Video database applications call for flexible and powerful modeling and querying facilities, which require an integration or interaction between database and knowledge base technologies. It is also necessary for many real life video database applications to incorporate uncertainty, which naturally occurs due to the complex and subjective semantic content of video data. In this thesis study, firstly, a fuzzy conceptual data model is introduced to represent the semantic content of video data. UML (Unified Modeling Language) is utilized and extended to represent uncertain information along with video specific properties at the conceptual level. Secondly, an intelligent fuzzy object-oriented database framework is presented for video database applications. The introduced fuzzy conceptual model is mapped to the presented framework, which is an adaptation of the previously proposed IFOOD architecture. The framework provides modeling and querying of complex and rich semantic content and knowledge of video data including uncertainty. Moreover, it allows (fuzzy) semantic, temporal, (fuzzy) spatial, hierarchical, regional and trajectory queries, based on the video data model. We think that the presented conceptual data model and framework can be adapted to any application domain related to video databases.

Data Mining deals with extracting valid, novel, easily understood by humans, potentially useful and general abstractions from large data. A data is large when number of patterns, number of features per pattern or both are large. Largeness of data is characterized by its size which is beyond the capacity of main memory of a computer. Data Mining is an interdisciplinary field involving database systems, statistics, machine learning, visualization and computational aspects. The focus of data mining algorithms is scalability and efficiency. Large data clustering and classification is an important activity in Data Mining. The clustering algorithms are predominantly iterative requiring multiple scans of dataset, which is very expensive when data is stored on the disk. In the current work we propose different schemes that have both theoretical validity and practical utility in dealing with such a large data. The schemes broadly encompass data compaction, classification, prototype selection, use of domain knowledge and hybrid intelligent systems. The proposed approaches can be broadly classified as (a) compressing the data by some means in a non-lossy manner; cluster as well as classify the patterns in their compressed form directly through a novel algorithm, (b) compressing the data in a lossy fashion such that a very high degree of compression and abstraction is obtained in terms of 'distinct subsequences'; classify the data in such compressed form to improve the prediction accuracy, (c) with the help of incremental clustering, a lossy compression scheme and rough set approach, obtain simultaneous prototype and feature selection, (d) demonstrate that prototype selection and data-dependent techniques can reduce number of comparisons in multiclass classification scenario using SVMs, and (e) by making use of domain knowledge of the problem and data under consideration, we show that we obtaina very high classification accuracy with less number of iterations with AdaBoost. The schemes have pragmatic utility. The prototype selection algorithm is incremental, requiring a single dataset scan and has linear time and space requirements. We provide results obtained with a large, high dimensional handwritten(hw) digit data. The compression algorithm is based on simple concepts, where we demonstrate that classification of the compressed data improves computation time required by a factor 5 with prediction accuracy with both compressed and original data being exactly the same as 92.47%. With the proposed lossy compression scheme and pruning methods, we demonstrate that even with a reduction of distinct sequences by a factor of 6 (690 to 106), the prediction accuracy improves. Specifically, with original data containing 690 distinct subsequences, the classification accuracy is 92.47% and with appropriate choice of parameters for pruning, the number of distinct subsequences reduces to 106 with corresponding classification accuracy as 92.92%. The best classification accuracy of 93.3% is obtained with 452 distinct subsequences. With the scheme of simultaneous feature and prototype selection, we improved classification accuracy to better than that obtained with kNNC, viz., 93.58%, while significantly reducing the number of features and prototypes, achieving a compaction of 45.1%. In case of hybrid schemes based on SVM, prototypes and domain knowledge based tree(KB-Tree), we demonstrated reduction in SVM training time by 50% and testing time by about 30% as compared to complete data and improvement of classification accuracy to 94.75%. In case of AdaBoost the classification accuracy is 94.48%, which is better than those obtained with NNC and kNNC on the entire data; the training timing is reduced because of use of prototypes instead of the complete data. Another important aspect of the work is to devise a KB-Tree (with maximum depth of 4), that classifies a 10-category data in just 4 comparisons. In addition to hw data, we applied the schemes to Network Intrusion Detection Data (10% dataset of KDDCUP99) and demonstrated that the proposed schemes provided less overall cost than the reported values.

In this thesis, research efforts are dedicated towards the development of practical adaptable techniques to be used in Smart Homes and Buildings, with the aim to improve energy management and conservation, while enhancing the learning capabilities of Programmable Communicating Thermostats (PCT) – “transforming” them into smart adaptable devices, i.e., “Smart Thermostats”. An Adaptable Hybrid Intelligent System utilizing Wireless Sensor Network (WSN) and Artificial Intelligence (AI) techniques is presented, based on which, a novel Adaptive Learning System (ALS) model utilizing WSN, a rule-based system and Adaptive Resonance Theory (ART) concepts is proposed. The main goal of the ALS is to adapt to the occupant’s pattern and/or schedule changes by providing comfort, while not ignoring the energy conservation aspect. The proposed ALS analytical model is a technique which enables PCTs to learn and adapt to user input pattern changes and/or other parameters of interest. A new algorithm for finding the global maximum in a predefined interval within a two dimensional space is proposed. The proposed algorithm is a synergy of reward/punish concepts from the reinforcement learning (RL) and agent-based technique, for use in small-scale embedded systems with limited memory and/or processing power, such as the wireless sensor/actuator nodes. An application is implemented to observe the algorithm at work and to demonstrate its main features. It was observed that the “RL and Agent-based Search”, versus the “RL only” technique, yielded better performance results with respect to the number of iterations and function evaluations needed to find the global maximum. Furthermore, a “House Simulator” is developed as a tool to simulate house heating/cooling systems and to assist in the practical implementation of the ALS model under different scenarios. The main building blocks of the simulator are the “House Simulator”, the “Smart Thermostat”, and a placeholder for the “Adaptive Learning Models”. As a result, a novel adaptive learning algorithm, “Observe, Learn and Adapt” (OLA) is proposed and demonstrated, reflecting the main features of the ALS model. Its evaluation is achieved with the aid of the “House Simulator”. OLA, with the use of sensors and the application of the ALS model learning technique, captures the essence of an actual PCT reflecting a smart and adaptable device. The experimental performance results indicate adaptability and potential energy savings of the single in comparison to the zone controlled scenarios with the OLA capabilities being enabled.

O uso da tecnologia sensorless é uma tendência crescente para acionamentos industriais aplicados em máquinas elétricas. A estimação dos parâmetros elétricos e mecânicos envolvidos com o controle da máquina elétrica são utilizados freqüentemente para se evitar medir todas as variáveis envolvidas no processo. A redução de custos em acionamentos industriais, além do incremento da robustez do sistema, são algumas das vantagens do uso de técnicas sensorless. Este trabalho propõe o uso de lógica fuzzy para estimar a velocidade de rotação de motores de indução trifásicos. Estão presentes resultados de simulações computacionais e comparação com outras técnicas inteligentes para validação da abordagem apresentada. / The use of sensorless technologies is an increasing tendency on industrial drives for electrical machines. The estimation of electrical and mechanical parameters involved with the electric machine control is used very frequently in order to avoid measurement of all variables from this process. The cost reduction may also be considered in industrial drives, besides the increasing robustness of the system, as advantages of the use of sensorless technologies. This work proposes the use of fuzzy logic to estimate the speed in three-phase induction motors. Simulation results are presented to validate the proposed approach and comparative analyses with other intelligent techniques are also outlined.

Os métodos convencionais para o controle de tensão concentram-se na média tensão. Em alguns casos não são suficientes para a correção da tensão na rede secundária. Este trabalho apresenta os problemas relacionados à regulação de tensão na baixa tensão, os métodos convencionais para correção da tensão e uma estratégia para o controle de tensão na rede secundária de sistema de distribuição de energia elétrica. A solução final proposta é um conjunto de transformador com taps no lado de baixa tensão, hardware e software que promovem a comutação das derivações do transformador de forma automática. Para o desenvolvimento dessa estratégia será abordada a aplicação de sistemas inteligentes, o sistema fuzzy, e a estimação de modelos elétricos dos transformadores de distribuição. O objetivo desse produto é ser uma solução prática, viável técnica e economicamente para a regulação da tensão em cenários onde os métodos convencionais não o são. Os protótipos dessa solução foram montados e testados em laboratório e em campo e os resultados atenderam ao objetivo proposto. / The conventional methods for voltage control concentrated in medium voltage. In some case, they are not enough to correct the voltage on secondary grid of distribuition. This paper presents a strategy to control the voltage on the low voltage of the distribution grid. As proposed for dealing of this problem is made the presentation of an architecture for the intelligent automatic control, that is composed of distribuition transformer with tap on the low side, hardware and software. For the development of this strategy will be approached the application the intelligent systems, Fuzzy Systems, and the estimation of electrical model of distribuition transformers. The goal with this design is to provide grants to set up a system for regulating the voltage on the low side which is technically and economically feasible to be deployed where conventional solutions, with the inclusion of line regulators, are not.

Incluir a Universidade Federal dos Vales do Jequitinhonha e Mucuri (UFVJM) como ag?ncia financiadora. / Submitted by Jos? Henrique Henrique (jose.neves@ufvjm.edu.br) on 2017-12-14T16:46:41Z No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) patrick_aurelio_luiz_ribeiro.pdf: 6159348 bytes, checksum: 5978e3ca5ff417ce94712c998e8c5c8a (MD5) / Approved for entry into archive by Rodrigo Martins Cruz (rodrigo.cruz@ufvjm.edu.br) on 2018-01-03T12:20:58Z (GMT) No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) patrick_aurelio_luiz_ribeiro.pdf: 6159348 bytes, checksum: 5978e3ca5ff417ce94712c998e8c5c8a (MD5) / Made available in DSpace on 2018-01-03T12:20:58Z (GMT). No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) patrick_aurelio_luiz_ribeiro.pdf: 6159348 bytes, checksum: 5978e3ca5ff417ce94712c998e8c5c8a (MD5) Previous issue date: 2017 / Na modalidade de ensino a dist?ncia, os Ambientes Virtuais de Aprendizagem (AVAs) s?o elementos fundamentais no processo de ensino e aprendizagem, atrav?s da disponibiliza??o de conte?dos e ?reas de discuss?o e comunica??o entre os atores do processo. Entretanto, tais ambientes, na sua maioria, caracterizam-se pelo fato de serem est?ticos, abordando m?todos pedag?gicos gen?ricos atrav?s dos quais estudantes com caracter?sticas e Estilos de Aprendizagem (EAs) diferentes buscam o conhecimento. Dessa maneira, ? importante que sejam levados em considera??o os EAs de cada estudante como forma de tornar a aprendizagem mais eficaz. Question?rios psicom?tricos na maioria das vezes s?o utilizados para que as caracter?sticas de aprendizagem do estudante sejam identificadas, por?m nem sempre tais question?rios apresentam resultados precisos quanto ao EAs de determinado estudante. Assim, faz-se necess?ria a utiliza??o de outras t?cnicas de detec??o, haja vista que uma identifica??o precisa ? capaz de melhorar o processo de aprendizagem por meio de escolhas de estrat?gias pedag?gicas melhores. Diante disso, surge a necessidade de utiliza??o de sistemas inteligentes que se adaptem ?s caracter?sticas de aprendizagem do estudante, utilizando como pressupostos as experi?ncias vivenciadas por ele e as an?lises estat?sticas dessas experi?ncias. Isso pode ser feito atrav?s de avalia??es dos EAs apresentados pelo estudante, em que a partir dos resultados um novo modelo de aprendizagem do estudante ? definido para que o conte?do seja disponibilizado de acordo com esse modelo. Nesse intuito a presente abordagem objetivou identificar e corrigir os EAs do estudante por meio da utiliza??o do conceito de M?dia M?vel Exponencialmente Ponderada no processo de decis?o sobre a aplica??o do refor?o de maneira a ajustar o Modelo do Estudante (ME), de modo que os resultados obtidos, ap?s a realiza??o do teste estat?stico n?o-param?trico de Mann-Whitney, mostraram-se significativamente melhores do que os resultados apresentados por Dor?a (2012), cujo trabalho foi refer?ncia para o desenvolvimento desta proposta. / Disserta??o (Mestrado Profissional) ? Programa de P?s-Gradua??o em Educa??o, Universidade Federal dos Vales do Jequitinhonha e Mucuri, 2017. / In Distance Learning, Learning Management Systems (LMS) are extremely important elements in teaching and learning process, because they can offer content and spaces of discussion and comunication between people who are part of that process. However they are static and do not consider students? Learning Styles (LS) to show the content, they just use the same pedagogical methods for all learners. It is important to consider students? Learning Styles because this can make the learning process more efective. Most of the time people use Psychometric Instruments to detect students? preferences, but sometimes the outcomes of those methods are not precise. Because of this other techniques of detection of LS can be used to identify precisely the student?s LS and consequently to choose better pedagogical strategies than when are used manual techniques of detection of LS. For this reason intelligent systems which adapt to students? learning characteristics get importance since they use experiences and statistical analysis over these experiences to be adaptive. It can be done based on learner?s Learning Styles that are adjusted by a part of the system, then these new LS are used by another part of the system to select a pedagogical strategy which fit to student?s characteristics. Thus, this work presents an approach which aimed to identify and to correct the Learning Styles of the learner using for this the Exponentially Weighted Moving Average (EWMA) concept. This concept was used to decide if reinforcement signs have to be used to make the student?s modeling. This approach was tested and the outcomes were submitted to non parametric test Mann-Whitney which pointed they were significantly better than the results of Dor?a (2012), whose work was the base of the work presented here.

Teaching model of the Police Negotiation Process Based on Petri Nets is a study that aims to help the teaching process of strategic negotiation in critical situations involving hostages, allowing the improvement of the Military Police of the State of Alagoas in non-routine police reports. Two types of research were used in the methodology construction of this study: literature research through printed and electronic sources, as well as field research, questionnaires and interviews with military policemen working at the Center of Crisis Management, Human Rights and Community Police (CGCDHPC) in the State of Alagoas. Initially, it discussed the Learning Environment, focusing them as a support system to the learning process. Strategies of persuasion are formally defined in order to be applied in the police negotiation process. We understood the process of strategic negotiation in situations where hostages are involved. We studied different teaching models of the strategic negotiation process. At the end of the study the following results were achieved: proposal for a working model of the persuasion techniques, propose an organizational model of strategic negotiation process and formalization of the model of negotiation process where hostages are involved, based on Petri Nets. / Modelo de Ensino do Processo de Negociação Policial Baseado em Redes de Petri é um estudo que tem como meta auxiliar o ensino do processo de negociação estratégica em situações críticas envolvendo reféns, permitindo o aprimoramento de Policiais Militares do Estado de Alagoas em ocorrências policiais não rotineiras. Na construção metodológica desse estudo foram adotados dois tipos de pesquisas: a pesquisa bibliográfica, mediante fontes impressas e fontes eletrônicas, e a pesquisa de campo, questionários e entrevistas com Policiais Militares lotados no Centro de Gerenciamento de Crise, Direitos Humanos e Polícia Comunitária (CGCDHPC) do Estado de Alagoas. Inicialmente, abordam-se os Ambientes de Aprendizagem, enfocando esses como um sistema de apoio em aprendizagem. Definem-se formalmente estratégias de persuasão com a finalidade de aplicá-las no processo de negociação policial. Compreende-se o processo de negociação estratégica em situações envolvendo reféns. Estudam-se os diversos modelos de ensino do processo de negociação estratégica. Ao final da pesquisa foram alcançados os seguintes resultados: proposta de um modelo de funcionamento das técnicas de persuasão, proposta de um modelo organizacional do processo de negociação estratégica e formalização do modelo do processo de negociação, envolvendo reféns, baseado em Redes de Petri.

Made available in DSpace on 2016-06-02T20:09:21Z (GMT). No. of bitstreams: 1 6223.pdf: 2983823 bytes, checksum: 5c5866e8a883e6646c642579dce1909d (MD5) Previous issue date: 2014-08-11 / Given the importance of forecasting costs in early stages of architectural projects, when it is possible to make changes in the product design and therefore obtain changes in the production costs, and also due to the difficulty of electrical-material consumption prognosis, this research proposes models for predicting electrical-material consumption used in buildings electrical installations. It was used artificial neural networks, an inteligent system, and conventional methods, such as linear regression and consumption rates for the prognostic models. The available data were collected from projects feasibility study and draft design. The research method includes the following steps: a) creation of a database with information collected in quantitatives used for estimates, b) data analysis and preprocessing for use in inteligent and conventional systems, c) attribute selection for best feature identification, i.e, for identifying features with high ability to influence the prognosis and d) development of the models and performance analysis, comparing the predicted values with the actual values. The developed models improves the consumption prognosis performance when compared with common prognostic tools. Current tools consists in multiplying quantitatives by a comsumption rate. Also, the novel models allows more cautious decision-making in projects early design phases, allowing greater awareness of costs impacts. It is expected that this metodology could be used for predicting other building materials. / Devido à importância da previsão de custos nas fases iniciais de empreendimentos, nas quais é possível intervir no produto de forma a obter impactos nos custos de produção; e também, devido à dificuldade de prognóstico do consumo de materiais das instalações prediais elétricas, que ainda não contam com ferramentas eficazes para tal tarefa, este trabalho propõe modelos para prognóstico do consumo de materiais de instalações prediais elétricas a partir de informações disponíveis na fase do estudo de viabilidade e estudo preliminar de empreendimentos. São utilizadas para o prognóstico sistemas inteligentes, mais especificamente, redes neurais artificiais, além de métodos convencionais, como regressão linear e índices de consumo. O método de pesquisa consta das etapas: a) constituição de um banco de dados com informações coletadas em quantitativos utilizados para orçamentos; b) análise e preparação dos dados para a devida utilização dos sistemas inteligentes e convencionais c) seleção de atributos para identificação das características com melhor capacidade de realizar o prognóstico e d) desenvolvimento dos modelos de prognóstico e análise de desempenho por meio de comparações entre os valores prognosticados e os valores reais. Os modelos desenvolvidos melhoram significativamente o prognóstico de consumo destes sistemas em relação a prática atual, que consiste em levantar quantitativos de projeto e multiplicá-los por índices de consumo. Os modelos propostos também permitem tomadas de decisões com maior consciência dos impactos nos custos, principalmente em fases iniciais dos empreendimentos. Espera-se também que a metodologia apresentada possa ser extrapolada para outros serviços de construção.

Made available in DSpace on 2016-06-02T19:05:49Z (GMT). No. of bitstreams: 1 3570.pdf: 760157 bytes, checksum: 38b3438fc00a0ce961ba30506837322f (MD5) Previous issue date: 2009-11-27 / There are many researches being developed in the Production Sequencing and Production Scheduling by the Tear team. In one of the researches focus, it has been investigated the use of simulation in cooperation with Artificial Intelligence to obtain a computer system as the solution of a reactive system in the entrance of production system. The main goal is to have a framework model to an intelligent system to help the products sequencing in the entrance of production systems. The framework should be a guide in the construction and customization of the intelligent system to be constructed. / No Tear (Laboratório de Pesquisa e Desenvolvimento de Tecnologia e Estratégias de Automação) vários trabalhos vêm sendo desenvolvidos na temática de seqüenciamento e programação da produção. Em um dos focos, o de seqüenciamento de produtos na entrada de um sistema produtivo, tem-se investigado o uso de simulação em cooperação com técnicas de Inteligência Artificial buscando-se uma aplicação como solução para condições de re-seqüenciamento ou de seqüenciamento reativo da entrada do sistema produtivo. A proposta deste trabalho é modelar um framework para um sistema de auxílio ao seqüenciamento na entrada do sistema produtivo. O framework deve servir como guia na construção e customização do sistema.

Submitted by Bruno Albuquerque de Castro null (bruno.castro@feb.unesp.br) on 2016-06-01T17:52:34Z No. of bitstreams: 1 Bruno_Albuquerque_de_Castro_MestradoEE.pdf: 3788528 bytes, checksum: 28b69717116921fa999eda08254ac160 (MD5) / Approved for entry into archive by Ana Paula Grisoto (grisotoana@reitoria.unesp.br) on 2016-06-02T19:10:47Z (GMT) No. of bitstreams: 1 castro_ba_me_bauru.pdf: 3788528 bytes, checksum: 28b69717116921fa999eda08254ac160 (MD5) / Made available in DSpace on 2016-06-02T19:10:47Z (GMT). No. of bitstreams: 1 castro_ba_me_bauru.pdf: 3788528 bytes, checksum: 28b69717116921fa999eda08254ac160 (MD5) Previous issue date: 2016-05-25 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / O monitoramento e a localização precoce de descargas parciais em aparelhos de alta tensão, como nos transformadores de potência, são de fundamental importância para a prevenção de problemas funcionais associados à degradação contínua dos materiais que compõe o isolamento elétrico destes tipos de aparelho. Alguns fatores críticos na operação dos transformadores, como o funcionamento em sobrecarga, superaquecimento, transitórios e sinais elétricos de grande conteúdo harmônico vinculados a sistemas chaveados, em longo prazo, fazem com que o sistema de isolação de um transformador apresente degradação de suas propriedades físicas e químicas intrínsecas aos diversos tipos de materiais utilizados para esta finalidade e, deste modo, surjam descargas parciais. Este trabalho teve como objetivo aplicar sensores piezelétricos de baixo custo para a identificação e localização de descargas parciais em transformadores de potência por meio de sistemas inteligentes do tipo Redes Neurais e sistema de inferência neuro fuzzy adaptativos. Ambos os sistemas foram treinados com algumas métricas de processamento de sinais e os resultados de erro médio de localização chegaram na casa dos milímetros. Variou-se o número de sensores acoplados e foi realizado um estudo sobre os resultados de localização obtidos. / Partial discharge damages in power transformers require high cost monitoring procedures based on corrective maintenance or even interruptions of the power system. The development of online non-invasive monitoring systems to detect partial discharges in power transformers has great relevance since it can reduce significant maintenance costs. Some critical factors in the operation of transformers such as overload, nonlinear loads, transient voltage surges by atmospheric origin and switching, can make the insulation system of transformers to lose their physical and chemical properties. Therefore, these operating conditions can cause early deterioration of the insulation, causing internal partial discharges that may develop into major defects and thus shorten the useful life of electrical equipment. This research aimed to apply a low cost piezoelectric sensors for partial discharge identification and location in power transformers through intelligent systems such as neural networks and adaptive fuzzy inference system. Both systems were trained with some signal processing metrics and the results for location error was in the region of millimeters. It was varied the number of coupled sensors and a study was conducted on the obtained location results.