Orientador: Renata Zotin Gomes de Oliveira / Resumo: Neste trabalho apresentamos um estudo introdutório à Teoria de Conjuntos Fuzzy e Lógica Fuzzy, mostrando o seu potencial de aplicação através do aperfeiçoamento do cálculo da estimativa do tempo de vida de um transformador de energia elétrica, informação de grande importância para a equipe de qualidade de uma empresa conseguir prever o melhor momento de manutenção. Foi desenvolvido um sistema computacional que consulta uma base de dados já existente e aplica o cálculo do tempo de vida utilizando três métodos diferentes, permitindo ao usuário comparar e selecionar o resultado mais próximo ao real, com base em fatores externos. Como energia elétrica é um tema amplo que aparece em diversos componentes curriculares, facilitando a construção de projetos interdisciplinares, apresentamos uma proposta de projeto onde será criada uma maquete de uma turbina hidrelétrica utilizando conceitos da construção geométrica e a modelagem matemática da tensão gerada por esta turbina, em função da altura. A proposta de projeto foi aplicada no 8◦ ano do Ensino Fundamental e os resultados estão presentes no trabalho. / Abstract: In this work we present an introductory study of Fuzzy Sets Theory and Fuzzy Logic, showing its potential for application through the improvement of the calculation of the life span estimation of an electrical power transformer, information of great importance for the company's quality team to predict the best moment of maintenance. It was developed a computational system that consults the existing database and applies the calculation of the life span using three di erent methods, allowing the user to compare and select the result closest to the actual based on external factors. Since electric energy is a broad theme that appears in several curricular components, facilitating the construction of interdisciplinary projects, we present a project proposal where a model of a hydroelectric turbine will be created using concepts of geometric construction and the mathematical modelling of the tension generated by this turbine, as a function of height. The project proposal was applied in the 8◦ year of elementary school and the results are present at work. / Mestre

Some disciplines are inherently complex and challenging to learn. This research attempts to design an instructional strategy for CBL systems to simplify learning certain complex domains. Firstly, problem transformation, a constructionist instructional technique, is used to promote active learning by encouraging students to construct more complex artefacts based on less complex ones. Scaffolding is used at the initial learning stages to alleviate the difficulty associated with complex transformation processes. The proposed instructional strategy brings various techniques together to enhance the learning experience. A functional prototype is implemented with Object-Z as the exemplar subject. Both objective and subjective evaluations using the prototype indicate that the proposed CBL system has a statistically significant impact on learning a complex domain. CBL systems include Learner models to provide adaptable support tailored to individual learners. Bayesian theory is used in general to manage uncertainty in Learner models. In this research, a fuzzy logic based locally intelligent Learner model is utilized. The fuzzy model is simple to design and implement, and easy to understand and explain, as well as efficient. Bayesian theory is used to complement the fuzzy model. Evaluation shows that the accuracy of the proposed Learner model is statistically significant. Further, opening Learner model reduces uncertainty, and the fuzzy rules are simple and resemble human reasoning processes. Therefore, it is argued that opening a fuzzy Learner model is both easy and effective. Scaffolding requires formative assessments. In this research, a confidence based multiple test marking scheme is proposed as traditional schemes are not suitable for measuring partial knowledge. Subjective evaluation confirms that the proposed schema is effective. Finally, a step-by-step methodology to transform simple UML class diagrams to Object-Z schemas is designed in order to implement problem transformation. This methodology could be extended to implement a semi-automated translation system for UML to Object Models.

A wireless Ad-hoc network is a group of wireless devices that communicate with each other without utilising any central management infrastructure. The operation of Ad-hoc networks depends on the cooperation among nodes to provide connectivity and communication routes. However, such an ideal situation may not always be achievable in practice. Some nodes may behave maliciously, resulting in degradation of the performance of the network or even disruption of its operation altogether. The ease of establishment, along with the mobility capabilities that these networks offer, provides many advantages. On the other hand, these very characteristics, as well as the lack of any centralised administration, are the root of several nontrivial challenges in securing such networks. One of the key objectives of this thesis is to achieve improvements in the performance of Ad-hoc networks in the presence of malicious nodes. In general, malicious nodes are considered as nodes that subvert the capability of the network to perform its expected functions. Current Ad-hoc routing protocols, such as the Ad-hoc On demand Distance Vector (AODV), have been developed without taking the effects of misbehaving nodes into consideration. In this thesis, to mitigate the effects of such nodes and to attain high levels of security and reliability, an approach that is based on the utilisation of the behaviour history of all member nodes is proposed. The aim of the proposed approach is to identify routes between the source and the destination, which enclose no, or if that is not possible, a minimal number, of malicious nodes. This is in contrast to traditional approaches that predominantly tend to use other criteria such as shortest path alone. Simulation and experimental results collected after applying the proposed approach, show significant improvements in the performance of Ad-hoc networks even in the presence of malicious nodes. However, to achieve further enhancements, this approach is expanded to incorporate trust levels between the nodes comprising the Ad-hoc network. Trust is an important concept in any relation among entities that comprise a group or network. Yet it is hard to quantify trust or define it precisely. Due to the dynamic nature of Ad-hoc networks, quantifying trust levels is an even more challenging task. This may be attributed to the fact that different numbers of factors can affect trust levels between the nodes of Ad-hoc networks. It is well established that fuzzy logic and soft computing offer excellent solutions for handling imprecision and uncertainties. This thesis expands on relevant fuzzy logic concepts to propose an approach to establish quantifiable trust levels between the nodes of Ad-hoc networks. To achieve quantification of the trust levels for nodes, information about the behaviour history of the nodes is collected. This information is then processed to assess and assign fuzzy trust levels to the nodes that make up the Ad-hoc network. These trust levels are then used in the routing decision making process. The performance of an Ad-hoc network that implements the behaviour history based approach using OPtimised NETwork (OPNET) simulator is evaluated for various topologies. The overall collected results show that the throughput, the packet loss rate, and the round trip delay are significantly improved when the behaviour history based approach is applied. Results also show further enhancements in the performance of the Ad-hoc network when the proposed fuzzy trust evaluation approach is incorporated with a slight increase in the routing traffic overhead. Given the improvements achieved when the fuzzy trust approach is utilised, for further enhancements of security and reliability of Ad-hoc networks, future work to combine this approach with other artificial intelligent approaches may prove fruitful. The learning capability of Artificial Neural Networks makes them a prime target for combination with fuzzy based systems in order to improve the proposed trust level evaluation approach. / Doctor of Philosophy (PhD)

This study investigates the relationships between firm's technology, absorptive capacity and the internationalization process in the high-tech SMEs. The research identifies the most influential factors that affect the international activities and expansion decisions of New Zealand high-tech SMEs with core capabilities in agro-technology. Mixed methods, qualitative and quantitative elements in the data collection and analysis, were employed in this research for a reason that a deeper understanding of the research subject and the analysis of complex issues such as the internationalization process and absorptive capacity required methodological variety. The use of qualitative and quantitative methods took place in parallel. Both methods were used to study the same subject but they had specific objective related purposes and they offered the possibility of developing rich empirical data as well as a more comprehensive understanding of the subject under the study. The findings show that it is absorptive capacity that explains internationalization process, not internationalization process that explains absorptive capacity. The practice of internationalizing is as much a reflection of a firm's absorptive capacity as it is its determinant. The research identifies that high-tech SMEs possess technological and non-core absorptive capacity which in a different way influence firms' strategies. The research suggests that firm's technological capabilities and the advantage of specialized knowledge along with their limited non-core absorptive capacity act as constraints to the development of the future international strategy in high-tech SMEs. The study expands the existing literature on internationalization by developing variables for evaluating absorptive capacity in firms. This helped develop an absorptive capacity model which can be used as a valuable tool for self-assessment by firms to facilitate gaining insight towards further growth and development. The research suggested that if firms were able to measure its absorptive capacity this may result in improved business activities and enhanced presence in the world market. The results of this study should encourage firms to identify, capture and articulate knowledge achieved by their ventures. Managers must develop and nurture skills that ensure effective integration of learning as their firms expand, particularly internationally. These findings and absorptive capacity model offered as a tool should encourage managers to explore when, where, and how to best use firm's resources in the business operations. This is particularly important in regards to the research context (high-tech SMEs) where scientists are managers as well.

The research presented in this thesis aims to demonstrate that fuzzy logic is a useful tool for developing mechanisms for controlling traffc flow in ATM based multimedia networks to maintain quality of service (QoS) requirements and maximize resource utilization. The study first proposes a hierarchical, multilevel control structure for ATM networks to exploit the reported strengths of fuzzy logic at various control levels. Then, an extensive development and evaluation is presented for a subset of the proposed control architecture at the congestion control level. An ATM based multimedia network must have quite sophisticated traffc control capabilities to effectively handle the requirements of a dynamically varying mixture of voice, video and data services while meeting the required levels of performance. Feedback control techniques have an essential role for the effective and efficient management of the resources of ATM networks. However, development of conventional feedback control techniques relies on the availability of analytical system models. The characteristics of ATM networks and the complexity of service requirements cause the analytical modeling to be very difficult, if not impossible. The lack of realistic dynamic explicit models leads to substantial problems in developing control solutions for B-ISDN networks. This limits the ability of conventional techniques to directly address the control objectives for ATM networks. In the literature, several connection admission and congestion control methods for B-ISDN networks have been reported, and these have achieved mixed success. Usually they either assume heavily simplified models, or they are too complicated to implement, mainly derived using probabilistic (steady-state) models. Fuzzy logic controllers, on the other hand, have been applied successfully to the task of controlling systems for which analytical models are not easily obtainable. Fuzzy logic control is a knowledge-based control strategy that can be utilized when an explicit model of a system is not available or, the model itself, if available, is highly complex and nonlinear. In this case, the problem of control system design is based on qualitative and/or empirically acquired knowledge regarding the operation of the system. Representation of qualitative or empirically acquired knowledge in a fuzzy logic controller is achieved by linguistic expressions in the form of fuzzy relational equations. By using fuzzy relational equations, classifications related to system parameters can be derived without explicit description. The thesis presents a new predictive congestion control scheme, Fuzzy Explicit Rate Marking (FERM), which aims to avoid congestion, and by doing so minimize the cell losses, attain high server utilization, and maintain the fair use of links. The performance of the FERM scheme is extremely competitive with that of control schemes developed using traditional methods over a considerable period of time. The results of the study demonstrate that fuzzy logic control is a highly effective design tool for this type of problems, relative to the traditional methods. When controlled systems are highly nonlinear and complex, it keeps the human insight alive and accessible at the lower levels of the control hierarchy, and so higher levels can be built on this understanding. Additionally, the FERM scheme has been extended to adaptively tune (A-FERM) so that continuous automatic tuning of the parameters can be achieved, and thus be more adaptive to system changes leading to better utilization of network bandwidth. This achieves a level of robustness that is not exhibited by other congestion control schemes reported in the literature. In this work, the focus is on ATM networks rather than IP based networks. For historical reasons, and due to fundamental philosophical differences in the (earlier) approach to congestion control, the research for control of TCP/IP and ATM based networks proceeded separately. However, some convergence between them has recently become evident. In the TCP/IP literature proposals have appeared on active queue management in routers, and Explicit Congestion Notication (ECN) for IP. It is reasonably expected that, the algorithms developed in this study will be applicable to IP based multimedia networks as well.

Improving network performance in terms of efficiency, fairness in the bandwidth, and system stability has been a research issue for decades. Current Internet traffic control maintains sophistication in end TCPs but simplicity in routers. In each router, incoming packets queue up in a buffer for transmission until the buffer is full, and then the packets are dropped. This router queue management strategy is referred to as Drop Tail. End TCPs eventually detect packet losses and slow down their sending rates to ease congestion in the network. This way, the aggregate sending rate converges to the network capacity. In the past, Drop Tail has been adopted in most routers in the Internet due to its simplicity of implementation and practicability with light traffic loads. However Drop Tail, with heavy-loaded traffic, causes not only high loss rate and low network throughput, but also long packet delay and lengthy congestion conditions. To address these problems, active queue management (AQM) has been proposed with the idea of proactively and selectively dropping packets before an output buffer is full. The essence of AQM is to drop packets in such a way that the congestion avoidance strategy of TCP works most effectively. Significant efforts in developing AQM have been made since random early detection (RED), the first prominent AQM other than Drop Tail, was introduced in 1993. Although various AQMs also tend to improve fairness in bandwidth among flows, the vulnerability of short-lived flows persists due to the conservative nature of TCP. It has been revealed that short-lived flows take up traffic with a relatively small percentage of bytes but in a large number of flows. From the user’s point of view, there is an expectation of timely delivery of short-lived flows. Our approach is to apply artificial intelligence technologies, particularly fuzzy logic (FL), to address these two issues: an effective AQM scheme, and preferential treatment for short-lived flows. Inspired by the success of FL in the robust control of nonlinear complex systems, our hypothesis is that the Internet is one of the most complex systems and FL can be applied to it. First of all, state of the art AQM schemes outperform Drop Tail, but their performance is not consistent under different network scenarios. Research reveals that this inconsistency is due to the selection of congestion indicators. Most existing AQM schemes are reliant on queue length, input rate, and extreme events occurring in the routers, such as a full queue and an empty queue. This drawback might be overcome by introducing an indicator which takes account of not only input traffic but also queue occupancy for early congestion notification. The congestion indicator chosen in this research is traffic load factor. Traffic load factor is in fact dimensionless and thus independent of link capacity, and also it is easy to use in more complex networks where different traffic classes coexist. The traffic load indicator is a descriptive measure of the complex communication network, and is well suited for use in FL control theory. Based on the traffic load indicator, AQM using FL – or FLAQM – is explored and two FLAQM algorithms are proposed. Secondly, a mice and elephants (ME) strategy is proposed for addressing the problem of the vulnerability of short-lived flows. The idea behind ME is to treat short-lived flows preferably over bulk flows. ME’s operational location is chosen at user premise gateways, where surplus processing resources are available compared to other places. By giving absolute priority to short-lived flows, both short and long-lived flows can benefit. One problem with ME is starvation of elephants or long-lived flows. This issue is addressed by dynamically adjusting the threshold distinguishing between mice and elephants with the guarantee that minimum capacity is maintained for elephants. The method used to dynamically adjust the threshold is to apply FL. FLAQM is deployed to control the elephant queue with consideration of capacity usage of mice packets. In addition, flow states in a ME router are periodically updated to maintain the data storage. The application of the traffic load factor for early congestion notification and the ME strategy have been evaluated via extensive experimental simulations with a range of traffic load conditions. The results show that the proposed two FLAQM algorithms outperform some well-known AQM schemes in all the investigated network circumstances in terms of both user-centric measures and network-centric measures. The ME strategy, with the use of FLAQM to control long-lived flow queues, improves not only the performance of short-lived flows but also the overall performance of the network without disadvantaging long-lived flows.

<p>In the railway industry, slip control has always been essential due to the low friction between the wheels and the rail. In this master’s thesis we have gathered several slip control methods and evaluated them. These evaluations were performed in Matlab-Simulink on a slip process model of a railway vehicle. The objective with these evaluations were to show advantages and disadvantages with the different slip control methods. </p><p>The results clearly show the advantage of using a slip optimizing control method, i.e. a method that finds the optimal slip and thereby maximizes the use of adhesion. We have developed two control strategies that we have found superior in this matter. These methods have a lot in common. For instance they both use an adhesion observer and non-linear gain, which enables fast optimization. The difference lies in how this non-linear gain is formed. One strategy uses an adaptive algorithm to estimate it and the other uses fuzzy logic. </p><p>A problem to overcome in order to have well functioning slip controllers is the formation of vehicle velocity. This is a consequence of the fact that most slip controllers use the velocity as a control signal.</p>

The simplest form of induction motors, known as AC squirrel cage motor, is the universal workhorse of industrial and commercial premises. For many years it was restricted to constant speed applications while DC motors were preferred for high-performance variable speed and servo drives. With modern advances in semiconductor and digital signal processing technologies, it is now possible to operate induction motors in high-performance drives at a reasonable cost with Field Oriented Control methods. The latter have made induction motor drives equivalent to DC drives in terms of independent control of flux and torque; and superior to them in terms of dynamic performance. In developing Field Oriented Control for induction motors engineers are faced with two major challenges: (1) the estimation of rotor data to compute for the slip gain, and (2) the compensation of changes in drive operating conditions and parameters in order to maintain the drive performance high at all time. This thesis addresses these issues by introducing two independent control systems. The first system is designed to estimate online the value of the slip gain in the entire torque-speed plane in order to maintain decoupled control of torque and flux despite the so-called detuning effects. It is based on evaluating the operating condition of the drive in terms frequency and load torque, and selecting the appropriate estimation method accordingly. A fuzzy controller is used to generate the distribution factor for the methods. The second system is a fuzzy self-tuning speed controller, with reduced sensitivity to motor parameters and operating condition changes. It has the ability to adjust its gains in real time according to the current trend of the drive system. It is designed to maintain tight control of speed and torque for high-performance applications. The performances of the two controllers are validated through a series of simulation and experimental tests using a 2HP 3-phase induction motor with an ADMC21992 160-MHz DSP microprocessor. / Power Engineering and Power Electronics

Thesis (M.Eng.)--Memorial University of Newfoundland, 2005. / Bibliography: leaves 107-116.

This dissertation introduces advanced artificial intelligence based algorithm for detecting and classifying faults on the power system transmission line. The proposed algorithm is aimed at substituting classical relays susceptible to possible performance deterioration during variable power system operating and fault conditions. The new concept relies on a principle of pattern recognition and detects the existence of the fault, identifies fault type, and estimates the transmission line faulted section. The approach utilizes self-organized, Adaptive Resonance Theory (ART) neural network, combined with fuzzy decision rule for interpretation of neural network outputs. Neural network learns the mapping between inputs and desired outputs through processing a set of example cases. Training of the neural network is based on the combined use of unsupervised and supervised learning methods. During training, a set of input events is transformed into a set of prototypes of typical input events. During application, real events are classified based on the interpretation of their matching to the prototypes through fuzzy decision rule. This study introduces several enhancements to the original version of the ART algorithm: suitable preprocessing of neural network inputs, improvement in the concept of supervised learning, fuzzyfication of neural network outputs, and utilization of on-line learning. A selected model of an actual power network is used to simulate extensive sets of scenarios covering a variety of power system operating conditions as well as fault and disturbance events. Simulation results show improved recognition capabilities compared to a previous version of ART neural network algorithm, Multilayer Perceptron (MLP) neural network algorithm, and impedance based distance relay. Simulation results also show exceptional robustness of the novel ART algorithm for all operating conditions and events studied, as well as superior classification capabilities compared to the other solutions. Consequently, it is demonstrated that the proposed ART solution may be used for accurate, high-speed distinction among faulted and unfaulted events, and estimation of fault type and fault section.