The burdensome capital cost of urban water distribution systems demands the use of efficient optimization methods capable of finding a relatively inexpensive design that guarantees a minimum functionality under all conditions of operation. The combinatorial and nonlinear nature of the optimization problem involved accepts no definitive method of solution. Adaptive search methods are well fitted for this type of problem (to which more formal methods cannot be applied), but their computational requirements demand the development and implementation of additional heuristics to find a satisfactory solution. This work seeks to employ adaptive search methods to enhance the search process used to find the optimal design of any water distribution system. A first study presented here introduces post-optimization heuristics that analyze the best design obtained by a genetic algorithm--arguably the most popular adaptive search method--and perform an ordered local search to maximize further cost savings. When used to analyze the best design found by a genetic algorithm, the proposed post-optimization heuristics method successfully achieved additional cost savings that the genetic algorithm failed to detect after an exhaustive search. The second study herein explores various ways to improve artificial neural networks employed as fast estimators of computationally intensive constraints. The study presents a new methodology for generating any large set of water supply networks to be used for the training of artificial neural networks. This dataset incorporates several distribution networks in the vicinity of the search space in which the genetic algorithm is expected to focus its search. The incorporation of these networks improved the accuracy of artificial neural networks trained with such a dataset. These neural networks consistently showed a lower margin of error than their counterparts trained with conventional training datasets populated by randomly generated distribution networks.

There is an undisputed need to increase accuracy of snow cover estimation in regions comprised of complex terrain, especially in areas dependent on winter snow accumulation for a substantial portion of their annual water supply, such as the Western United States, Central Asia, and the Andes. Presently, the most pertinent monitoring and research needs related to alpine snow cover area (SCA) are: (1) to improve SCA monitoring by providing detailed fractional snow cover (FSC) products which perform well in temporal/spatial heterogeneous forested and/or alpine terrains; and (2) to provide accurate measurements of FSC at the watershed scale for use in snow water equivalent (SWE) estimation for regional water management. To address the above, the presented research approach is based on Landsat Fractional Snow Cover (Landsat-FSC), as a measure of the temporal/spatial distribution of alpine SCA. A fusion methodology between remotely sensed multispectral input data from Landsat TM/ETM+, terrain information, and IKONOS are utilized at their highest respective spatial resolutions. Artificial Neural Networks (ANNs) are used to capture the multi-scale information content of the input data compositions by means of the ANN training process, followed by the ANN extracting FSC from all available information in the Landsat and terrain input data compositions. The ANN Landsat-FSC algorithm is validated (RMSE ~ 0.09; mean error ~ 0.001-0.01 FSC) in watersheds characterized by diverse environmental factors such as: terrain, slope, exposition, vegetation cover, and wide-ranging snow cover conditions. ANN input data selections are evaluated to determine the nominal data information requirements for FSC estimation. Snow/non-snow multispectral and terrain input data are found to have an important and multi-faced impact on FSC estimation. Constraining the ANN to linear modeling, as opposed to allowing unconstrained function shapes, results in a weak FSC estimation performance and therefore provides evidence of non-linear bio-geophysical and remote sensing interactions and phenomena in complex mountain terrains. The research results are presented for rugged areas located in the San Juan Mountains of Colorado, and the hilly regions of Black Hills of Wyoming, USA.

At the moment social exclusion is a topical problem in a whole Europe. That’s why innovative decisions are prompted for social exclusive group of people in order to facilitate their integration process into the labour market. The stepping-stone of this work is e-advisor for choosing possibilities within social isolation environment. This e-advisor is created in accordance with artificial neural network and considering to individual person’s features give suggestions for the most suitable professions. Also in this work is presented disease diagnostic model, which is defined by artificial neural network.

Over the past decade, the popularity of regional bus services has grown in large North American cities owing to more people living in suburban areas and commuting to the Central Business District to work every day. Estimating journey time for regional buses is challenging because of the low frequencies and long commuting distances that typically characterize such services. This research project developed a mathematical model to estimate regional bus travel time using artificial neural networks (ANN). ANN outperformed other forecasting methods, namely historical average and linear regression, by an average of 35 and 26 seconds respectively. The ANN results showed, however, overestimation by 40% to 60%, which can lead to travellers missing the bus. An operational strategy is integrated into the model to minimize stakeholders’ costs when the model’s forecast time is later than the scheduled bus departure time. This operational strategy should be varied as the commuting distance decreases.

Over the past decade, the popularity of regional bus services has grown in large North American cities owing to more people living in suburban areas and commuting to the Central Business District to work every day. Estimating journey time for regional buses is challenging because of the low frequencies and long commuting distances that typically characterize such services. This research project developed a mathematical model to estimate regional bus travel time using artificial neural networks (ANN). ANN outperformed other forecasting methods, namely historical average and linear regression, by an average of 35 and 26 seconds respectively. The ANN results showed, however, overestimation by 40% to 60%, which can lead to travellers missing the bus. An operational strategy is integrated into the model to minimize stakeholders’ costs when the model’s forecast time is later than the scheduled bus departure time. This operational strategy should be varied as the commuting distance decreases.

Dissolved oxygen (DO) is an important water quality parameter that is used to assess the health of aquatic ecosystems. Typically physically-based numerical models are used to predict DO, however, these models do not capture the complexity and uncertainty seen in highly urbanised riverine environments. To overcome these limitations, an alternative approach is proposed in this dissertation, that uses a combination of data-driven methods and fuzzy numbers to improve DO prediction in urban riverine environments. A major issue of implementing fuzzy numbers is that there is no consistent, transparent and objective method to construct fuzzy numbers from observations. A new method to construct fuzzy numbers is proposed which uses the relationship between probability and possibility theory. Numerical experiments are used to demonstrate that the typical linear membership functions used are inappropriate for environmental data. A new algorithm to estimate the membership function is developed, where a bin-size optimisation algorithm is paired with a numerical technique using the fuzzy extension principle. The developed method requires no assumptions of the underlying distribution, the selection of an arbitrary bin-size, and has the flexibility to create different shapes of fuzzy numbers. The impact of input data resolution and error value on membership function are analysed. Two new fuzzy data-driven methods: fuzzy linear regression and fuzzy neural network, are proposed to predict DO using real-time data. These methods use fuzzy inputs, fuzzy outputs and fuzzy model coefficients to characterise the total uncertainty. Existing methods cannot accommodate fuzzy numbers for each of these variables. The new method for fuzzy regression was compared against two existing fuzzy regression methods, Bayesian linear regression, and error-in-variables regression. The new method was better able to predict DO due to its ability to incorporate different sources of uncertainty in each component. A number of model assessment metrics were proposed to quantify fuzzy model performance. Fuzzy linear regression methods outperformed probability-based methods. Similar results were seen when the method was used for peak flow rate prediction. An existing fuzzy neural network model was refined by the use of possibility theory based calibration of network parameters, and the use of fuzzy rather than crisp inputs. A method to find the optimum network architecture was proposed to select the number of hidden neurons and the amount of data used for training, validation and testing. The performance of the updated fuzzy neural network was compared to the crisp results. The method demonstrated an improved ability to predict low DO compared to non-fuzzy techniques. The fuzzy data-driven methods using non-linear membership functions correctly identified the occurrence of extreme events. These predictions were used to quantify the risk using a new possibility-probability transformation. All combination of inputs that lead to a risk of low DO were identified to create a risk tool for water resource managers. Results from this research provide new tools to predict environmental factors in a highly complex and uncertain environment using fuzzy numbers. / Graduate / 0543 / 0775 / 0388

Irrigation of large agricultural areas represents a significant portion of energy consumption in food producing countries. The electric power of irrigation pumps in rice crops, for example, may exceed 800 HP, while these systems are continually used during the harvest. In Brazil, non-technical losses in this type of consumer are generally due to fraud or error in power meters, or illegal connections. This type of problem is difficult to control because of the large length of rural feeder, in addition to access difficulties in many cases. This work presents a proposal for identifying non-technical losses in rural feeders containing pumping systems for irrigation of rice crops. The proposed methodology is based on the correlation of patterns of energy consumption, characteristics of the irrigated area and climatic conditions of the irrigation period. The developed system employs Artificial Neural Network technique, and it has as input a dataset of rainfall, temperature, solar irradiation, humidity, installed power and irrigated area of rice cultivation. The final result of the analysis indicates the percentage risk of each set of data and inconsistencies that can result in non-technical losses. The results of the developed methodology were obtained and validated from a real data base of crops of the period between 2009 and 2014, in the State of Rio Grande do Sul, Brazil. / A irrigação de grandes áreas agrícolas representa uma porção significativa do consumo de energia elétrica em países produtores de alimentos. A potência das bombas de irrigação em lavouras de arroz, por exemplo, pode ser superior a 800 CV, sendo que esses sistemas são utilizados continuamente no período da safra. No Brasil, as perdas não técnicas neste tipo de consumidor geralmente são devidas a fraudes, erros em medidores de energia e ligações clandestinas. Esse tipo de problema é de difícil fiscalização devido à grande extensão das linhas rurais, além da dificuldade de acesso, em muitos casos. Este trabalho apresenta uma proposta para a identificação de perdas não técnicas em alimentadores rurais contendo sistemas de bombeamento para irrigação de lavouras de arroz. A metodologia proposta é baseada na correlação dos padrões de consumo de energia elétrica, das características da área irrigada e das condições climáticas do período de irrigação. A metodologia emprega a técnica de Rede Neural Artificial, e tem como entrada um conjunto de dados de precipitação pluviométrica, temperatura, incidência solar, umidade do ar, carga instalada e área de solo irrigado característico do cultivo de arroz. O resultado final das análises indica o risco percentual de cada conjunto de dados e inconsistências que possam implicar em perdas não técnicas. Os resultados do trabalho foram obtidos e validados a partir de uma base de dados reais de safras do período entre 2009 e 2014, de lavouras do Estado do Rio do Grande do Sul.

A rede PROFIBUS DP é o barramento de campo mais utilizado na indústria mundial atualmente. Com o uso cada vez maior desta rede de \"chão de fábrica\" nas plantas industriais, o diagnóstico rápido de falhas tornou-se extremamente necessário e importante, o que permite minimizar os tempos de parada da instalação e consequentes prejuízos no processo produtivo. Este trabalho apresenta o estudo de técnicas baseadas em Redes Neurais Artificiais (RNA) que serão utilizadas em um trabalho futuro para se fazer um diagnóstico rápido de uma rede PROFIBUS em caso de falha, permitindo através dos conceitos básicos apresentados, uma análise criteriosa do seu desempenho. Inicialmente são apresentados conceitos básicos sobre as redes PROFIBUS DP, tais como arquitetura e versões do protocolo, a descrição da Camada Física, dentre outros assuntos relevantes. Nas etapas seguintes, serão apresentados outros tópicos importantes para entendimento do projeto como, conceitos sobre RNA e métodos de pré-processamento do sinal colhido no osciloscópio. / The PROFIBUS (Process Field Bus) DP is the most popular fieldbus communication used in the worldwide industry. With the increasing use of this network fieldbus in industrial plants, the rapid faults diagnosis has become extremely necessary and important, which minimizes the lodgments time and consequent losses in the production process. This work is based into a Artificial Neural Networks (ANN) tool that will be used to make a rapidly diagnosis of a PROFIBUS network in case of failure, allowing through basic concepts, a careful review of their performance. First the basics of PROFIBUS DP networks are presented, such as architecture and protocol versions, a description of the Physical Layer, among other relevant issues. In the following steps, others important topics to understand the project will be shown, such as, concepts of ANN and pre-processing methods from collected signal.

Made available in DSpace on 2014-06-11T19:30:50Z (GMT). No. of bitstreams: 0 Previous issue date: 2010-10-05Bitstream added on 2014-06-13T18:40:59Z : No. of bitstreams: 1 moretti_jf_dr_ilha.pdf: 2346139 bytes, checksum: 4b4bfcffd24744e4627ebd26e46f3196 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / O concreto é o material estrutural mais utilizado na construção civil. Há mais de um século e meio ele vem sendo estudado e aperfeiçoado. É confeccionado utilizando-se de matérias primas regionais, com características técnicas diferentes de outras regiões. O cimento também se apresenta com diversas formulações. Quantificar adequadamente esses materiais é o objetivo do estudo da dosagem do concreto, de tal modo a se obter um concreto que atenda às necessidades estruturais exigidas. Sendo a principal delas a resistência à compressão. A dosagem do concreto é uma prática essencialmente laboratorial quando se pensa em resultados aceitáveis. Através de experimentos são idealizados ábacos e diagramas que podem fornecer a resistência do concreto endurecido com diversas combinações de matérias primas utilizadas. Não há uma formulação matemática abrangente e bem definida para um processo generalizado de dosagem. A complexidade aumenta quando se adicionam outros componentes a mais no concreto simples e tradicional. Obter a relação entre eles é um trabalho contínuo. As redes neurais vêm sendo utilizadas na solução de problemas da engenharia civil, com ênfase na aplicação da técnica da retropropagação. Ela realiza satisfatoriamente as iterações entre as diversas variáveis, num processo de treinamento e aprendizagem, e tem sido capaz de generalizar soluções aceitáveis. Nesta pesquisa de doutorado é utilizada uma rede neural feedfoward com algoritmo retropropagação para prever a resistência e o módulo de elasticidade do concreto. Os dados de entrada são quantidades de materiais utilizadas para confeccionar 1 m3 de concreto adensado, de forma semelhante a dosagem por diagramas. Será aplicada na interpretação de diagramas de dosagem. Tais diagramas são amplamente utilizados por empresas na confecção de concretos,... / Concrete is the most widely used structural material in construction, for more than a century and a half it has been studied and improved. It's prepared using regional raw materials with different technical characteristics of other regions. The cement also performs with various formulations. Adequately quantify these materials is the goal of the study of the concrete mixtures proportion, to obtain a concrete that meets the structural needs required. The main one being the compressive strength. The strength of concrete is essentially a practice laboratory when one considers acceptable results. Through experiments are idealized abacus and diagrams that can provide the strength of hardened concrete with various combinations of raw materials used. There is no mathematical formulation of comprehensive and well defined for a generalized process of mixes. The complexity increases when other components is added in the most simple and traditional concrete. Obtain the relationship between them is a work in progress. Neural networks have been used in solving engineering problems, with emphasis on applying the technique of backpropagation. It performs satisfactorily iterations between the different variables in a process of training and learning, and has been able to generalize acceptable solutions. In this work is used a feedforward neural network with backpropagation algorithm to predict the compressive strength and modulus of elasticity of the concrete. The input data are quantities of materials used to fabricate 1,0 m3 of concrete hardened, similar dosing for diagrams and abacus. Such diagrams are widely used by companies in the manufacturing of concrete, yielding good precision in the final results. They are produced on the vast experience with the same materials and are highly regional representative to provide subsidies for training neural networks. This... (Complete abstract click electronic access below)

Submitted by RICARDO FERNANDO NUNES null (ricardofnes@gmail.com) on 2016-05-09T19:20:22Z No. of bitstreams: 1 Ricardo_Nunes_Pos_Defesa_FINAL.pdf: 4578185 bytes, checksum: 7d5f2601e5d33327aef16ab69587d77f (MD5) / Approved for entry into archive by Felipe Augusto Arakaki (arakaki@reitoria.unesp.br) on 2016-05-12T19:25:25Z (GMT) No. of bitstreams: 1 nunes_rf_me_ilha.pdf: 4578185 bytes, checksum: 7d5f2601e5d33327aef16ab69587d77f (MD5) / Made available in DSpace on 2016-05-12T19:25:25Z (GMT). No. of bitstreams: 1 nunes_rf_me_ilha.pdf: 4578185 bytes, checksum: 7d5f2601e5d33327aef16ab69587d77f (MD5) Previous issue date: 2016-03-31 / Neste trabalho apresenta-se uma abordagem para o mapeamento da cinemática inversa utilizando Redes Neurais Artificiais do tipo Perceptron Multicamadas na configuração em paralelo, tendo como referência o protótipo de um manipulador robótico de 5 graus de liberdade, composto por sete servomotores controlado pela plataforma de desenvolvimento Intel® Galileo Gen 2. As equações da cinemática inversa, normalmente apresentam múltiplas soluções, desta forma, uma solução interessante e frequentemente encontrada na literatura são as Redes Neurais Artificiais (RNA) em razão da sua flexibilidade e capacidade de aprendizado por meio do treinamento. As Redes Neurais são capazes de entender a relação cinemática entre o sistema de coordenadas das juntas e a posição final da ferramenta do manipulador. Para avaliar a eficiência do método proposto foram realizadas simulações no software MATLAB, as quais demostram pelos resultados obtidos e comparações a uma RNA do tipo MLP simples, aproximadamente redução das médias dos erros das juntas em até 87,8% quando aplicado à trajetória e 80% quando aplicado a pontos distribuídos no volume de trabalho. / This paper presents an approach to the mapping of inverse kinematics using Artificial Neural Networks Multilayer Perceptron in parallel configuration, in the prototype of a robotic manipulator 5 degrees of freedom, as reference, composed of seven servomotors controlled by development board Intel® Galileo Gen 2. The equations of inverse kinematics, usually have multiple solutions, therefore, an interesting solution and often found in the literature are the Artificial Neural Networks (ANN) because of their flexibility and learning capacity through training. Neural Networks are able to understand the kinematic relationship between the coordinate system of the joints and the final position of the manipulator tool. To evaluate the efficiency of the proposed, simulations in MATLAB software are performaded, that demonstrate by the results obtained and compared to a simple MLP type RNA, one reduction in mean errors of the joints by up to 87.8% when applied to the path and 80% when applied to points distributed in the work space.