Contribuição para o estudo do embarque de uma rede neural artificial em field programmable gate array (FPGA)

This study shows the implementation and the embedding of an Artificial Neural Network (ANN) in hardware, or in a programmable device, as a field programmable gate array (FPGA). This work allowed the exploration of different implementations, described in VHDL, of multilayer perceptrons ANN. Due to the parallelism inherent to ANNs, there are disadvantages in software implementations due to the sequential nature of the Von Neumann architectures. As an alternative to this problem, there is a hardware implementation that allows to exploit all the parallelism implicit in this model. Currently, there is an increase in use of FPGAs as a platform to implement neural networks in hardware, exploiting the high processing power, low cost, ease of programming and ability to reconfigure the circuit, allowing the network to adapt to different applications. Given this context, the aim is to develop arrays of neural networks in hardware, a flexible architecture, in which it is possible to add or remove neurons, and mainly, modify the network topology, in order to enable a modular network of fixed-point arithmetic in a FPGA. Five synthesis of VHDL descriptions were produced: two for the neuron with one or two entrances, and three different architectures of ANN. The descriptions of the used architectures became very modular, easily allowing the increase or decrease of the number of neurons. As a result, some complete neural networks were implemented in FPGA, in fixed-point arithmetic, with a high-capacity parallel processing

Reconhecimento de padrões de falhas em motores trifásicos utilizando redes neurais

This work presents a diagnosis faults system (rotor, stator, and contamination) of three-phase induction motor through equivalent circuit parameters and using techniques patterns recognition. The technology fault diagnostics in engines are evolving and becoming increasingly important in the field of electrical machinery. The neural networks have the ability to classify non-linear relationships between signals through the patterns identification of signals related. It is carried out induction motor´s simulations through the program Matlab R & Simulink R , and produced some faults from modifications in the equivalent circuit parameters. A system is implemented with multiples classifying neural network two neural networks to receive these results and, after well-trained, to accomplish the identification of fault´s pattern

Sistema inteligente para detecção de vazamentos em dutos de petróleo usando transformada Wavelet e redes neurais

This work consists in the use of techniques of signals processing and artificial neural networks to identify leaks in pipes with multiphase flow. In the traditional methods of leak detection exists a great difficulty to mount a profile, that is adjusted to the found in real conditions of the oil transport. These difficult conditions go since the unevenly soil that cause columns or vacuum throughout pipelines until the presence of multiphases like water, gas and oil; plus other components as sand, which use to produce discontinuous flow off and diverse variations. To attenuate these difficulties, the transform wavelet was used to map the signal pressure in different resolution plan allowing the extraction of descriptors that identify leaks patterns and with then to provide training for the neural network to learning of how to classify this pattern and report whenever this characterize leaks. During the tests were used transient and regime signals and pipelines with punctures with size variations from ½' to 1' of diameter to simulate leaks and between Upanema and Estreito B, of the UN-RNCE of the Petrobras, where it was possible to detect leaks. The results show that the proposed descriptors considered, based in statistical methods applied in domain transform, are sufficient to identify leaks patterns and make it possible to train the neural classifier to indicate the occurrence of pipeline leaks

Técnicas inteligentes hídridas para o controle de sistemas não lineares

A neuro-fuzzy system consists of two or more control techniques in only one structure. The main characteristic of this structure is joining one or more good aspects from each technique to make a hybrid controller. This controller can be based in Fuzzy systems, artificial Neural Networks, Genetics Algorithms or rein forced learning techniques. Neuro-fuzzy systems have been shown as a promising technique in industrial applications. Two models of neuro-fuzzy systems were developed, an ANFIS model and a NEFCON model. Both models were applied to control a ball and beam system and they had their results and needed changes commented. Choose of inputs to controllers and the algorithms used to learning, among other information about the hybrid systems, were commented. The results show the changes in structure after learning and the conditions to use each one controller based on theirs characteristics

Estimação da demanda de instalações elétricas de edificações com múltiplas unidades de consumo

This work aims to predict the total maximum demand of a transformer that will be used in power systems to attend a Multiple Unit Consumption (MUC) in design. In 1987, COSERN noted that calculation of maximum total demand for a building should be different from that which defines the scaling of the input protection extension in order to not overestimate the power of the transformer. Since then there have been many changes, both in consumption habits of the population, as in electrical appliances, so that this work will endeavor to improve the estimation of peak demand. For the survey, data were collected for identification and electrical projects in different MUCs located in Natal. In some of them, measurements were made of demand for 7 consecutive days and adjusted for an integration interval of 30 minutes. The estimation of the maximum demand was made through mathematical models that calculate the desired response from a set of information previously known of MUCs. The models tested were simple linear regressions, multiple linear regressions and artificial neural networks. The various calculated results over the study were compared, and ultimately, the best answer found was put into comparison with the previously proposed model

Localização de um robô móvel usando odometria e marcos naturais

Several methods of mobile robot navigation request the mensuration of robot position and orientation in its workspace. In the wheeled mobile robot case, techniques based on odometry allow to determine the robot localization by the integration of incremental displacements of its wheels. However, this technique is subject to errors that accumulate with the distance traveled by the robot, making unfeasible its exclusive use. Other methods are based on the detection of natural or artificial landmarks present in the environment and whose location is known. This technique doesnt generate cumulative errors, but it can request a larger processing time than the methods based on odometry. Thus, many methods make use of both techniques, in such a way that the odometry errors are periodically corrected through mensurations obtained from landmarks. Accordding to this approach, this work proposes a hybrid localization system for wheeled mobile robots in indoor environments based on odometry and natural landmarks. The landmarks are straight lines de.ned by the junctions in environments floor, forming a bi-dimensional grid. The landmark detection from digital images is perfomed through the Hough transform. Heuristics are associated with that transform to allow its application in real time. To reduce the search time of landmarks, we propose to map odometry errors in an area of the captured image that possesses high probability of containing the sought mark

Aplicação de técnicas de aprendizado de máquina no reconhecimento de classes estruturais de proteínas

Nowadays, classifying proteins in structural classes, which concerns the inference of patterns in their 3D conformation, is one of the most important open problems in Molecular Biology. The main reason for this is that the function of a protein is intrinsically related to its spatial conformation. However, such conformations are very difficult to be obtained experimentally in laboratory. Thus, this problem has drawn the attention of many researchers in Bioinformatics. Considering the great difference between the number of protein sequences already known and the number of three-dimensional structures determined experimentally, the demand of automated techniques for structural classification of proteins is very high. In this context, computational tools, especially Machine Learning (ML) techniques, have become essential to deal with this problem. In this work, ML techniques are used in the recognition of protein structural classes: Decision Trees, k-Nearest Neighbor, Naive Bayes, Support Vector Machine and Neural Networks. These methods have been chosen because they represent different paradigms of learning and have been widely used in the Bioinfornmatics literature. Aiming to obtain an improvment in the performance of these techniques (individual classifiers), homogeneous (Bagging and Boosting) and heterogeneous (Voting, Stacking and StackingC) multiclassification systems are used. Moreover, since the protein database used in this work presents the problem of imbalanced classes, artificial techniques for class balance (Undersampling Random, Tomek Links, CNN, NCL and OSS) are used to minimize such a problem. In order to evaluate the ML methods, a cross-validation procedure is applied, where the accuracy of the classifiers is measured using the mean of classification error rate, on independent test sets. These means are compared, two by two, by the hypothesis test aiming to evaluate if there is, statistically, a significant difference between them. With respect to the results obtained with the individual classifiers, Support Vector Machine presented the best accuracy. In terms of the multi-classification systems (homogeneous and heterogeneous), they showed, in general, a superior or similar performance when compared to the one achieved by the individual classifiers used - especially Boosting with Decision Tree and the StackingC with Linear Regression as meta classifier. The Voting method, despite of its simplicity, has shown to be adequate for solving the problem presented in this work. The techniques for class balance, on the other hand, have not produced a significant improvement in the global classification error. Nevertheless, the use of such techniques did improve the classification error for the minority class. In this context, the NCL technique has shown to be more appropriated

Análise de desempenho da rede neural artificial do tipo multilayer perceptron na era multicore

Artificial neural networks are usually applied to solve complex problems. In problems with more complexity, by increasing the number of layers and neurons, it is possible to achieve greater functional efficiency. Nevertheless, this leads to a greater computational effort. The response time is an important factor in the decision to use neural networks in some systems. Many argue that the computational cost is higher in the training period. However, this phase is held only once. Once the network trained, it is necessary to use the existing computational resources efficiently. In the multicore era, the problem boils down to efficient use of all available processing cores. However, it is necessary to consider the overhead of parallel computing. In this sense, this paper proposes a modular structure that proved to be more suitable for parallel implementations. It is proposed to parallelize the feedforward process of an RNA-type MLP, implemented with OpenMP on a shared memory computer architecture. The research consistes on testing and analizing execution times. Speedup, efficiency and parallel scalability are analyzed. In the proposed approach, by reducing the number of connections between remote neurons, the response time of the network decreases and, consequently, so does the total execution time. The time required for communication and synchronization is directly linked to the number of remote neurons in the network, and so it is necessary to investigate which one is the best distribution of remote connections

Contribuições a técnicas de agrupamento e visualização de dados multivariados utilizando mapas auto-organizáveis

Self-organizing maps (SOM) are artificial neural networks widely used in the data mining field, mainly because they constitute a dimensionality reduction technique given the fixed grid of neurons associated with the network. In order to properly the partition and visualize the SOM network, the various methods available in the literature must be applied in a post-processing stage, that consists of inferring, through its neurons, relevant characteristics of the data set. In general, such processing applied to the network neurons, instead of the entire database, reduces the computational costs due to vector quantization. This work proposes a post-processing of the SOM neurons in the input and output spaces, combining visualization techniques with algorithms based on gravitational forces and the search for the shortest path with the greatest reward. Such methods take into account the connection strength between neighbouring neurons and characteristics of pattern density and distances among neurons, both associated with the position that the neurons occupy in the data space after training the network. Thus, the goal consists of defining more clearly the arrangement of the clusters present in the data. Experiments were carried out so as to evaluate the proposed methods using various artificially generated data sets, as well as real world data sets. The results obtained were compared with those from a number of well-known methods existent in the literature

Air-assisted boom sprayer and spray deposition on bean plants

A possibilidade do desenvolvimento de técnicas de aplicação de produtos fitossanitários mais seguras, com menores volumes de calda, número de aplicações e deriva, aliados à necessidade de se obter melhores níveis de controle dos agentes nocivos às plantas cultivadas, justificam o uso da assistência de ar junto à barra de pulverização. Com o objetivo de avaliar a deposição da pulverização na cultura do feijoeiro (Phaseolus vulgaris), em presença e ausência da assistência de ar junto à barra de pulverização, com diferentes pontas de pulverização e volumes de calda, foi conduzido um experimento em delineamento inteiramente casualizado, utilizando-se como traçador o óxido cuproso. Alvos artificiais (papel filtro com 3 x 3 cm) foram afixados nas superfícies adaxial e abaxial de folíolos posicionados nos terços superior e inferior de plantas, selecionadas ao acaso, distribuídas perpendicularmente ao deslocamento do pulverizador. Após a aplicação do traçador os coletores foram lavados individualmente em solução extratora de ácido nítrico a 1,0 mol L-1. A determinação quantitativa dos depósitos foi realizada com o uso da espectrofotometria de absorção atômica. A assistência de ar junto à barra de pulverização não aumentou a deposição do traçador em folíolos de feijoeiro, aos 48 dias após a emergência da cultura.

Equalização neural aplicada a sistemas com modulação bidimensional em fibra óptica

Nowadays, optic fiber is one of the most used communication methods, mainly due to the fact that the data transmission rates of those systems exceed all of the other means of digital communication. Despite the great advantage, there are problems that prevent full utilization of the optical channel: by increasing the transmission speed and the distances involved, the data is subjected to non-linear inter symbolic interference caused by the dispersion phenomena in the fiber. Adaptive equalizers can be used to solve this problem, they compensate non-ideal responses of the channel in order to restore the signal that was transmitted. This work proposes an equalizer based on artificial neural networks and evaluates its performance in optical communication systems. The proposal is validated through a simulated optic channel and the comparison with other adaptive equalization techniques

Influência do estresse hídrico e salino na germinação de urochloa decumbens e urochloa ruziziensis

Studies of the germination response of seeds subjected to artificial stresses are provided tools for better understanding of the survivability and adaptation of these species in natural stress conditions such as drought or saline soils, common in agricultural and forest regions, contributing significantly to the development of management strategies. Thus, the purpose of this study was to evaluate the possible effects of water and salt stress on germination of Urochloa decumbens and Urochloa ruziziensis. The test was conducted at the Faculty of Technology of São Paulo, campus of Capon Bonito. The seeds were sown with four replicates of 50 seeds in paper soaked in solutions with the potentials of 0.0, -0.2, -0.4 and -0.8 MPa, induced with polyethylene glycol (PEG 6000) and NaCl. The germination test was conducted at 25 degrees C in the presence of light, evaluating the first test score at seven days after sowing, and weekly germination (normal seedlings) until 35 days. We calculated the index of germination rate. The results allowed the conclusion that water stress causes a greater reduction in force, speed of germination and cumulative germination of seeds of U. decumbens and U. ruziziensis than salt stress. The species U. decumbens showed higher tolerance to water and salt stresses.

Velocidade do ar em barra de pulverização na deposição da calda fungicida, severidade da ferrugem asiática e produtividade da soja

O trabalho objetivou comparar a deposição da pulverização e o controle da ferrugem asiática após o tratamento com fungicidas sob quatro velocidades da assistência de ar junto à barra de pulverização na cultura da soja. Dois experimentos foram conduzidos na FCA/UNESP - Campus de Botucatu, safra 2006/07. Alvos artificiais foram fixados na superfície adaxial e abaxial de folíolos posicionados nas partes superior e inferior das plantas selecionadas e distribuídas perpendicularmente ao deslocamento do pulverizador. O oxicloreto de cobre (50% de cobre metálico) foi o marcador utilizado em pulverização e a determinação quantitativa dos depósitos feita com o uso de espectrofotometria de absorção atômica. Após a aplicação do fungicida piraclostrobina + epoxiconazole sob diferentes velocidades da assistência de ar junto à barra de pulverização (0, 9, 11 e 29 km h-1) procedeu-se a avaliação da severidade da doença e produtividade da soja. Na parte superior das plantas os maiores níveis de depósitos foram encontrados na pulverização sem assistência de ar. Já na parte inferior da planta foram encontrados os maiores níveis de deposição quando foram utilizadas as maiores velocidades da assistência de ar. No geral, a severidade da doença foi mais acentuada nos tratamentos sem o uso da assistência de ar. em relação à produtividade não houve diferenças entre os tratamentos com aplicação de fungicidas, porém houve incremento na produtividade para os tratamentos com assistência de ar.

Desempenho de pontas de pulverização em Brachiaria brizantha cv. MG-4 para controle de ninfas de cigarrinhas das pastagens

O trabalho teve como objetivo estudar o desempenho de pontas de pulverização na deposição da calda inseticida para o controle de ninfas de cigarrinhas das pastagens em Brachiaria brizantha cv. MG-4. Doze tratamentos foram estudados em esquema fatorial 6x2, constituídos pelo contraste de seis pontas de pulverização e pressões de 196 e 392 kPa: TF-VP2 (336 L ha-1 e 467 L ha-1); AI11002-VS (184 L ha-1 e 200 L ha-1); XR11002-VS (200 L ha-1 e 280 L ha-1); TT11002-VP (200 L ha-1 e 280 L ha-1); TJ60-11002VS (208 L ha-1 e 280 L ha-1) e TX-VK4 (72 L ha-1 e 97 L ha-1). Para monitorar a deposição das caldas de pulverização, utilizaram-se os traçadores Azul Brilhante FD&C-1 (0,3% p/v) e Amarelo de Tartrasina FD&C-5 (0,6% p/v). Alvos artificiais, constituídos de lâminas de vidro, foram posicionados na base das plantas, próximos à superfície do solo, e os depósitos por unidade de área das soluções pulverizadas foram quantificados por espectrofotometria. As pontas TF-VP2, XR11002-VS e AI11002-VS, nas pressões de 196 e 392 kPa, proporcionam as maiores deposições da calda de pulverização na região das espumas das cigarrinhas das pastagens, apesar de apresentarem menor uniformidade na distribuição dos depósitos em relação a TX-VK4, XR110.02-VS e TJ110.02-VS. O aumento da pressão de 196 para 392 kPa promoveu aumento na deposição da calda de pulverização sobre a Brachiaria brizantha e na região onde se encontram as espumas das cigarrinhas para todos os tipos de pontas estudadas.

Aspectos metodológicos para análise de depósitos de pulverizações pela determinação dos depósitos pontuais

Objetivou-se com este trabalho avaliar a variabilidade dos depósitos de traçadores, simulando herbicidas aplicados em pós-emergência, em populações de Brachiaria plantaginea e Commelina benghalensis infestantes da cultura da soja. Os depósitos dos traçadores foram também avaliados em plantas da cultura, utilizando-se o pulverizador de barra tratorizado, com pontas de jato plano da série 110-SF-03, aplicando o volume de 250 L ha-1 de calda preparada com 0,18% de corante Azul Brilhante e 0,18% de Amarelo Saturn Yellow. Os alvos naturais utilizados foram: plantas de soja com 150 repetições; B. plantaginea no estádio de duas a oito folhas, coletadas na linha da cultura com 141 repetições; e B. plantaginea e C. benghalensis nas entrelinhas, com 150 e 50 repetições, respectivamente. Os alvos artificiais foram constituídos por lâminas distribuídas a 0, 12,2 e 22,5 cm da linha da cultura. Após a aplicação, os alvos foram coletados individualmente e lavados com 30, 20 e 15 mL de água deionizada, para soja, lâminas e plantas daninhas, respectivamente. Estas originaram as amostras analisadas em espectrofotômetro, estimando-se o depósito de calda em µL por planta e µL cm² de área foliar. Foram ajustadas curvas de regressão entre os depósitos unitários e as freqüências acumuladas, utilizando-se o modelo de Gompertz. As relações entre os depósitos máximos e mínimos foram de 7, 4, 10 e 6 para soja, C. benghalensis e B. plantaginea na linha e na entrelinha, respectivamente. As plantas de B. plantaginea da entrelinha receberam, em média, 34% a mais de depósito do que as plantas da linha.