Orientador: Odivaldo José Seraphim / Banca: Luiz Francisco da Cruz / Banca: Eduardo Antonio Sleiman / Resumo: O uso de energias alternativas e renováveis tem sido cada vez mais discutido em todos os setores da sociedade. O interesse nessas fontes de energias, alternativas e renováveis, é de fundamental importância, tanto no que se refere à escassez, quanto ao preço do petróleo, além das questões ambientais envolvidas no uso dessas energias. Sistemas de geração eólica e fotovoltaica com armazenamento de energia em baterias tem-se apresentado como uma forma de geração alternativa de energia. A variabilidade na intensidade da energia eólica e solar pode ser contornada pela complementação entre uma fonte e outra, ou pela maior estabilidade configurada à geração do sistema. O objetivo desta dissertação é criar uma metodologia fuzzy e simular seu uso no controle inteligente de um sistema híbrido de geração de energia elétrica, utilizando as energias solar-fotovoltaica e eólica. Quando utilizamos um sistema de controle baseado na lógica fuzzy, é atingido o ponto de máxima geração de energia, desta forma transferindo toda a energia gerada à partir das fontes alternativas, solar-fotovoltaica e eólica, à carga e/ou as baterias quando seu uso não imediato. O modelo utilizado adota três variáveis de entrada, que são: velocidade do vento, radiação solar e carga do banco de baterias.Serão utilizados para as simulações alguns softwares como o MATLAB e outros, que serão citados ao longo do trabalho. Nestes ambientes serão analisadas e simuladas todas as modelagens matemáticas, regras e demais variáveis descritas no sistema fuzzy. Este modelo foi utilizado para a implementação de um sistema de controle de sistemas híbridos de geração de energia, proporcionando o melhor aproveitamento das fontes de energia, sol e vento, de modo que possamos extrair o máximo de energia possível dessas fontes alternativas sem nenhum prejuízo ao meio ambiente...(Resumo completo, clicar acesso eletrônico abaixo) / Abstract: The use of alternative and renewable energy has been increasingly discussed in all sectors of society. The interest in these sources of energy, alternative and renewable, it is of fundamental importance, both in terms of shortage, as the price of oil, in addition to the environmental issues involved in the use of these energies. Systems with photovoltaic and wind generation storage of energy in batteries has been presented as a form of alternative energy generation. The variability in the intensity of solar and wind energy can be circumvented by complementation between a source and another, or the more stable set to the generation of the system. The objective of this dissertation is to create a methodology fuzzy and simulate their use in intelligent control of a hybrid system of generating power, using the energy solar-photovoltaic and wind.When you use a system of control based on fuzzy logic, is reached the point of maximum generation of energy, thus transferring all the energy generated on the basis of alternative sources, solar, photovoltaic and wind, to the load and / or batteries when its use not immediately. The model takes three variables of entry, which are: wind speed, solar radiation and loading the bank of batteries. The simulations will be used for some software such as MATLAB and others, that will be cited during the work. In these environments will be analyzed and simulated all mathematical modeling, rules and other variables in the system described fuzzy. This model could be used in the implementation of a system of control of hybrid systems to generate energy, providing the best use of energy sources, sun and wind, so we can extract the maximum energy possible these alternative sources without any prejudice to the environment. The work was developed in the Laboratory of Energização Rural, of the Department of Rural Engineering of UNESP - FCA...(Complete abstract, click electronic access below) / Mestre

目標：本文旨在建構一種新型的模糊回歸模式，解決一类較複雜的模糊回歸問題。 研究方法:推廣局部加權回歸的思想，先從理論上構建新模型；然後借由模拟數據，從多個方面考察新模型的性質，并和其他模型做比較。 發現：局部加權回歸方法結合模糊隸屬度概念，使模糊回歸理論有更多的應用場合。 原創性：目前在模糊回歸領域的主流思想是通過線性規劃等方法來構建模型，而本文另闢蹊徑，首次從局部加權的角度構建了模糊回歸的新模型。 關鍵字: 模糊理論 模糊回歸分析 局部加權 / Objective: This paper aims to construct a new fuzzy regression model to solve a more complex fuzzy regression problem. Method: Build a new model by promoting the idea of locally weighted regression; Using simulated data to compare the new model with other models. Conclusion: The fuzzy membership degree concept combined with the locally weighted regression method makes the fuzzy regression theory have more applications. Originality: At present, the main idea in the field of fuzzy regression is to construct models by means of linear programming. In this paper, a new model of fuzzy regression is constructed from the perspective of locally weighted method for the first time. Keyword: Fuzzy theory、 Fuzzy regression、Locally weighted method

Esta dissertação apresenta uma metodologia baseada em Algoritmos Genéticos e Teoria dos Conjuntos Fuzzy que determina o número ótimo de medidores de Qualidade de Energia, bem como os pontos onde os mesmos devem ser instalados para monitorar Afundamentos e Elevações de Tensão em redes de transmissão de energia elétrica. A metodologia desenvolvida representa uma ferramenta para diversas áreas de uma empresa de energia elétrica, porque ela indica a melhor maneira para se alocar medidores de Qualidade de Energia, considerando as restrições econômicas a que as empresas estão sujeitas. Além disto, a metodologia também determina a melhor configuração para um sistema de medição para um sistema de potência, quando há barras que devem ser monitoradas e/ou o número de medidores disponíveis é menor que o mínimo necessário para atingir a completa Observabilidade. O trabalho desenvolvido também representa uma importante aplicação dos Algoritmos Genéticos em conjunto com a Teoria dos Conjuntos Fuzzy. Ele define aspectos a serem considerados no problema de monitoração de Qualidade de Energia Elétrica e sugere rapidamente a melhor forma de atender diferentes objetivos e restrições operacionais. A fim de validar o modelo proposto, nesta dissertação, três redes elétricas são avaliadas para determinar o número mínimo de medidores de Qualidade de Energia, bem como a localização onde esses equipamentos deveriam ser instalados. Os níveis de Redundância e de Observabilidade são apresentados, bem como as formulações usadas para atingir os resultados apresentados. / This dissertation presents a methodology based on Genetic Algorithms and Fuzzy Mathematical Programming to determine the optimum number of power quality monitors and the locations they should be installed to measure Voltage Sags and Swells in transmission power networks. The methodology developed represents a tool for planning power networks, because it indicates the best way to allocate the power quality monitors, considering economical constraints. Similar to this, the methodology also determines the best configuration for a monitoring system when there are important busses to monitor and the number of monitors available is lower than the minimal needed to accomplish completely Observability. The work presented here also represents an important Genetic Algorithms application. With fuzzy mathematical programming it easily allows the definition of aspects to be considered in the monitoring problem and quickly suggests the best way to attend different objectives and operational restrictions. In order to validate the developed model proposed in this dissertation three electric power networks are assessed to determine the minimum number of power quality monitors as well as the locations where these devices should be installed. The levels of Redundancy and Observability are presented as well as the formulations used to achieve the results.

Machine tool components are widely used in many industrial applications. In accordance with their usage, a reliable health monitoring system is necessary to detect defects in these components in order monitor machinery performance and avoid malfunction. Even though several techniques have been reported for fault detection and diagnosis, it is a challenging task to implement a condition monitoring system in real world applications due to their complexity in structure and noisy operating environment. The primary objective of this thesis is to develop novel intelligent algorithms for a reliable fault diagnosis of machine tool components. Another objective is to use Micro Electro Mechanical System (MEMS) sensor and interface it with Raspberry pi hardware for the real time condition monitoring. Primarily knowledge based approach with morphological operators and Fuzzy Inference System is proposed, the e˙ectiveness of this approach lies in the selection of structuring elements(SEs). When this is evaluated with di˙erent classes of bearing fault signals, it is able to detect the fault frequencies e˙ectively. Secondarily, An analytical approach with multi class support machine is proposed, this method has uniqueness of learning on its own with out any prior knowledge, the e˙ectiveness of this method lies on selected features and used kernel for converging. Results have shown that RBF (Radial Bias Function) kernel, which is commonly known as gauss kernel has good performance in identifying faults with less computation time. An idea of prototyping these methods has triggered in using Micro Electro Mechanical System (MEMS) sensor for data acquisition and real time Condition Monitoring. LIS3DH accelerometer sensor is used for the data acquisition of spindle for capturing high frequency fault signals. The measured data is analyzed and compared with the industrial sensor k-shear accelerometer type 8792A.

在資訊爆炸的時代，處於日趨複雜的環境及多重資訊來源管道之下，如何從大量及瑣碎的資訊中找出「重要且有用」的部份，藉以輔助企業或個人制定正確的決策，並降低資訊取得的成本，是資訊人員在設計資訊系統時所必須考量的重要因素之一，因此，資訊篩選(Information filtering)已成為當務之急，更顯示出其重要性。 本研究之主要目的在於整合類神經網路與模糊理論以建立一個通用型資訊篩選之演算法，藉由此演算法可篩選出重要之決策變數，減少資訊的使用量，達到相同或類似的決策結果，進而降低後續資訊蒐集之成本。最後並以四個XOR實驗及國內上市公訂股價預測為例，以測試本研究所開發出來之演算法的正確性及實用性。就XOR實驗結果顯示均能迅速且正確的篩選出重要的輸入資訊；而在股價預測方面，結合基本面分析及技術面分析，根據個別公司的特性及不同的時間點，能夠篩選出其重要的預測變數，可作為股市投資者之重要參考依據。因此，藉由本演算法所開發出來的系統，可以達到資訊篩選的目的。 / At the time of information explosion, how to filter the important and useful parts from a large and trivial information pool is one of the most important factors considering in designing information systems which are used to assist users making right decisions by MIS managers. The purpose of this research is to integrate two technologies. Artificial Neural Network and Fuzzy Theory, to develop a generalized algorithm to filter important information. We hope that using this algorithm we can (1)filter the important decision variables, (2)decrease the information usage, and (3)reduce the cost of information collection. Finally, we made four experiments on the XOR system and stock market forecasting to test the accuracy and practicability of the information filter algorithm. The results of experiments showed that the algorithm could filter the important information correctly and quickly.

Die vorliegende Arbeit beschreibt ein generisches Verfahren zur wahrscheinlichkeitsbasierten Erkennung von Fahrsituationen und Fahrmanövern für sicherheitsrelevante Fahrerassistenzsysteme. Fahrsituationen und Manöver unterliegen einer gewissen Unsicherheit basierend auf der unterschiedlichen Situationswahrnehmung bzw. Manöverdurchführung der Fahrzeugführer. Diese Unsicherheitskomponente wird in den Ansatz zur Situations- und Manövererkennung mit einbezogen. Ein weiterer Unsicherheitsaspekt beruht auf den ungenauen Umgebungsinformationen auf denen die Situations- und Manövererkennung basiert. Beide Unsicherheitsursachen sind völlig unabhängig voneinander und werden aus diesem Grund separat betrachtet und modelliert. Zur Modellierung dieser beiden Unsicherheitsaspekte bedient sich der vorgestellte Ansatz der Fuzzy-Theorie, der Theorie der probabilistischen Netzen sowie Verfahren zur Fehlerfortpflanzung und Sensitivitätsanalyse. Nach der theoretischen Vorstellung dieser Methodiken wird in der Arbeit detailliert auf den Einsatz und das Zusammenspiel der einzelnen Verfahren zur Erkennung der Fahrsituationen und Fahrmanöver eingegangen. Die Umsetzbarkeit des vorgestellten Verfahrens wird am Beispiel der Notbremssituation gezeigt. Die Notbremssituation setzt sich aus unterschiedlichen Teilsituationen und Manövern zusammen. Die Erkennung der einzelnen Situationen und Manöver sowie die Zusammenführung zur übergeordneten Notbremssituation wurden mit Hilfe des vorgestellten Verfahrens realisiert. Zur Evaluierung der Erkennungsgüte wurden sowohl Messdaten aus dem Straßenverkehr als auch realitätsnahe Daten, aufgezeichnet auf Versuchsstrecken, herangezogen. / The present work describes a generic method for the probabilistic identification of driving situations and driving manoeuvres for safety relevant driver assistance systems. Driving situations and driving manoeuvres underlie a certain uncertainty based on the different situation perception and manoeuvre execution of the driver. This uncertainty component is considered in the approach for the situation and manoeuvre identification. An additional uncertainty aspect is based on the inaccurate environment information, the identification of driving situations and manoeuvres depend on. Both uncertainty aspects are completely independent and are considered and modelled separately for this reason. For modelling both of these uncertainty aspects the present approach is using the fuzzy theory, probabilistic networks, as well as methods for error propagation and sensitivity analysis. After introducing these techniques theoretically, the application and the interaction of the single methods to identify the driving situations and manoeuvres is described in detail. The practicability of the introduced proceeding is shown exemplarily on the emergency brake situation. The emergency brake situation consists of several situation and manoeuvre components. The identification of the single situations and manoeuvres as well as the combination to the higher emergency brake situation is realised with the introduced proceeding. Measuring data gathered on road traffic and close to reality data measured on a test track were used to evaluate the identification quality.

自Zadeh(1965)提出模糊集合論，將集合論之二值邏輯擴充至多值邏輯後，使得集合論更為完善，能更精確地描述實際問題之特性。由於教師評鑑的規準建構與測量方法皆深具模糊性，因此有必要針對評鑑問題之模糊特性，選擇適當之模糊理論。本研究以模糊德菲法以整合專家意見以建構評鑑指標，續以模糊層級分析法計算各指標間的相對權重。教師評鑑除考量量化指標外，由於質化指標常以語意詞句(linguistic terms)表達，故採取Chen和Hwang (1992)所提之轉換法將模糊評語集轉換為明確得點值(crisp score)。並利用模糊綜合評估以從事實際教師評鑑之結果分析，其中涉及不同運算算子間的差異比較。 本研究結果發現：教師評鑑指標體系包括教師評鑑指標共可區分為九大類主指標四十一項次指標。其主指標分別為「專業知識」、「教學準備能力」、「教學策略與實施能力」、「教學評量能力」、「運用教學資源能力」、「班級經營能力」、「專業責任」、「校務參與及服務績效」與「人際溝通能力」等九類；「專業知識類」次指標包括「任教科目的專門知識（含提供完整的知識架構）」、「教學方法的專業知識（含清楚教導概念）」、「課程與教材方面的知識（包括清楚目前的學習內容與先前的學習內容及未來的學習內容之間的關連）」、「教學情境的專業知識」、「輔導方面的知識（包括了解學生的心理）」與「學習與發展方面的知識（包括精熟學生背景知識和經驗）」等六項，「教學準備能力類」次指標包括「訂定教學計畫，妥善準備教具」、「根據學生學習需求及課程標準，訂定適合的教學目標」等兩項，「教學策略與實施能力類」次指標包括「教導認知、情意及動作技能的學習與遷移」、「教材展示精確又清楚」、「運用適當教學方法的技巧」、「對教學內容的解釋、舉例之能力」、「教學內容組織能力」、「引起並維持學生的學習動機與注意力」、「教學表達能力」、「傾聽兒童說話的技巧」、「促使學生延展其思考」等九項，「教學評量能力類」次指標包括「評量兒童表現並提供回饋與指導」、「根據評量結果改進教學歷程」、「適時而正確地評估學生進步情形」等三項，「運用教學資源能力類」次指標包括「妥善佈置教學環境」、「運用多樣教學資源」等兩項，「班級經營能力類」包括「輔導學生的能力（包括輔導學生因學業及日常生活所產生的困擾）」、「指導學生遵守生活常規，養成良好行為習慣」、「教室管理的技巧（包括建立愉快的班級氣氛、建立教室常規和程序）」、「有效運用獎懲手段」、「訂定適切的班級規範」與「創造適當的學習環境」等六項，「專業責任類」次指標包括「關懷與瞭解學生」、「工作勤惰」、「教學認真」、「教育信念」、「研究進修」與「敬業精神」等六項，「校務參與及服務績效類」次指標包括「主動積極協助校務推展」、「配合學校行政詳建各項學生檔案」、「對學校活動積極參與」等三項，「人際溝通能力類」包括「同儕教師溝通交流分享互動之能力」、「和家長保持溝通互動以維持良好的親師關係」、「建立和維持師生關係」、「教師能公平、公正地對待全部學生」等四項，合計四十一項次指標。 / Since Zadeh(1965) proposed the fuzzy sets theorem and further developed the theorem from　two-value logic to continues multi-value logic, the assembly theorem had become more completed and was capable of depicting the characteristics of real problem more specifically. Because the standard structure and measuring method of teacher appraisal are both ambiguous, it is necessary to adopt an suitable fuzzy theorem for this ambiguous characteristic of teacher appraisal. The study incorporated the opinions of specialists into structural evaluation indicator based upon Fuzzy Delphi and further calculated the relative weights among each indicator by Fuzzy Analytic Hierarchy Process. The teacher appraisal not only need to quantify the indicator but also to qualify them using linguistic terms to express. Therefore, the study adopted the Chen-Hwang (1992) transfer method to transform the ambiguous evaluation terms into specific crisp scores and used the composite of fuzzy evaluation to engage the analysis of real teacher appraisal involving differentiated comparison among individual operants. The outcomes of the study demonstrated that the indicator system of teacher appraisal including teacher appraisal indicator consist of 9 major indicator and 41 minor indicator. The 9 major indicator were professional knowledge, capability of teaching preparation, teaching strategy and implementation ability, capability of teaching assessment , capability of employing teaching resource, capability of class management, professional duty, involvement of school managing and service achievement, and capability of interpersonal communication. The 6 sub-indicator in the professional knowledge category includes professional knowledge of the specific discipline with a comprehensive knowledge structure, professional knowledge of teaching methods included a lucid teaching concept, professional knowledge of class and teaching material included understanding of the connection between the current and future learning contents. professional knowledge of teaching situations, professional knowledge of consulting included understanding of student psychology and professional knowledge of learning and developing included awareness of student background knowledge and experiences. There were two sub-indicator in the capability of teaching preparation (making a teaching agenda and preparing teaching tools, and planing a proper teaching goal based upon student learning need and class standard. The 9 sub-indicator in the teaching strategy and implementation ability were, understanding of guidance and learning and transferring of mental and physical skills, demonstration of teaching material precisely and clearly, employment of proper teaching skills, the interpretation of teaching contents and exemplification faculty, ability to organize teaching contents, inspire and maintain learning motivation and attention of student, faculty of teaching expression, listening skills concerning the child talk, and stimulate the extension of student thinking. The 3 sub-indicator in capability of teaching evaluation were, evaluation of child perfor-mance in order to render guidance and feedback, rescheduling the teaching agenda according to the evaluation outcomes, and accurately and timely evaluating student progression . The capability of employing teaching resource category consisted of implementing teaching environment congruently and adopting diversified teaching resources. The 6 sub-indicator in capability of class management were faculty of student guidance (consulting the student problems causing by class-learning and everyday activities), lead students to fellow the normal rules and to keep righteous behavior habits, class room managing skills ( establish an amicable classroom atermosphere and the classroom guidelines and procedures), proper employing award/ punish system, making the proper classroom paradigm, and create suitable learning environment. The 6 subindicator in professional duty were concerning and understanding students, working habits, teaching sincerity, teaching believe, advancing research, reverence to occupation, The 3 subin-dicator in involvement of school administration and service achievement were, actively assisting the school administration to implement, cooperating the school administration to establish the student records, actively engaging the school activities. The capability of interpersonal communication category composed of 4 subindicator, capability of communicating and interacting with teaching colleagues, capability of communicating and interacting with student parents and maintain a good teacher-parent relationship, establishing and keeping the teacher-student relationship, treating all student with equality and fairness. The grand total of sub-indicator were 41.

Die vorliegende Arbeit beschreibt ein generisches Verfahren zur wahrscheinlichkeitsbasierten Erkennung von Fahrsituationen und Fahrmanövern für sicherheitsrelevante Fahrerassistenzsysteme. Fahrsituationen und Manöver unterliegen einer gewissen Unsicherheit basierend auf der unterschiedlichen Situationswahrnehmung bzw. Manöverdurchführung der Fahrzeugführer. Diese Unsicherheitskomponente wird in den Ansatz zur Situations- und Manövererkennung mit einbezogen. Ein weiterer Unsicherheitsaspekt beruht auf den ungenauen Umgebungsinformationen auf denen die Situations- und Manövererkennung basiert. Beide Unsicherheitsursachen sind völlig unabhängig voneinander und werden aus diesem Grund separat betrachtet und modelliert. Zur Modellierung dieser beiden Unsicherheitsaspekte bedient sich der vorgestellte Ansatz der Fuzzy-Theorie, der Theorie der probabilistischen Netzen sowie Verfahren zur Fehlerfortpflanzung und Sensitivitätsanalyse. Nach der theoretischen Vorstellung dieser Methodiken wird in der Arbeit detailliert auf den Einsatz und das Zusammenspiel der einzelnen Verfahren zur Erkennung der Fahrsituationen und Fahrmanöver eingegangen. Die Umsetzbarkeit des vorgestellten Verfahrens wird am Beispiel der Notbremssituation gezeigt. Die Notbremssituation setzt sich aus unterschiedlichen Teilsituationen und Manövern zusammen. Die Erkennung der einzelnen Situationen und Manöver sowie die Zusammenführung zur übergeordneten Notbremssituation wurden mit Hilfe des vorgestellten Verfahrens realisiert. Zur Evaluierung der Erkennungsgüte wurden sowohl Messdaten aus dem Straßenverkehr als auch realitätsnahe Daten, aufgezeichnet auf Versuchsstrecken, herangezogen. / The present work describes a generic method for the probabilistic identification of driving situations and driving manoeuvres for safety relevant driver assistance systems. Driving situations and driving manoeuvres underlie a certain uncertainty based on the different situation perception and manoeuvre execution of the driver. This uncertainty component is considered in the approach for the situation and manoeuvre identification. An additional uncertainty aspect is based on the inaccurate environment information, the identification of driving situations and manoeuvres depend on. Both uncertainty aspects are completely independent and are considered and modelled separately for this reason. For modelling both of these uncertainty aspects the present approach is using the fuzzy theory, probabilistic networks, as well as methods for error propagation and sensitivity analysis. After introducing these techniques theoretically, the application and the interaction of the single methods to identify the driving situations and manoeuvres is described in detail. The practicability of the introduced proceeding is shown exemplarily on the emergency brake situation. The emergency brake situation consists of several situation and manoeuvre components. The identification of the single situations and manoeuvres as well as the combination to the higher emergency brake situation is realised with the introduced proceeding. Measuring data gathered on road traffic and close to reality data measured on a test track were used to evaluate the identification quality.