Multiclassificador inteligente de falhas no domínio do tempo em motores de indução trifásicos alimentados por inversores de frequência

Os motores de indução desempenham um importante papel na indústria, fato este que destaca a importância do correto diagnóstico e classificação de falhas ainda em fase inicial de sua evolução, possibilitando aumento na produtividade e, principalmente, eliminando graves danos aos processos e às máquinas. Assim, a proposta desta tese consiste em apresentar um multiclassificador inteligente para o diagnóstico de motor sem defeitos, falhas de curto-circuito nos enrolamentos do estator, falhas de rotor e falhas de rolamentos em motores de indução trifásicos acionados por diferentes modelos de inversores de frequência por meio da análise das amplitudes dos sinais de corrente de estator no domínio do tempo. Para avaliar a precisão de classificação frente aos diversos níveis de severidade das falhas, foram comparados os desempenhos de quatro técnicas distintas de aprendizado de máquina; a saber: (i) Rede Fuzzy Artmap, (ii) Rede Perceptron Multicamadas, (iii) Máquina de Vetores de Suporte e (iv) k-Vizinhos-Próximos. Resultados experimentais obtidos a partir de 13.574 ensaios experimentais são apresentados para validar o estudo considerando uma ampla faixa de frequências de operação, bem como regimes de conjugado de carga em 5 motores diferentes.

On-line system for faults detection in induction motors based on PCA

Dissertation to obtain the degree of Master in Electrical and Computer Engineering

Evaluation of the applicability of FRA for inter-turn fault detection in stator windings

In this paper, the applicability of the FRA technique is discussed as a method for detecting inter-turn faults in stator windings. Firstly, this method is tested in an individual medium-voltage stator coil with satisfactory results. Secondly, the tests are extended to a medium-voltage induction motor stator winding, in which inter-turn faults are performed in every coil end of one phase. Results of the frequency response in case of inter-turn faults are evaluated in both cases for different fault resistance values. The experimental setup is also described for each experiment. The results of the application of this technique to the detection of inter-turn faults justify further research in optimizing this technique for preventive maintenance.

Evaluation of the Grounding Circuit Measurements for Stator Ground-Fault Location of Synchronous Generators

Locating stator-winding ground faults accurately is a very difficult task. In this paper the grounding circuit measurements are evaluated in order to obtain information about the stator ground-fault location in synchronous generators. In power generators grounded through a high impedance, the relation between the neutral voltage and the phase voltage provide a first estimation of the fault location. The location error by using this ratio depends on the fault resistance and the value of the capacitance to ground of the stator winding. However, the error added by ignoring the value of the fault resistance is the most relevant term. This location estimation and the location error have been evaluated through the data of a real synchronous machine.

Hybrid Method to Assess Sensitive Process Interruption Costs Due to Faults in Electric Power Distribution Networks

This paper shows a new hybrid method for risk assessment regarding interruptions in sensitive processes due to faults in electric power distribution systems. This method determines indices related to long duration interruptions and short duration voltage variations (SDVV), such as voltage sags and swells in each customer supplied by the distribution network. Frequency of such occurrences and their impact on customer processes are determined for each bus and classified according to their corresponding magnitude and duration. The method is based on information regarding network configuration, system parameters and protective devices. It randomly generates a number of fault scenarios in order to assess risk areas regarding long duration interruptions and voltage sags and swells in an especially inventive way, including frequency of events according to their magnitude and duration. Based on sensitivity curves, the method determines frequency indices regarding disruption in customer processes that represent equipment malfunction and possible process interruptions due to voltage sags and swells. Such approach allows for the assessment of the annual costs associated with each one of the evaluated power quality indices.

Evolution of stress deficit and changing rates of seismicity in cellular automaton models of earthquake faults

We investigate the internal dynamics of two cellular automaton models with heterogeneous strength fields and differing nearest neighbour laws. One model is a crack-like automaton, transferring ail stress from a rupture zone to the surroundings. The other automaton is a partial stress drop automaton, transferring only a fraction of the stress within a rupture zone to the surroundings. To study evolution of stress, the mean spectral density. f(k(r)) of a stress deficit held is: examined prior to, and immediately following ruptures in both models. Both models display a power-law relationship between f(k(r)) and spatial wavenumber (k(r)) of the form f(k(r)) similar tok(r)(-beta). In the crack model, the evolution of stress deficit is consistent with cyclic approach to, and retreat from a critical state in which large events occur. The approach to criticality is driven by tectonic loading. Short-range stress transfer in the model does not affect the approach to criticality of broad regions in the model. The evolution of stress deficit in the partial stress drop model is consistent with small fluctuations about a mean state of high stress, behaviour indicative of a self-organised critical system. Despite statistics similar to natural earthquakes these simplified models lack a physical basis. physically motivated models of earthquakes also display dynamical complexity similar to that of a critical point system. Studies of dynamical complexity in physical models of earthquakes may lead to advancement towards a physical theory for earthquakes.

Effect of material anisotropy on the onset of convective flow in three-dimensional fluid-saturated faults

In order to investigate the effect of material anisotropy on convective instability of three-dimensional fluid-saturated faults, an exact analytical solution for the critical Rayleigh number of three-dimensional convective flow has been obtained. Using this critical Rayleigh number, effects of different permeability ratios and thermal conductivity ratios on convective instability of a vertically oriented three-dimensional fault have been examined in detail. It has been recognized that (1) if the fault material is isotropic in the horizontal direction, the horizontal to vertical permeability ratio has a significant effect on the critical Rayleigh number of the three-dimensional fault system, but the horizontal to vertical thermal conductivity ratio has little influence on the convective instability of the system, and (2) if the fault material is isotropic in the fault plane, the thermal conductivity ratio of the fault normal to plane has a considerable effect on the critical Rayleigh number of the three-dimensional fault system, but the effect of the permeability ratio of the fault normal to plane on the critical Rayleigh number of three-dimensional convective flow is negligible.

Diagnóstico de avarias em motores elétricos

Trabalho Final de Mestrado para obtenção do grau de Mestre em Engenharia Mecânica

How a cyber-physical system can efficiently obtain a snapshot of physical information even in the presence of sensor faults

We present a distributed algorithm for cyber-physical systems to obtain a snapshot of sensor data. The snapshot is an approximate representation of sensor data; it is an interpolation as a function of space coordinates. The new algorithm exploits a prioritized medium access control (MAC) protocol to efficiently transmit information of the sensor data. It scales to a very large number of sensors and it is able to operate in the presence of sensor faults.

Analyzing the effects of transient faults into applications

As computer chips implementation technologies evolve to obtain more performance, those computer chips are using smaller components, with bigger density of transistors and working with lower power voltages. All these factors turn the computer chips less robust and increase the probability of a transient fault. Transient faults may occur once and never more happen the same way in a computer system lifetime. There are distinct consequences when a transient fault occurs: the operating system might abort the execution if the change produced by the fault is detected by bad behavior of the application, but the biggest risk is that the fault produces an undetected data corruption that modifies the application final result without warnings (for example a bit flip in some crucial data). With the objective of researching transient faults in computer system’s processor registers and memory we have developed an extension of HP’s and AMD joint full system simulation environment, named COTSon. This extension allows the injection of faults that change a single bit in processor registers and memory of the simulated computer. The developed fault injection system makes it possible to: evaluate the effects of single bit flip transient faults in an application, analyze an application robustness against single bit flip transient faults and validate fault detection mechanism and strategies.

Weakening processes in thrust faults: insights from the Monte Perdido thrust fault (southern Pyrenees, Spain)

The Monte Perdido thrust fault (southern Pyrenees) consists of a 6-m-thick interval of intensely deformed clay-bearing rocks. The fault zone is affected by a pervasive pressure solution seam and numerous shear surfaces. Calcite extensional-shear veins are present along the shear surfaces. The angular relationships between the two structures indicate that shear surfaces developed at a high angle (70°) to the local principal maximum stress axis r1. Two main stages of deformation are present. The first stage corresponds to the development of calcite shear veins by a combination of shear surface reactivation and extensional mode I rupture. The second stage of deformation corresponds to chlorite precipitation along the previously reactivated shear surfaces. The pore fluid factor k computed for the two deformation episodes indicates high fluid pressures during the Monte Perdido thrust activity. During the first stage of deformation, the reactivation of the shear surface was facilitated by a suprahydrostatic fluid pressure with a pore fluid factor kv equal to 0.89. For the second stage, the fluid pressure remained still high (with a k value ranging between 0.77 and 0.84) even with the presence of weak chlorite along the shear surfaces. Furthermore, evidence of hydrostatic fluid pressure during calcite cement precipitation supports that incremental shear surface reactivations are correlated with cyclic fluid pressure fluctuations consis- tent with a fault-valve model.

Fluid migration through thrust faults in the Helvetic nappes (Western Swiss Alps)

The delta(18)O, delta(13)C and Sr-87/Sr-86 values of calcite and organic matter were measured for carbonate mylonites from numerous thrusts in the Helvetic Alps. Carbonate mylonites in most of the thrusts retain essentially unaltered protolith delta(18)O and delta(13)C values, consistent with there having been little to no advection of isotopically distinct fluid through these faults. Only carbonate mylonites from the basal thrusts of the largest nappes have delta(18)O and/or delta(13)C values that differ from those of their protoliths. The zone of isotopic alteration/exchange is confined to c. 10 to 20 meters of these fault contacts. We propose the fluids that migrated through these faults contained variable amounts of organically derived carbon and radiogenic strontium, and were probably derived from dewatering of the sedimentary rocks and prograde metamorphic reactions in the nappes' root zones. Apart from the basal thrusts of the largest nappes that behaved as narrow, laterally extensive conduits for fluids, there is little isotopic evidence that large quantities of fluids passed through most of the carbonate-hosted thrusts in the Helvetic Alps.

The Quaternary Active Faults Database of Iberia (QAFI v.2.0)

The Quaternary Active Faults Database of Iberia (QAFI) is an initiative lead by the Institute of Geology and Mines of Spain (IGME) for building a public repository of scientific data regarding faults having documented activity during the last 2.59 Ma (Quaternary). QAFI also addresses a need to transfer geologic knowledge to practitioners of seismic hazard and risk in Iberia by identifying and characterizing seismogenic fault-sources. QAFI is populated by the information freely provided by more than 40 Earth science researchers, storing to date a total of 262 records. In this article we describe the development and evolution of the database, as well as its internal architecture. Aditionally, a first global analysis of the data is provided with a special focus on length and slip-rate fault parameters. Finally, the database completeness and the internal consistency of the data are discussed. Even though QAFI v.2.0 is the most current resource for calculating fault-related seismic hazard in Iberia, the database is still incomplete and requires further review.

On the Condition Monitoring of Induction machines

The research of condition monitoring of electric motors has been wide for several decades. The research and development at universities and in industry has provided means for the predictive condition monitoring. Many different devices and systems are developed and are widely used in industry, transportation and in civil engineering. In addition, many methods are developed and reported in scientific arenas in order to improve existing methods for the automatic analysis of faults. The methods, however, are not widely used as a part of condition monitoring systems. The main reasons are, firstly, that many methods are presented in scientific papers but their performance in different conditions is not evaluated, secondly, the methods include parameters that are so case specific that the implementation of a systemusing such methods would be far from straightforward. In this thesis, some of these methods are evaluated theoretically and tested with simulations and with a drive in a laboratory. A new automatic analysis method for the bearing fault detection is introduced. In the first part of this work the generation of the bearing fault originating signal is explained and its influence into the stator current is concerned with qualitative and quantitative estimation. The verification of the feasibility of the stator current measurement as a bearing fault indicatoris experimentally tested with the running 15 kW induction motor. The second part of this work concentrates on the bearing fault analysis using the vibration measurement signal. The performance of the micromachined silicon accelerometer chip in conjunction with the envelope spectrum analysis of the cyclic bearing faultis experimentally tested. Furthermore, different methods for the creation of feature extractors for the bearing fault classification are researched and an automatic fault classifier using multivariate statistical discrimination and fuzzy logic is introduced. It is often important that the on-line condition monitoring system is integrated with the industrial communications infrastructure. Two types of a sensor solutions are tested in the thesis: the first one is a sensor withcalculation capacity for example for the production of the envelope spectra; the other one can collect the measurement data in memory and another device can read the data via field bus. The data communications requirements highly depend onthe type of the sensor solution selected. If the data is already analysed in the sensor the data communications are needed only for the results but in the other case, all measurement data need to be transferred. The complexity of the classification method can be great if the data is analysed at the management level computer, but if the analysis is made in sensor itself, the analyses must be simple due to the restricted calculation and memory capacity.