Most signal processing methods were developed for continuous signals. Digital devices, such as the computer, process only discrete signals. This dissertation proposes new techniques to accurately define and efficiently implement an important signal processing method---the time--frequency distribution (TFD)---using discrete signals. The TFD represents a signal in the joint time--frequency domain. Because these distributions are a function of both time and frequency they, unlike traditional signal processing methods, can display frequency content that changes over time. TFDs have been used successfully in many signal processing applications as almost all real-world signals have time-varying frequency content. Although TFDs are well defined for continuous signals, defining and computing a TFD for discrete signals is problematic. This work overcomes these problems by making contributions to the definition, computation, and application of discrete TFDs. The first contribution is a new discrete definition of TFDs. A discrete TFD (DTFD) should be free from the sampling-related distortion known as aliasing and satisfy all the important mathematical properties that the continuous TFD satisfies. Many different DTFD definitions exist but none come close to attaining this ideal. I propose three new components which make up the DTFD: 1) a new discrete Wigner--Ville distribution (DWVD) definition which satisfies all properties, 2) a new discrete analytic signal which minimises aliasing in the DWVD, and 3) a new method to define and convolve the discrete kernel with the DWVD to produce the DTFD. The result: a DTFD definition that, relative to the existing definitions, better approximates the ideal DTFD. The second contribution is two sets of computationally efficient algorithms to compute the proposed DTFD. The first set of algorithms computes the DTFD exactly; the second set requires less memory than the first set by computing time- and, or frequency-decimated versions of the DTFD. Both sets of algorithms reduce the computational load by exploiting symmetries in the DTFD and by constructing kernel-specific algorithms for four different kernel types. The third, and final, contribution is a biomedical application for the proposed DTFD and algorithms. This application is to accurately detect seizure events in newborn electroencephalogram (EEG) signals. Existing detection methods do not perform well enough for use in a clinical setting. I propose a new method which is more robust than existing methods and show how using the proposed DTFD, comparative to an existing DTFD, improves detection performance for this method. In summary, this dissertation makes practical contributions to the area of time--frequency signal processing by proposing an improved DTFD definition, efficient DTFD algorithms, and an improved newborn EEG seizure detection method using DTFDs.

Most signal processing methods were developed for continuous signals. Digital devices, such as the computer, process only discrete signals. This dissertation proposes new techniques to accurately define and efficiently implement an important signal processing method---the time--frequency distribution (TFD)---using discrete signals. The TFD represents a signal in the joint time--frequency domain. Because these distributions are a function of both time and frequency they, unlike traditional signal processing methods, can display frequency content that changes over time. TFDs have been used successfully in many signal processing applications as almost all real-world signals have time-varying frequency content. Although TFDs are well defined for continuous signals, defining and computing a TFD for discrete signals is problematic. This work overcomes these problems by making contributions to the definition, computation, and application of discrete TFDs. The first contribution is a new discrete definition of TFDs. A discrete TFD (DTFD) should be free from the sampling-related distortion known as aliasing and satisfy all the important mathematical properties that the continuous TFD satisfies. Many different DTFD definitions exist but none come close to attaining this ideal. I propose three new components which make up the DTFD: 1) a new discrete Wigner--Ville distribution (DWVD) definition which satisfies all properties, 2) a new discrete analytic signal which minimises aliasing in the DWVD, and 3) a new method to define and convolve the discrete kernel with the DWVD to produce the DTFD. The result: a DTFD definition that, relative to the existing definitions, better approximates the ideal DTFD. The second contribution is two sets of computationally efficient algorithms to compute the proposed DTFD. The first set of algorithms computes the DTFD exactly; the second set requires less memory than the first set by computing time- and, or frequency-decimated versions of the DTFD. Both sets of algorithms reduce the computational load by exploiting symmetries in the DTFD and by constructing kernel-specific algorithms for four different kernel types. The third, and final, contribution is a biomedical application for the proposed DTFD and algorithms. This application is to accurately detect seizure events in newborn electroencephalogram (EEG) signals. Existing detection methods do not perform well enough for use in a clinical setting. I propose a new method which is more robust than existing methods and show how using the proposed DTFD, comparative to an existing DTFD, improves detection performance for this method. In summary, this dissertation makes practical contributions to the area of time--frequency signal processing by proposing an improved DTFD definition, efficient DTFD algorithms, and an improved newborn EEG seizure detection method using DTFDs.

Most signal processing methods were developed for continuous signals. Digital devices, such as the computer, process only discrete signals. This dissertation proposes new techniques to accurately define and efficiently implement an important signal processing method---the time--frequency distribution (TFD)---using discrete signals. The TFD represents a signal in the joint time--frequency domain. Because these distributions are a function of both time and frequency they, unlike traditional signal processing methods, can display frequency content that changes over time. TFDs have been used successfully in many signal processing applications as almost all real-world signals have time-varying frequency content. Although TFDs are well defined for continuous signals, defining and computing a TFD for discrete signals is problematic. This work overcomes these problems by making contributions to the definition, computation, and application of discrete TFDs. The first contribution is a new discrete definition of TFDs. A discrete TFD (DTFD) should be free from the sampling-related distortion known as aliasing and satisfy all the important mathematical properties that the continuous TFD satisfies. Many different DTFD definitions exist but none come close to attaining this ideal. I propose three new components which make up the DTFD: 1) a new discrete Wigner--Ville distribution (DWVD) definition which satisfies all properties, 2) a new discrete analytic signal which minimises aliasing in the DWVD, and 3) a new method to define and convolve the discrete kernel with the DWVD to produce the DTFD. The result: a DTFD definition that, relative to the existing definitions, better approximates the ideal DTFD. The second contribution is two sets of computationally efficient algorithms to compute the proposed DTFD. The first set of algorithms computes the DTFD exactly; the second set requires less memory than the first set by computing time- and, or frequency-decimated versions of the DTFD. Both sets of algorithms reduce the computational load by exploiting symmetries in the DTFD and by constructing kernel-specific algorithms for four different kernel types. The third, and final, contribution is a biomedical application for the proposed DTFD and algorithms. This application is to accurately detect seizure events in newborn electroencephalogram (EEG) signals. Existing detection methods do not perform well enough for use in a clinical setting. I propose a new method which is more robust than existing methods and show how using the proposed DTFD, comparative to an existing DTFD, improves detection performance for this method. In summary, this dissertation makes practical contributions to the area of time--frequency signal processing by proposing an improved DTFD definition, efficient DTFD algorithms, and an improved newborn EEG seizure detection method using DTFDs.

Most signal processing methods were developed for continuous signals. Digital devices, such as the computer, process only discrete signals. This dissertation proposes new techniques to accurately define and efficiently implement an important signal processing method---the time--frequency distribution (TFD)---using discrete signals. The TFD represents a signal in the joint time--frequency domain. Because these distributions are a function of both time and frequency they, unlike traditional signal processing methods, can display frequency content that changes over time. TFDs have been used successfully in many signal processing applications as almost all real-world signals have time-varying frequency content. Although TFDs are well defined for continuous signals, defining and computing a TFD for discrete signals is problematic. This work overcomes these problems by making contributions to the definition, computation, and application of discrete TFDs. The first contribution is a new discrete definition of TFDs. A discrete TFD (DTFD) should be free from the sampling-related distortion known as aliasing and satisfy all the important mathematical properties that the continuous TFD satisfies. Many different DTFD definitions exist but none come close to attaining this ideal. I propose three new components which make up the DTFD: 1) a new discrete Wigner--Ville distribution (DWVD) definition which satisfies all properties, 2) a new discrete analytic signal which minimises aliasing in the DWVD, and 3) a new method to define and convolve the discrete kernel with the DWVD to produce the DTFD. The result: a DTFD definition that, relative to the existing definitions, better approximates the ideal DTFD. The second contribution is two sets of computationally efficient algorithms to compute the proposed DTFD. The first set of algorithms computes the DTFD exactly; the second set requires less memory than the first set by computing time- and, or frequency-decimated versions of the DTFD. Both sets of algorithms reduce the computational load by exploiting symmetries in the DTFD and by constructing kernel-specific algorithms for four different kernel types. The third, and final, contribution is a biomedical application for the proposed DTFD and algorithms. This application is to accurately detect seizure events in newborn electroencephalogram (EEG) signals. Existing detection methods do not perform well enough for use in a clinical setting. I propose a new method which is more robust than existing methods and show how using the proposed DTFD, comparative to an existing DTFD, improves detection performance for this method. In summary, this dissertation makes practical contributions to the area of time--frequency signal processing by proposing an improved DTFD definition, efficient DTFD algorithms, and an improved newborn EEG seizure detection method using DTFDs.

Most signal processing methods were developed for continuous signals. Digital devices, such as the computer, process only discrete signals. This dissertation proposes new techniques to accurately define and efficiently implement an important signal processing method---the time--frequency distribution (TFD)---using discrete signals. The TFD represents a signal in the joint time--frequency domain. Because these distributions are a function of both time and frequency they, unlike traditional signal processing methods, can display frequency content that changes over time. TFDs have been used successfully in many signal processing applications as almost all real-world signals have time-varying frequency content. Although TFDs are well defined for continuous signals, defining and computing a TFD for discrete signals is problematic. This work overcomes these problems by making contributions to the definition, computation, and application of discrete TFDs. The first contribution is a new discrete definition of TFDs. A discrete TFD (DTFD) should be free from the sampling-related distortion known as aliasing and satisfy all the important mathematical properties that the continuous TFD satisfies. Many different DTFD definitions exist but none come close to attaining this ideal. I propose three new components which make up the DTFD: 1) a new discrete Wigner--Ville distribution (DWVD) definition which satisfies all properties, 2) a new discrete analytic signal which minimises aliasing in the DWVD, and 3) a new method to define and convolve the discrete kernel with the DWVD to produce the DTFD. The result: a DTFD definition that, relative to the existing definitions, better approximates the ideal DTFD. The second contribution is two sets of computationally efficient algorithms to compute the proposed DTFD. The first set of algorithms computes the DTFD exactly; the second set requires less memory than the first set by computing time- and, or frequency-decimated versions of the DTFD. Both sets of algorithms reduce the computational load by exploiting symmetries in the DTFD and by constructing kernel-specific algorithms for four different kernel types. The third, and final, contribution is a biomedical application for the proposed DTFD and algorithms. This application is to accurately detect seizure events in newborn electroencephalogram (EEG) signals. Existing detection methods do not perform well enough for use in a clinical setting. I propose a new method which is more robust than existing methods and show how using the proposed DTFD, comparative to an existing DTFD, improves detection performance for this method. In summary, this dissertation makes practical contributions to the area of time--frequency signal processing by proposing an improved DTFD definition, efficient DTFD algorithms, and an improved newborn EEG seizure detection method using DTFDs.

Nous nous intéressons aux méthodes de contrôle de vibrations de modèles réduits de structures à n degrés de liberté, sismiquement isolées au niveau de la base par des systèmes d'isolation mixte.Le mouvement provoqué par une sollicitation sismique horizontale a lieu dans le plan vertical.Nous avons construit un problème de contrôle semi-actif de systèmes incertains soumis à des perturbations inconnues, mais bornées. Dans le langage de l'automatique, il s'agit d'un problème d'atténuation de perturbations.Le résultat principal de cette thèse porte sur la construction d'une version modifiée des résultats de Leitmann et de ses collaborateurs sur la stabilisation de systèmes non linéaires incertains. Le théorème proposé repose sur une loi de commande par retour d'état qui assure en boucle fermée les propriétés de "uniform boundedness" et "uniform ultimate boundedness".En particulier, il peut être appliqué à la résolution de problèmes de contrôle semi-actif, qui sont actuellement traités en génie parasismique.L'objectif du contrôle est d'améliorer le comportement (i.e. la réponse) de structures isolées pour faire face aux perturbations externes, c'est-à-dire les séismes. Plusieurs points différencient notre problème de la majorité que l'on trouve dans la littérature: (i) on ne s'intéresse pas seulement à la protection de la structure isolée, mais aussi aux équipements situés à l'intérieur de la structure, et (ii) au lieu d'utiliser des indicateurs de performance habituels exprimés en termes de déplacement relatif de la base versus des accélérations absolues des planchers, nous utilisons uniquement le spectre de plancher en pseudo-accélération, comme il a été proposé dans des travaux précédents par Politopoulos et Pham. Ce travail est une tentative d'utiliser explicitement les spectres de plancher comme critère de performance.Concernant la procédure d'application, plusieurs étapes intermédiaires ont été détaillées:(i) modélisation de signaux sismiques;(ii) réglage des paramètres de la loi de commande utilisant la théorie des vibrations;(iii) validation et test du comportement en boucle fermée à travers des simulations numériques: pour des raisons de simplicité, on se limite au cas n=2.Cette procédure peut être utilisée sur des structures en industrie nucléaire, mais aussi en génie civil.D'autres sujets traités incluent une tentative d'utiliser les outils temps-fréquence, et en particulier la distribution de Wigner-Ville, pour la synthèse de lois de commande, en espérant pouvoir mieux contrôler les composants transitoires des signaux de perturbation (les entrées) et des variables d'état (les sorties). / Vibration attenuation control designs are proposed for reduced plant models consisting of n-degree-of-freedom base seismically-isolated structures (i.e., a specific type of earthquake-resistant design), modeled by uncertain nonlinear systems and subjected to one-dimensional horizontal ground acceleration (i.e. the earthquake signal), treated as unknown disturbance but assumed to be bounded.In control systems literature, this is a perturbation attenuation problem.The main result of this PhD is the development of a modified version of Leitmann and co-authors' classical result on the stabilization of uncertain nonlinear systems. The proposed theorem consists of a bounded nonlinear feedback control law that is capable of ensuring uniform boundedness and uniform ultimate boundedness in closed-loop. In particular, it can be applied to solving semi-active control design problems, which are currently dealt with in earthquake engineering.The control objective is to improve the behavior (i.e. response) of mixed base-isolated structures to external disturbance, namely earthquakes. What differentiates our problem from the majority to be found in the literature is that: (i) attention is being paid to the protection of equipment placed inside the structure an not only to the structure itself; (ii) instead of using regular performance indicators expressed in terms of relative base displacement versus floors accelerations, we use solely the pseudo-acceleration floor response spectra, as it was proposed in previous recent works by Politopoulos and Pham.Actually, this work is an attempt to explicitly use floor response spectra as performance criterion.Concerning the application procedure, some of the topics that were detailed are:(i) modeling of earthquake signals;(ii) tuning of control law parameters based on vibration theory;(iii) validation and testing of the closed-loop behavior using numerical simulations: for simplicity reasons, we take n=2.This procedure can be used on structures of both nuclear industry as well as civil engineering.Other topics include an attempt to using time-frequency concepts and in particular the Wigner-Ville distribution to the control law design procedure, in order to better control transitory components of both perturbation (the input) and state variables signals (the output).

Les entraînements électriques à base de machine asynchrone sont largement utilisés dans les applications industrielles en raison de leur faible coût, de leurs performances et de leur robustesse. Cependant, des modes de fonctionnement dégradés peuvent apparaître durant la vie de la machine. L'une des raisons principales de ces défaillances reste les défauts de roulements à billes. Afin d'améliorer la sûreté de fonctionnement des entraînements, des schémas de surveillance peuvent être mis en place afin d'assurer une maintenance préventive. Ce travail de thèse traite de la détection et du diagnostic des défauts mécaniques et plus particulièrement des défauts de roulements dans une machine asynchrone. Généralement, une surveillance vibratoire peut être mise en place. Cette méthode de surveillance est cependant souvent chère du fait de la chaîne de mesure. Une approche, basée sur l'analyse et le traitement des courants statoriques, est alors proposée, afin de suppléer à l'analyse vibratoire. L'étude est basée sur l'existence et la caractérisation des effets des oscillations du couple de charge sur les courants d'alimentation. Un schéma de détection est alors introduit pour détecter différents types de défauts de roulements. De plus, des variables mécaniques, telles que la vitesse ou le couple, sont également reconstruites afin de fournir une indication sur la présence de défauts de roulements. Par ailleurs, un diagnostic des modulations des courants statoriques est proposé, en régime permanent et en régime transitoire, quel que soit le rapport entre les fréquences porteuse et modulante. Les méthodes étudiées sont la transformée de Hilbert, la transformée de Concordia, l'amplitude et la fréquence instantanées ainsi que la distribution de Wigner-Ville. / Asynchronous drives are widely used in many industrial applications because of their low cost, high performance and robustness. However, faulty operations may appear during the lifetime of the system. The most frequently encountered faults in asynchronous drives come from rolling bearings. To improve the availability and reliability of the drives, a condition monitoring may be implemented to favor the predictive maintenance. This Ph.D. thesis deals with detection and diagnosis of mechanical faults, particularly rolling bearings defects in induction motors. Traditionally, bearing monitoring is supervised using vibration analysis. Measuring such quantities is often expensive due to the measurement system. An other approach, based on stator current analysis, is then proposed. The characterization of load torque oscillation effects on stator currents is studied. A detection scheme is then proposed to detect several types of bearing faults. Moreover, mechanical variables, such as rotating speed or torque, are estimated in order to detect bearings defects. In addition, a diagnosis of stator currents modulations is proposed, in steady and transient state, whatever the career and modulation frequencies. Hilbert transform, Concordia transform, instantaneous amplitude and frequency are studied. The Wigner-Ville distribution is used in transient state.

Die Arbeit befaßt sich mit Methoden der Identifikation von Schwingungsursachen anhand gemessener Schwingungssignale an Maschinen mit ungleichförmig übersetzenden Getrieben. Den technischen Hintergrund bilden die Verarbeitungsmaschinen, die als ein Haupteinsatzgebiet der auch als Mechanismen bezeichneten Getriebe gelten. Es werden eine ganze Reihe mechanismentypischer Schwingungsursachen untersucht, die im wesentlichen auf Elastizitäten, veränderliche Parameter, Spiel, Reibung und Unstetigkeiten in den Lagefunktionen zurückzuführen sind. Neben den traditionellen Methoden der Signalanalyse gilt das Hauptaugenmerk den neuen Methoden der Zeit-Frequenz-Analyse, die dem zumeist instationären Charakter der durch die Mechanismen verursachten Schwingungen gerecht werden. Dazu zählen vor allem lineare Transformationen wie die Kurzzeit-Fourier- oder die Wavelet-Transformation, quadratische Verteilungen aus Cohen's Klasse wie die Wigner-Ville und die Choi-Williams Verteilung oder auch höhere Transformationen wie die Adaptive Optimal Kernel oder die Reassignment Methode. Die Untersuchungen münden in einer tabellarischen Zusammenstellung der gefundenen Merkmale. Diese soll der Unterstützung der praktischen Signalanalyse bei Mechanismenschwingungen dienen. Die Anwendung der zur Verfügung stehenden Mittel und Methoden wird an zahlreichen Beispielen aus dem Bereich der Schwingungsdiagnose an Verarbeitungsmaschinen demonstriert. / This thesis deals with methods to identify causes of vibrations by investigation of messured signals at machines with mechanisms (with varying velocity ratio). The technical background is the field of manufacturing machines, the main application of such mechanisms. Inspected are the most typical causes of vibrations at mechanisms that comes with elasticity, variable parameters, clearance, friction or nonsmooth transferfunctions. There are used both, traditional methods of signal analysis and time-frequency-analysis methods that are especially advantageous for the often instationary vibrations at this kind of machines. Linear transformations like the Short-Time-Fourier- or the Wavelet-Transformation, quadratic distributions from Cohe's Class like the Wigner-Ville- or the Choi-Williams-Distribution and higher order transformations like the Adaptive Optimal Kernel or the Reassignment Method are used in this Paper. The investigation leads into a tabular form of characteristics wich should be used for identification of vibrational causes during vibrational analysis. The use of available methods is demonstrated by solving various examples on real manufactoring machines.

Die Arbeit befaßt sich mit Methoden der Identifikation von Schwingungsursachen anhand gemessener Schwingungssignale an Maschinen mit ungleichförmig übersetzenden Getrieben. Den technischen Hintergrund bilden die Verarbeitungsmaschinen, die als ein Haupteinsatzgebiet der auch als Mechanismen bezeichneten Getriebe gelten. Es werden eine ganze Reihe mechanismentypischer Schwingungsursachen untersucht, die im wesentlichen auf Elastizitäten, veränderliche Parameter, Spiel, Reibung und Unstetigkeiten in den Lagefunktionen zurückzuführen sind. Neben den traditionellen Methoden der Signalanalyse gilt das Hauptaugenmerk den neuen Methoden der Zeit-Frequenz-Analyse, die dem zumeist instationären Charakter der durch die Mechanismen verursachten Schwingungen gerecht werden. Dazu zählen vor allem lineare Transformationen wie die Kurzzeit-Fourier- oder die Wavelet-Transformation, quadratische Verteilungen aus Cohen's Klasse wie die Wigner-Ville und die Choi-Williams Verteilung oder auch höhere Transformationen wie die Adaptive Optimal Kernel oder die Reassignment Methode. Die Untersuchungen münden in einer tabellarischen Zusammenstellung der gefundenen Merkmale. Diese soll der Unterstützung der praktischen Signalanalyse bei Mechanismenschwingungen dienen. Die Anwendung der zur Verfügung stehenden Mittel und Methoden wird an zahlreichen Beispielen aus dem Bereich der Schwingungsdiagnose an Verarbeitungsmaschinen demonstriert. - - - Diese Version ersetzt eine ältere Version, Grund: Formatkonvertierung postscript zu pdf. Der Inhalt unterscheidet sich in der Qualität der Abb. 4.23a und 4.23b auf Seite 97 sowie aller Abbildungen auf Seite 111. Der Zugriff auf die Original-Abb. ist über die ältere Version oder die gedruckte Ausgabe möglich. Die gedruckte Ausgabe der Dissertation ist in der Universitätsbibliothek Chemnitz entleihbar. / This thesis deals with methods to identify causes of vibrations by investigation of messured signals at machines with mechanisms (with varying velocity ratio). The technical background is the field of manufacturing machines, the main application of such mechanisms. Inspected are the most typical causes of vibrations at mechanisms that comes with elasticity, variable parameters, clearance, friction or nonsmooth transferfunctions. There are used both, traditional methods of signal analysis and time-frequency-analysis methods that are especially advantageous for the often instationary vibrations at this kind of machines. Linear transformations like the Short-Time-Fourier- or the Wavelet-Transformation, quadratic distributions from Cohe's Class like the Wigner-Ville- or the Choi-Williams-Distribution and higher order transformations like the Adaptive Optimal Kernel or the Reassignment Method are used in this Paper. The investigation leads into a tabular form of characteristics wich should be used for identification of vibrational causes during vibrational analysis. The use of available methods is demonstrated by solving various examples on real manufactoring machines. - - - This version replaces an older version, reason is conversion of format from postscript to pdf. Content is different in quality of pictures 4.23a and 4.23b on page 97 and all pictures on page 111. Please see the original pictures in the older version or the printed version. The printed version is borrowable at the Library of Chemnitz University of Technology.

This thesis deals with time-frequency analysis of electrograms measured on isolated guinea pig hearts perfused according to Langendorff. Time-frequency analysis is based on algorithms Matching Pursuit and Wigner-Ville Distribution. The theoretical part describes the basics of electrocardiography, measurement on isolated hearts, the theory of approximation method Matching Pursuit and its combination with the Wigner-Ville distribution spectrum showing the energy density of the signal. Also other common approaches of time-frequency analysis are presented including the theory of continuous wavelet transform. The presented algorithms were tested on a set of electrograms, on which were induced ischemia within measurement followed by reperfusion. The proposed method allows for the fast detection of ischemia without any a priori knowledge of the signal, and also serves as a tool for measurement of EG important points and intervals. In the conclusion efficacy of the method was presented and its possible uses has been discussed.