appraisal profession. The methodologies illustrate how regression based appraisal models, previously restricted to homogeneous data, can be applied to heterogeneous data without significant loss in accuracy. The majority of research, previous to this, has addressed this problem by manually selecting homogeneous sub-regions from a heterogeneous parent region. However, the main drawback with this approach is that the segregation of parent regions into sub-regions relies upon a significant amount of a priori knowledge pertaining to the location of the property being valued. The requirement for a commercial residential property appraisal system is one that given sufficient training evidence can automatically learn how to value a property in any region and be able to modify this knowledge over time. Two methodologies are proposed within the thesis to address this requirement. The first, using a technique known as the Kohonen Self Organising Map, makes an assumption that residential properties that share sufficient characteristics can be appraised using the same function The Kohonen Self Organising Map is used to cluster properties with respect to their property characteristics and locational characteristics represented using a mortgage transaction database and UK Census statistics. Aptness of each cluster to define a homogeneous subset suitable to train a regression model, such as multiple regression analysis or a neural network, is estimated using a form of 'nearest neighbour' analysis. The second methodology, improves on the previous by transforming the static 'cluster then observe' solution to a more dynamic one using a Genetic Algorithm to evolve good clusters from those that at first inspection were mediocre. Another issue that has hindered the development of intelligent residential property appraisal systems has been the inability of such models to express their underlying functional form. This is addressed from two perspectives in this thesis: Rules are derived that describe the characteristic make-up of the formed clusters and, alternative modelling techniques are used to generate the final training models that are able to express their functional form as a set of induced rules. The work contained within this thesis demonstrates the feasibility of such an automatic stratification approach. Also, the research illustrates that by observing the characteristics of the generated clusters formed, a useful insight into both the underlying reasoning of the generated models and also of the locational and financial makeup of the subject location can be gained.

Made available in DSpace on 2015-03-03T11:52:33Z (GMT). No. of bitstreams: 0 Previous issue date: 2014-10-24Bitstream added on 2015-03-03T12:06:30Z : No. of bitstreams: 1 000808114.pdf: 869921 bytes, checksum: d49e859927277a17ca43ba9e2d635677 (MD5) / Neste trabalho é proposta a resolução de problemas multimodais usando duas diferentes meta-heurísticas: Algoritmo Genético de Chu-Beasley modificado e o Algoritmo Genético de Chaves Aleatórias Viciadas (BRKGA), com foco principal nos resultados obtidos por esta última. É feita especificamente a implementação das meta-heurísticas e comparação dos resultados obtidos por estas diferentes técnicas. Uma característica muito importante do BRKGA é a estruturação que permite separar o algoritmo em duas parcelas claramente diferenciadas, uma parcela que depende exclusivamente das características do BRKGA e, portanto, independente do problema que se pretende resolver e outra parcela que depende exclusivamente das características especificas do problema que pretendemos resolver. Essa característica geral do BRKGA permite que ele seja facilmente aplicado a uma grande variedade de problemas, já que a primeira parcela pode ser integralmente aproveitada na resolução de um novo problema. Por outro lado, o Algoritmo Genético de Chu-Beasley (AGCB) é caracterizado pela substituição de um único indivíduo no ciclo geracional e pelo controle máximo de diversidade, mas isto não é suficiente para resolução de problemas complexos e multimodais, sendo assim, é apresentado o AGCB modificado, onde o critério de diversidade é estendido, a população inicial e o descendente gerado no ciclo geracional passa por uma melhoria local. Essas características tornam-o competitivo justificando a comparação com o BRKGA / In this work it is proposed the resolution of multimodal problems using two different meta- heuristics: Chu-Beasley’s Genetic Algorithm and Biased Random Key Genetic Algorithm (BRKGA), focusing mainly on the results obtained by the latter. Specifically the imple- mentation and comparison of results obtained by these different techniques is made. There are several metaheuristics, each with its own specific characteristics which have advan- tages and disadvantages for the resolution of certain problems and in several ways in the implementation and results. A very important feature of the BRKGA is the structure that allows to separate the algorithm into two clearly different parts, one part that depends exclusively on the characteristics of BRKGA and therefore independent of the problem to be solved and another part that depends exclusively on the specific characteristics of the problem we intend to solve. This general feature of the BRKGA allows it to be readily applied to a variety of problems, because the first component part can be fully utilized to solve a new problem. On the other hand, Chu-Beasley’s Genetic Algorithm (AGCB) is characterized by the replacement of a single individual in the generation cycle and by maximum control of diversity, but this is not enough to solve complex and multimodal problems, therefore it is presented the modified AGCB, where the diversity criterion is extended, the initial population and the descendant generated in the generational cycle passes through a local improvement. These features make it competitive, justifying the comparison with BRKGA

Resumo: Este projeto tem como objetivo descrever, implementar e avaliar um modelo de Unit Commitment (UC) hidrotérmico para o ambiente de mercados de energia. O modelo deve considerar aspectos que têm sido negligenciados na abordagem atualmente vigente no Brasil, ou seja, o modelo deve apresentar as seguintes contribuições: i) a discretização do problema deve ser feita em base horária e não semanal, de modo a permitir o estabelecimento de um mercado de curtíssimo prazo efetivo; ii) o modelo deve levar em conta os custos de partida/parada de máquinas, comparando a solução do modelo proposta com o modelo em que essses custos não são considerados; iii) as inter-relações entre os mercados pool e bilateral são descritas de forma explícita em um único problema de otimização de UC; iv) a inserção dos custos de oportunidade hidráulica propostos no âmbito desse trabalho na função de custos / Abstract: This project aims to describe, implement and evaluate a model of Unit Commitment (UC) for the hydrothermal environment of energy markets. The model takes into account aspects that have been neglected in the approach currently used in Brazil, and present the following contributions: i) discretization of the problem is performed in an hourly basis, instead of the weekly-based approach currently used, to allow the establishment of an effective market for short term generation planning; ii) the model takes into account unit start-up and shut down costs; the outcomes of the proposed model are compared with those of a model in which these costs are not considered; iii) the interrelationships between pool and bilateral markets are described explicity within a single optimization problem in the proposed UC model; iv) the inclusion of opportunity costs associated with hydraulic utility, proposed in the context of this research / Orientador: Leonardo Nepomuceno / Coorientador: Paulo Sérgio da Silva / Banca: Eduardo Nobuhiro Asada / Banca: Antonio Roberto Balbo / Mestre

Orientador: Gerson Francisco / Banca: Rogério Rosenfeld / Banca: Rodrigo de Losso da Silveira Bueno / Resumo: Neste trabalho é apresentado um estudo do Jogo da Minoria, um modelo que visa simular o comportamento coletivo dos agentes no mercado financeiro. As propriedades deste sistema, bem como a resolução analítica do mesmo, são tratadas. Por fim, faz-se uma discussão das relações do jogo com o mercado real e reproduz-se um método que busca a utilização deste sistema para fins de modelagem e previsão de séries temporais / Abstract: In this work we study the Minority Game, a model which tries to simulate the collective behavior of the agents in the financial market. The properties of the system, as well as its analytical resolution, are shown. A discussion of the relations between this game and the real market, and also a reproduction of a method, which uses this system to look for modeling and prediction of temporal series, are made / Mestre

Orientador: Rubén Augusto Romero Lázaro / Banca: Marina Lavorato de Oliveira / Banca: Marcelo Escobar de Oliveira / Resumo: Neste trabalho é proposta a resolução de problemas multimodais usando duas diferentes meta-heurísticas: Algoritmo Genético de Chu-Beasley modificado e o Algoritmo Genético de Chaves Aleatórias Viciadas (BRKGA), com foco principal nos resultados obtidos por esta última. É feita especificamente a implementação das meta-heurísticas e comparação dos resultados obtidos por estas diferentes técnicas. Uma característica muito importante do BRKGA é a estruturação que permite separar o algoritmo em duas parcelas claramente diferenciadas, uma parcela que depende exclusivamente das características do BRKGA e, portanto, independente do problema que se pretende resolver e outra parcela que depende exclusivamente das características especificas do problema que pretendemos resolver. Essa característica geral do BRKGA permite que ele seja facilmente aplicado a uma grande variedade de problemas, já que a primeira parcela pode ser integralmente aproveitada na resolução de um novo problema. Por outro lado, o Algoritmo Genético de Chu-Beasley (AGCB) é caracterizado pela substituição de um único indivíduo no ciclo geracional e pelo controle máximo de diversidade, mas isto não é suficiente para resolução de problemas complexos e multimodais, sendo assim, é apresentado o AGCB modificado, onde o critério de diversidade é estendido, a população inicial e o descendente gerado no ciclo geracional passa por uma melhoria local. Essas características tornam-o competitivo justificando a comparação com o BRKGA / Abstract: In this work it is proposed the resolution of multimodal problems using two different meta- heuristics: Chu-Beasley's Genetic Algorithm and Biased Random Key Genetic Algorithm (BRKGA), focusing mainly on the results obtained by the latter. Specifically the imple- mentation and comparison of results obtained by these different techniques is made. There are several metaheuristics, each with its own specific characteristics which have advan- tages and disadvantages for the resolution of certain problems and in several ways in the implementation and results. A very important feature of the BRKGA is the structure that allows to separate the algorithm into two clearly different parts, one part that depends exclusively on the characteristics of BRKGA and therefore independent of the problem to be solved and another part that depends exclusively on the specific characteristics of the problem we intend to solve. This general feature of the BRKGA allows it to be readily applied to a variety of problems, because the first component part can be fully utilized to solve a new problem. On the other hand, Chu-Beasley's Genetic Algorithm (AGCB) is characterized by the replacement of a single individual in the generation cycle and by maximum control of diversity, but this is not enough to solve complex and multimodal problems, therefore it is presented the modified AGCB, where the diversity criterion is extended, the initial population and the descendant generated in the generational cycle passes through a local improvement. These features make it competitive, justifying the comparison with BRKGA / Mestre

This thesis presents research into the possibility of developing a computerised system that can evaluate the aesthetics and engineering aspects of solid shapes. One of the research areas is also to include such an evaluation system into an existing evolutionary CAD system which utilizes the Genetic Algorithms (GAs) technology. An extensive literature survey has been carried out to better understand and clarify the vagueness and subjectivity of the concept of aesthetics, which leads to the work of defining and quantifying a set of aesthetic parameters. This research achieves its novelty in aiming to assist designers in evaluating the aesthetics and functional aspects of designs early in the conceptual design stage, and its inclusion into an evolutionary CAD system. The field of Computer Aided Design (CAD) lacks the aesthetics aspect of the design, which is very crucial in evaluating designs especially considering the trend towards virtual prototypes replacing physical prototypes. This research has managed to suggest, define and quantify a set of aesthetic and functional elements or parameters, which will be the basis of solid shape evaluation. This achievement will help designers in determining the fulfilment of design targets, where the designers will have a full control to determine the priority of each evaluation element in the developed system. In achieving this, computer software including a programming language package and CAD software are involved, which eventually led to the development of a prototype system called Computer Aided Aesthetics and Functions Evaluation (CAAFE). An evolutionary CAD system called Evolutionary Form Design (EFD), which utilizes GAs, has been available for few years now. It evolves shapes for quick and creative suggestions, however it lacks the automated evaluation and aesthetics aspects of the design. This research has worked into the integrating of CAAFE into EFD, which led to a system that could evolve objects based on a selected and weighed aesthetic and functional elements. Finally, surveys from users have also been presented in this thesis to offer improvement to the scoring system within the CAAFE system.

Kyoto University (京都大学) / 0048 / 新制・課程博士 / 博士(工学) / 甲第14577号 / 工博第3045号 / 新制||工||1453(附属図書館) / 26929 / UT51-2009-D289 / 京都大学大学院工学研究科航空宇宙工学専攻 / (主査)教授 吉村 允孝, 教授 椹木 哲夫, 教授 松原 厚 / 学位規則第4条第1項該当