Parameters Inversion of Fluid-Saturated Porous Media Based on Neural Networks

The multi-layers feedforward neural network is used for inversion of material constants of fluid-saturated porous media. The direct analysis of fluid-saturated porous media is carried out with the boundary element method. The dynamic displacement responses obtained from direct analysis for prescribed material parameters constitute the sample sets training neural network. By virtue of the effective L-M training algorithm and the Tikhonov regularization method as well as the GCV method for an appropriate selection of regularization parameter, the inverse mapping from dynamic displacement responses to material constants is performed. Numerical examples demonstrate the validity of the neural network method.

Coupled plasmonic systems and devices : applications in visible metamaterials, nanophotonic circuits, and CMOS imaging

With the size of transistors approaching the sub-nanometer scale and Si-based photonics pinned at the micrometer scale due to the diffraction limit of light, we are unable to easily integrate the high transfer speeds of this comparably bulky technology with the increasingly smaller architecture of state-of-the-art processors. However, we find that we can bridge the gap between these two technologies by directly coupling electrons to photons through the use of dispersive metals in optics. Doing so allows us to access the surface electromagnetic wave excitations that arise at a metal/dielectric interface, a feature which both confines and enhances light in subwavelength dimensions - two promising characteristics for the development of integrated chip technology. This platform is known as plasmonics, and it allows us to design a broad range of complex metal/dielectric systems, all having different nanophotonic responses, but all originating from our ability to engineer the system surface plasmon resonances and interactions. In this thesis, we demonstrate how plasmonics can be used to develop coupled metal-dielectric systems to function as tunable plasmonic hole array color filters for CMOS image sensing, visible metamaterials composed of coupled negative-index plasmonic coaxial waveguides, and programmable plasmonic waveguide network systems to serve as color routers and logic devices at telecommunication wavelengths.

Optical interconnection for neural networks by use of a self-imaging function

A grating-lens combination unit is developed to form a scaling self-transform function that can self-image on scale. Then an array of many such grating-lens units is used for the optical interconnection of a two-dimensional neural network, and experiments are carried out. We find that our idea is feasible, the optical interconnection system is simple, and optical adjustment is easy. (C) 1998 Optical Society of America.

Identificação da região periorbital utilizando a técnica SIFT em conjunto com rede neural híbrida

Nesta dissertação, foi utilizada a técnica SIFT (Scale Invariant Feature Transform) para o reconhecimento de imagens da área dos olhos (região periorbital). Foi implementada uma classificação das imagens em subgrupos internos ao banco de dados, utilizando-se das informações estatísticas provenientes dos padrões invariantes produzidos pela técnica SIFT. Procedeu-se a uma busca categorizada pelo banco de dados, ao invés da procura de um determinado padrão apresentado, através da comparação deste com cada padrão presente no banco de dados. A tais padrões foi aplicada uma abordagem estatística, através da geração da matriz de covariâncias dos padrões gerados, sendo esta utilizada para a categorização, tendo por base uma rede neural híbrida. A rede neural classifica e categoriza o banco de dados de imagens, criando uma topologia de busca. Foram obtidos resultados corretos de classificação de 76,3% pela rede neural híbrida, sendo que um algoritmo auxiliar determina uma hierarquia de busca, onde, ocorrendo uma errônea classificação, a busca segue em grupos de pesquisas mais prováveis.

Estimating relative abundance from catch and effort data, using neural networks

We develop and test a method to estimate relative abundance from catch and effort data using neural networks. Most stock assessment models use time series of relative abundance as their major source of information on abundance levels. These time series of relative abundance are frequently derived from catch-per-unit-of-effort (CPUE) data, using general linearized models (GLMs). GLMs are used to attempt to remove variation in CPUE that is not related to the abundance of the population. However, GLMs are restricted in the types of relationships between the CPUE and the explanatory variables. An alternative approach is to use structural models based on scientific understanding to develop complex non-linear relationships between CPUE and the explanatory variables. Unfortunately, the scientific understanding required to develop these models may not be available. In contrast to structural models, neural networks uses the data to estimate the structure of the non-linear relationship between CPUE and the explanatory variables. Therefore neural networks may provide a better alternative when the structure of the relationship is uncertain. We use simulated data based on a habitat based-method to test the neural network approach and to compare it to the GLM approach. Cross validation and simulation tests show that the neural network performed better than nominal effort and the GLM approach. However, the improvement over GLMs is not substantial. We applied the neural network model to CPUE data for bigeye tuna (Thunnus obesus) in the Pacific Ocean.

Exploring the use of neural networks for biomass forecasts in the Peruvian upwelling ecosystem

A pilot study was conducted to study the ability of an artificial neural network to predict the biomass of Peruvian anchoveta Engraulis ringens, given time series of earlier biomasses, and of environmental parameters (ocenographic data and predator abundances). Acceptable predictions of three months or more appear feasible after thorough scrutiny of the input data set.

Localização de faltas de curta duração em redes de distribuição.

O objetivo deste trabalho é contribuir com o desenvolvimento de uma técnica baseada em sistemas inteligentes que possibilite a localização exata ou aproximada do ponto de origem de uma Variação de Tensão de Curta Duração (VTCD) (gerada por uma falta) em um sistema de distribuição de energia elétrica. Este trabalho utiliza um Phase-Locked Loop (PLL) com o intuito de detectar as faltas. Uma vez que a falta é detectada, os sinais de tensão obtidos durante a falta são decompostos em componentes simétricas instantâneas por meio do método proposto. Em seguida, as energias das componentes simétricas são calculadas e utilizadas para estimar a localização da falta. Nesta pesquisa, são avaliadas duas estruturas baseadas em Redes Neurais Artificiais (RNAs). A primeira é projetada para classificar a localização da falta em um dos pontos possíveis e a segunda é projetada para estimar a distância da falta ao alimentador. A técnica aqui proposta aplica-se a alimentadores trifásicos com cargas equilibradas. No desenvolvimento da mesma, considera-se que há disponibilidade de medições de tensões no nó inicial do alimentador e também em pontos esparsos ao longo da rede de distribuição. O banco de dados empregado foi obtido através de simulações de um modelo de alimentador radial usando o programa PSCAD/EMTDC. Testes de sensibilidade empregando validação-cruzada são realizados em ambas as arquiteturas de redes neurais com o intuito de verificar a confiabilidade dos resultados obtidos. Adicionalmente foram realizados testes com faltas não inicialmente contidas no banco de dados a fim de se verificar a capacidade de generalização das redes. Os desempenhos de ambas as arquiteturas de redes neurais foram satisfatórios e demonstram a viabilidade das técnicas propostas para realizar a localização de faltas em redes de distribuição.

Sistemas inteligentes adaptativos aplicados a um robô auto-equilibrante de duas rodas.

Com o avanço no desenvolvimento e utilização de veículos e robôs autoequilibrantes, faz-se necessário a investigação de controladores capazes de atender os diversos desafios relacionados à utilização desses sistemas. Neste trabalho foi estudado o controle de equilíbrio e posição de um robô auto-equilibrante de duas rodas. O interesse particular nesta aplicação vem da sua estrutura e da riqueza de sua dinâmica física. Por ser um problema complexo e não trivial há grande interesse em avaliar os controladores inteligentes. A primeira parte da dissertação aborda o desenvolvimento de um controle clássico do tipo PID, para em seguida ser comparado com a implementação de dois tipos de controladores inteligentes: On-line Neuro Fuzzy Control (ONFC) e Proportional-Integral-Derivative Neural-Network (PIDNN). Também é apresentada a implementação dos controladores em uma plataforma de hardware, utilizando o kit LEGO Mindstorm, e numa plataforma de simulação utilizando o MATLAB-Simulink. Em seguida, dois estudos de casos são desenvolvidos visando comparar o desempenho dos controladores. O primeiro caso avalia o controle de equilíbrio e posição do robô auto-equilibrante de duas rodas sobre um terreno plano tendo como interesse observar o desempenho intrínseco do sistema sob ausência de fatores externos. O segundo caso estuda o controle de equilíbrio e posição do robô em terrenos irregulares visando investigar a resposta do sistema sob influência de condições adversas em seu ambiente. Finalmente, o desempenho de cada um dos controladores desenvolvidos é discutido, verificando-se resultados competitivos no controle do robô auto-equilibrante de duas rodas.

Application of neural networks to the ranking of perinatal variables influencing birthweight.

In this paper we compare Multi-Layer Perceptrons (a neural network type) with Multivariate Linear Regression in predicting birthweight from nine perinatal variables which are thought to be related. Results show, that seven of the nine variables, i.e., gestational age, mother's body-mass index (BMI), sex of the baby, mother's height, smoking, parity and gravidity, are related to birthweight. We found no significant relationship between birthweight and each of the two variables, i.e., maternal age and social class.

Development of low entropy coding in a recurrent network.

In this paper we present an unsupervised neural network which exhibits competition between units via inhibitory feedback. The operation is such as to minimize reconstruction error, both for individual patterns, and over the entire training set. A key difference from networks which perform principal components analysis, or one of its variants, is the ability to converge to non-orthogonal weight values. We discuss the network's operation in relation to the twin goals of maximizing information transfer and minimizing code entropy, and show how the assignment of prior probabilities to network outputs can help to reduce entropy. We present results from two binary coding problems, and from experiments with image coding.

Language model cross adaptation for LVCSR system combination

State-of-the-art large vocabulary continuous speech recognition (LVCSR) systems often combine outputs from multiple sub-systems that may even be developed at different sites. Cross system adaptation, in which model adaptation is performed using the outputs from another sub-system, can be used as an alternative to hypothesis level combination schemes such as ROVER. Normally cross adaptation is only performed on the acoustic models. However, there are many other levels in LVCSR systems' modelling hierarchy where complimentary features may be exploited, for example, the sub-word and the word level, to further improve cross adaptation based system combination. It is thus interesting to also cross adapt language models (LMs) to capture these additional useful features. In this paper cross adaptation is applied to three forms of language models, a multi-level LM that models both syllable and word sequences, a word level neural network LM, and the linear combination of the two. Significant error rate reductions of 4.0-7.1% relative were obtained over ROVER and acoustic model only cross adaptation when combining a range of Chinese LVCSR sub-systems used in the 2010 and 2011 DARPA GALE evaluations. © 2012 Elsevier Ltd. All rights reserved.

An example of the use of neural computing techniques in materials science - The modelling of fatigue thresholds in Ni-base superalloys

Two adaptive numerical modelling techniques have been applied to prediction of fatigue thresholds in Ni-base superalloys. A Bayesian neural network and a neurofuzzy network have been compared, both of which have the ability to automatically adjust the network's complexity to the current dataset. In both cases, despite inevitable data restrictions, threshold values have been modelled with some degree of success. However, it is argued in this paper that the neurofuzzy modelling approach offers real benefits over the use of a classical neural network as the mathematical complexity of the relationships can be restricted to allow for the paucity of data, and the linguistic fuzzy rules produced allow assessment of the model without extensive interrogation and examination using a hypothetical dataset. The additive neurofuzzy network structure means that redundant inputs can be excluded from the model and simple sub-networks produced which represent global output trends. Both of these aspects are important for final verification and validation of the information extracted from the numerical data. In some situations neurofuzzy networks may require less data to produce a stable solution, and may be easier to verify in the light of existing physical understanding because of the production of transparent linguistic rules. © 1999 Elsevier Science S.A.

Preliminary model development for predicting strength and stiffness of cement-stabilized soils using artificial neural networks

This paper presents ongoing work on data collection and collation from a large number of laboratory cement-stabilization projects worldwide. The aim is to employ Artificial Neural Networks (ANN) to establish relationships between variables, which define the properties of cement-stabilized soils, and the two parameters determined by the Unconfined Compression Test, the Unconfined Compressive Strength (UCS), and stiffness, using E50 calculated from UCS results. Bayesian predictive neural network models are developed to predict the UCS values of cement-stabilized inorganic clays/silts, as well as sands as a function of selected soil mix variables, such as grain size distribution, water content, cement content and curing time. A model which can predict the stiffness values of cement-stabilized clays/silts is also developed and compared to the UCS model. The UCS model results emulate known trends better and provide more accurate estimates than the results from the E50 stiffness model. © 2013 American Society of Civil Engineers.