In environmental studies, measurements of interest are often taken on multiple variables. The results of spatial data analyses can be substantially affected by the spatial configuration of the sites where measurements are taken. Hence, optimal designs which result in data guaranteeing efficient statistical inferences need to be studied. We study optimal designs on two large classes of spatial regions with respect to three design criteria, which were prediction, covariance parameter estimation, and empirical prediction. The first class of regions includes those in the plane, where Euclidean distance is used. The performance of the optimal designs is compared to that of randomly chosen designs. Optimal designs for a small example and a relatively large example are obtained. For the small example, complete enumeration of all possible designs is computationally feasible. For the large example, the computational difficulty in searching for the optimal spatial sampling design is overcome by a simulated annealing algorithm. The second class of spatial regions includes streams and rivers, where the distance is defined as distance along the stream network. A moving average construction is used to establish valid covariance and cross-covariance models using stream distance. Optimal designs for small and large examples are obtained. An application of our methodology to a real stream network is included. We discuss the impact of asymmetry in the cross covariance function on the spatial multivariate design. The relationship between multivariate optimal design and univariate optimal design if the multivariate design is restricted to be completely collocated is studied. The efficiency lost if we consider the design that is optimal within the class of collocated designs is discussed.

Submitted by Aelson Maciera (aelsoncm@terra.com.br) on 2017-05-19T18:15:12Z No. of bitstreams: 1 TesePLMA.pdf: 8101922 bytes, checksum: 456faa861edf6a27b2e7de9a7a271429 (MD5) / Approved for entry into archive by Ronildo Prado (ronisp@ufscar.br) on 2017-05-23T20:32:54Z (GMT) No. of bitstreams: 1 TesePLMA.pdf: 8101922 bytes, checksum: 456faa861edf6a27b2e7de9a7a271429 (MD5) / Approved for entry into archive by Ronildo Prado (ronisp@ufscar.br) on 2017-05-23T20:33:05Z (GMT) No. of bitstreams: 1 TesePLMA.pdf: 8101922 bytes, checksum: 456faa861edf6a27b2e7de9a7a271429 (MD5) / Made available in DSpace on 2017-05-25T14:21:19Z (GMT). No. of bitstreams: 1 TesePLMA.pdf: 8101922 bytes, checksum: 456faa861edf6a27b2e7de9a7a271429 (MD5) Previous issue date: 2016-04-29 / Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / This work deals with modeling applications, systematic and reliable optimization methodologies of global search, and other computational tools. It is expected that existing computational intelligence methods, encoded in an appropriate tool for the application of process engineering assisted by computer, can lead to useful numerical results for the modeling and optimization of different processes, including biotechnological processes (focus of this work). Thus, different types of methodologies suitable for computer applications, were studied here. The proposed methodologies were implemented and evaluated for the development and optimization of culture media for the fermentation process of Clostridium novyi type B, besides the fermentation process and enzymatic hydrolysis of bagasse associated with the production of bioethanol (1G and 2G). Thus, the potential application of these computational techniques was evaluated to biotechnological systems in different approaches. More specifically, it was performed: Classification of biotechnological systems ( "clustering") in kinetically similar regions to produce cellulosic ethanol (2G ethanol) using fuzzy logic; estimation by global search of kinetic parameters to an alcoholic fermentation model using Simmulated Annealing algorithm (SA) (Contributions to the thematic project FAPESP 2011 / 51902-9); formulation and optimization of economically viable culture media for Clostridium novyi type B using neuro-fuzzy data modeling followed by global search which maximize productivity, also utilizing SA algorithm as a search engine (this step of the project was conducted in partnership with the veterinary pharmaceutical company Vallée SA). The computational tools presented in this work were highly effective for modeling and optimization of the bioprocesses studied. / Este trabalho aborda aplicações de modelagem, metodologias sistemáticas e confiáveis de otimização por busca global, além de outras ferramentas computacionais. Espera-se que métodos de inteligência computacional existentes, codificados em uma ferramenta apropriada para a aplicação da engenharia de processos assistida por computador, resultem em resultados numéricos úteis para a modelagem e otimização de diferentes processos, incluindo-se os processos biotecnológicos (foco deste trabalho). Assim, diferentes tipos de metodologias, apropriadas para aplicações em computador, foram aqui estudadas. Os métodos propostos foram aplicados e avaliados ao desenvolvimento e otimização de meios de cultura para o processo fermentativo do microrganismo Clostridium novyi tipo B, além dos processos de fermentação alcoólica e hidrolise enzimática de bagaço de cana, associados à produção de bioetanol (1G e 2G). Desta forma, foi avaliado o potencial de aplicação destas técnicas computacionais aos sistemas biotecnológicos, em diversas abordagens. Mais especificamente, foram realizadas: classificação (“clustering”) de sistemas em regiões cineticamente semelhantes para a produção de etanol celulósico (Etanol 2G) utilizando lógica Fuzzy; estimação por busca global de parâmetros cinéticos do modelo para uma fermentação alcoólica utilizando o algoritmo Simmulated Annealing (SA) (Contribuições ao projeto temático FAPESP 2011/51902-9); formulação e otimização do meio de cultura economicamente viável para o Clostridium novyi tipo B utilizando a modelagem de dados por neuro-fuzzy seguido de busca global da composição de meio que maximize a produtividade utilizando também o algoritmo SA como ferramenta de busca global (esta etapa do projeto foi realizado em parceria com a empresa farmacêutica veterinária Vallée S.A). As ferramentas computacionais apresentadas neste trabalho se mostraram altamente efetivas para a modelagem e otimização dos bioprocessos estudados." / FAPESP: 2011/51902-9. / FAPESP: 2008/56246-0.