<p style="text-align: justify;">U radu je razvijen jedan matematički model za simuliranje ponašanja nekih važnih radnih parametara motora SUS, kada u njegovom usisnom sistemu postoje neispravnosti tipa: &bdquo;nepredviđeni ulaz vazduha u usisni kolektor&ldquo;, &bdquo;pogrešno očitavanje senzora masenog protoka vazduha&ldquo;, &bdquo;pogrešno očitavanje senzora pritiska u usisnom kolektoru&ldquo;, &bdquo;pogrešno očitavanje senzora temperature u usisnom kolektoru&ldquo; i &bdquo;umanjen EGR protok&ldquo;. Na osnovu rezultata ovog modela predložen je novi dijagnostički koncept, u okviru kojeg je razvijen jedan model za prepoznavanje pomenutih neispravnosti. Predloženi koncept je proveren na realnim podacima, prikupljenim ispitivanjem jednog stvarnog motora u laboratorijskim uslovima, pri čemu su dobijeni zadovoljavajući rezultati.</p> / <p>A mathematical model capable of simulating some important IC engine operating parameters behavior when a fault in its intake air path exists. The faults considered are of the following types: &bdquo;air leakage in the intake path&ldquo;, &bdquo;faulty mass air flow sensor&ldquo;, &bdquo;faulty manifold absolute pressure sensor&ldquo;, &bdquo;faulty intake air temperature sensor&ldquo; and &bdquo;clogged EGR pipe&ldquo;. Relying on the data obtained by the fault simulator, a novel diagnosis concept is proposed. A model for fault detection and diagnosis was developed in the scope of the concept. The proposed concept was tested on the real data collected from an automobile IC engine in the laboratory conditions and satisfying results were obtained.</p>

Induction motors (IM) find widespread use in modern industry and for this reason they have been subject to a significant amount of research interest in recent times. One particular aspect of this research is the fault detection and diagnosis (FDD) of induction motors for use in a condition based maintenance (CBM) strategy; by effectively tracking the condition of the motor, maintenance action need only be carried out when necessary. This type of maintenance strategy minimises maintenance costs and unplanned downtime. The benefits of an effective FDD for IM is clear and there have been numerous studies in this area but few which consider the problem in a practical sense with the aim of developing a single system that can be used to monitor motor condition under a range of different conditions, with different motor specifications and loads. This thesis aims to address some of these problems by developing a general FDD system for induction motor. The solution of this problem involved the development and testing of a new approach; the adaptive mixed-residual approach (AMRA). The main aim of the AMRA system is to avoid the vast majority of unplanned failures of the machine and therefore as opposed to tackling a single induction motor fault, the system is developed to detect all four of the most statistically prevalent induction motor fault types; rotor fault, stator fault, air-gap fault and bearing fault. The mixed-residual fault detection algorithm is used to detect these fault types which includes a combination of spectral and model-based techniques coupled with particle swarm optimisation (PSO) for automatic identification of motor parameters. The AMRA residuals are analysed by a fuzzy-logic classifier and the system requires only current and voltage inputs to operate. Validation results indicate that the system performs well under a range of load torques and different coupling methods proving it to have significant potential for use in industrial applications.

<div> <p>Sound and vibration analysis is a prominent tool used for scientific investigations in various fields such as structural model identification or dynamic behavior studies. In manufacturing fields, the vibration signals collected through commercial sensors are utilized to monitor machine health, for sustainable and cost-effective manufacturing.</p> <p> Recently, the development of commercial sensors and computing environments have encouraged researchers to combine gathered data and Machine Learning (ML) techniques, which have been proven to be efficient for categorical classification problems. These discriminative algorithms have been successfully implemented in monitoring problems in factories, by simulating faulty situations. However, it is difficult to identify all the sources of anomalies in a real environment. </p> <p>In this paper, a Neural Network (NN) application on a KUKA KR6 robot arm is introduced, as a solution for the limitations described above. Specifically, the autoencoder architecture was implemented for anomaly detection, which does not require the predefinition of faulty signals in the training process. In addition, stethoscopes were utilized as alternative sensing tools as they are easy to handle, and they provide a cost-effective monitoring solution. To simulate the normal and abnormal conditions, different load levels were assigned at the end of the robot arm according to the load capacity. Sound signals were recorded from joints of the robot arm, then meaningful features were extracted from spectrograms of the sound signals. The features were utilized to train and test autoencoders. During the autoencoder process, reconstruction errors (REs) between the autoencoder’s input and output were computed. Since autoencoders were trained only with features corresponding to normal conditions, RE values corresponding to abnormal features tend to be higher than those of normal features. In each autoencoder, distributions of the RE values were compared to set a threshold, which distinguishes abnormal states from the normal states. As a result, it is suggested that the threshold of RE values can be utilized to determine the condition of the robot arm.</p> </div> <br>

In 2016 the average industry downtime cost was estimated to $260.000 every hour, and with Swedish industries being an important part of the national economy it would be desirable to reduce the amount of unplanned downtime to a minimum. There are currently many different solutions for system supervision for monitoring system health but none which analyse data with machine learning in an industrial gateway.   The aim for this thesis is to test, compare and evaluate three different algorithms to find a classifier suitable for a gateway environment. The evaluated algorithms were Random Forest, K-Nearest Neighbour and Linear Discriminant Analysis. Load imbalance detection was used as a case study for evaluating these algorithms. The gateway received data from a Modbus ATV32 frequency converter, which measured specific features from an induction motor. The imbalance was created with loads that were attached on a fly-wheel at different angles to simulate different imbalances. The classifiers were compared on their accuracy, memory usage, CPU usage and execution time. The result was evaluated with tables, confusion matrices and AUC- ROC curves.  Although all algorithms performed well LDA was best based on the criteria set.

Identificar e localizar faltas em alimentadores de distribuição representa um passo importante para a melhoria da qualidade de energia, pois proporciona impactos diretos sobre o tempo de inspeção. Na verdade, a duração da inspeção implica consideravelmente no intervalo em que os consumidores estão sem energia elétrica, quando ocorre uma interrupção não programada. O objetivo deste trabalho é fornecer um sistema de detecção automática de curtos-circuitos, permitindo aos profissionais das companhias de distribuição acompanhar e monitorar de maneira on-line a ocorrência de possíveis faltas e transitórios eletromagnéticos observados na rede primária de distribuição. A abordagem de detecção utiliza um sistema híbrido que combina ferramentas inteligentes e convencionais para identificar e localizar faltas em redes primárias. Os resultados que foram compilados demonstram grande potencialidade de aplicação da proposta em sistemas de distribuição. / Efficient faults identification and location in power distribution lines constitute an important step for power quality improvement, since they provide direct impacts on the inspection time. In fact, the duration of inspection implies directly in the time interval where consumers are without power, considering here the occurrence of a non-programmed interruption. The objective of this work is to provide an automated fault detection system, allowing to the power companies engineers to online track and monitor the possible occurrence of faults and electromagnetic transients observed in the primary network for the distribution circuits. The detection approach uses a hybrid system, which combines a set of intelligent and conventional tools to identify and locate faults in the primary networks. Validation results show great application potential in distribution systems.

A demanda crescente na complexidade, eficiência e confiabilidade nos sistemas industriais modernos têm estimulado os estudos da teoria de controle aplicada no desenvolvimento de sistemas de Monitoração e Detecção de Falhas. Neste trabalho foi desenvolvida uma metodologia inédita de Monitoração e Detecção de Falhas através do algoritmo GMDH e Redes Neurais Artificiais (RNA) que foi aplicada ao reator de pesquisas do IPEN, IEA-R1. O desenvolvimento deste trabalho foi dividido em duas etapas: sendo a primeira etapa dedicada ao pré-processamento das informações, realizada através do algoritmo GMDH; e a segunda o processamento das informações através de RNA. O algoritmo GMDH foi utilizado de duas maneiras diferentes: primeiramente, o algoritmo GMDH foi utilizado para gerar uma melhor estimativa da base de dados, tendo como resultado uma matriz denominada matriz_z, que foi utilizada no treinamento das RNA. Logo após, o GMDH foi utilizado no estudo das variáveis mais relevantes, sendo estas variáveis utilizadas no processamento das informações. Para realizar as simulações computacionais, foram propostos cinco modelos: Modelo 1 (Modelo Teórico) e Modelos 2, 3, 4 e 5 (Dados de operação do reator). Após a realização de um estudo exaustivo dedicado a Monitoração, iniciou-se a etapa de Detecção de Falhas em sensores, onde foram simuladas falhas na base de dados dos sensores. Para tanto as leituras dos sensores tiveram um acréscimo dos seguintes valores: 5%, 10%, 15% e 20%. Os resultados obtidos utilizando o algoritmo GMDH na escolha das melhores variáveis de entrada para as RNA foram melhores do que aqueles obtidos utilizando apenas RNA, o que viabiliza o uso da nova metodologia de Monitoração e Detecção de Falhas em sensores apresentada. / The increasing demand in the complexity, efficiency and reliability in modern industrial systems stimulated studies on control theory applied to the development of Monitoring and Fault Detection system. In this work a new Monitoring and Fault Detection methodology was developed using GMDH (Group Method of Data Handling) algorithm and Artificial Neural Networks (ANNs) which was applied to the IEA-R1 research reactor at IPEN. The Monitoring and Fault Detection system was developed in two parts: the first was dedicated to preprocess information, using GMDH algorithm; and the second part to the process information using ANNs. The GMDH algorithm was used in two different ways: firstly, the GMDH algorithm was used to generate a better database estimated, called matrix_z, which was used to train the ANNs. After that, the GMDH was used to study the best set of variables to be used to train the ANNs, resulting in a best monitoring variable estimative. The methodology was developed and tested using five different models: one Theoretical Model and four Models using different sets of reactor variables. After an exhausting study dedicated to the sensors Monitoring, the Fault Detection in sensors was developed by simulating faults in the sensors database using values of 5%, 10%, 15% and 20% in these sensors database. The results obtained using GMDH algorithm in the choice of the best input variables to the ANNs were better than that using only ANNs, thus making possible the use of these methods in the implementation of a new Monitoring and Fault Detection methodology applied in sensors.

Faults and failures in system components are the two main reasons for the instability and the degradation in control performance. In recent decades, fault-tolerant control (FTC) approaches were introduced to improve the resiliency of the control system against faults and failures. In general, FTC techniques are classified into two major groups: passive and active. Passive FTC systems do not rely on the fault information to control the system and are closely related to the robust control techniques while an active FTC system performs based on the information received from the fault detection and isolation (FDI) system, and the fault problem will be tackled more intelligently without affecting other parts of the system. This dissertation technically reviews fault and failure causes in control systems and finds solutions to compensate for their effects. Recent achievements in FDI approaches, and active and passive FTC designs are investigated. Thorough comparisons of several different aspects are conducted to understand the advantages and disadvantages of different FTC techniques to motivate researchers to further developing FTC, and FDI approaches. Then, a novel active FTC system framework based on online FDI is presented which has significant advantages in comparison with other state of the art FTC strategies. To design the proposed active FTC, a new FDI approach is introduced which uses the artificial neural network (ANN) and a model based observer to detect and isolate faults and failures in sensors and actuators. In addition, the extended Kalman filter (EKF) is introduced to tune ANN weights and improve the ANN performance. Then, the FDI signal combined with a nonlinear dynamic inversion (NDI) technique is used to compensate for the faults in the actuators and sensors of a nonlinear system. The proposed scheme detects and accommodates faults in the actuators and sensors of the system in real-time without the need of controller reconfiguration. The proposed active FTC approach is used to design a control system for three different applications: Unmanned aerial vehicle (UAV), load frequency control system, and proton exchange membrane fuel cell (PEMFC) system. The performance of the designed controllers are investigated through numerical simulations by comparison with conventional control approaches, and their advantages are demonstrated.

<p>The problem of fault diagnosis consists of detecting and isolating faults present in a system. As technical systems become more and more complex and the demands for safety, reliability and environmental friendliness are rising, fault diagnosis is becoming increasingly important. One example is automotive systems, where fault diagnosis is a necessity for low emissions, high safety, high vehicle uptime, and efficient repair and maintenance.</p><p>One approach to fault diagnosis, providing potentially good performance and in which the need for additional hardware is minimal, is model-based fault diagnosis with residuals. A residual is a signal that is zero when the system under diagnosis is fault-free, and non-zero when particular faults are present in the system. Residuals are typically generated by using a mathematical model of the system and measurements from sensors and actuators. This process is referred to as residual generation.</p><p>The main contributions in this thesis are two novel methods for residual generation. In both methods, systems described by Differential-Algebraic Equation (DAE) models are considered. Such models appear in a large class of technical systems, for example automotive systems. The first method consider observer-based residual generation for linear DAE-models. This method places no restrictions on the model, such as e.g. observability or regularity, in comparison with other previous methods. If the faults of interest can be detected in the system, the output from the design method is a residual generator, in state-space form, that is sensitive to the faults of interest. The method is iterative and relies on constant matrix operations, such as e.g. null-space calculations and equivalence transformations.</p><p>In the second method, non-linear DAE-models are considered. The proposed method belongs to a class of methods, in this thesis referred to as sequential residual generation, which has shown to be successful for real applications. This method enables simultaneous use of integral and derivative causality, and is able to handle equation sets corresponding to algebraic and differential loops in a systematic manner. It relies on a formal framework for computing unknown variables in the model according to a computation sequence, in which the analytical properties of the equations in the model as well as the available tools for equation solving are taken into account. The method is successfully applied to complex models of an automotive diesel engine and a hydraulic braking system.</p>

<p>Diagnosis means detection and isolation of faults. A model based diagnosis system is built on a mathematical model of the system. The difficulty when constructing the diagnosis system depends om how the model is formulated. In this report, a method is described that rewrites the model on such a form that the construction of the diagnosis algoritm is easy. The model is transformed by two state space transformations and the result will be a system on state space form where one part of the system becomes easy to supervise. </p><p>The main part of the report describes the procedure to create these transformations, which can be done in seven steps, based on differential geometric methods. </p><p>The aim of this masters thesis was to create an implementation in Mathematica (a computer tool for symbolic formula manipulation) of the creation of the two transformations and the system transformation. The created functions are described and examples of these are given. </p><p>A further aim was to evaluate if Mathematica could be a good support to rewrit a model. This was done by studying examples, and on the basis of the examples, identify difficult and easy steps. </p><p>The program has shown to be a good aid. Two of the seven steps have been identified as difficult and proposals for improvements have been given.</p>

<p>Within the next years, the amount of consumption data will increase rapidly as old meters will be exchanged in favor of meters with hourly remote reading. A new refined supervision system must be developed. The main objective of this thesis is to investigate mathematical methods that can be used to find incorrect hourly measurements in district heat and electricity consumption, for each consumer. </p><p>A simulation model and a statistical model have been derived. The model parameters in the simulation model are estimated by using historical data of consumption and outdoor temperature. By using the outdoor temperature as input, the consumption can be simulated and compared to the actual consumption. Faults are detected by using a residual with a sliding window. The second model uses the fact that consumers with similar consumption patterns can be grouped into a collective. By studying the correlation between the consumers, incorrect measurements can be found. </p><p>The performed simulations show that the simulation model is best suited for consumers whose consumption is mostly affected by the outdoor temperature. These consumers are district heat consumers and electricity consumers that use electricity for space heating. The fault detection performance of the statistical model is highly dependent on finding a collective that is well correlated. If these collectives can be found, the model can be used on district heat consumers as well as electricity consumers.</p>