Recent advances in power electronics have lead to the wide spread adoption of advanced power supplies and energy efficient devices. This has lead to increased levels of harmonic currents in power systems, degrading the performance of electrical machinery and interfering with telecommunication services. Active filters provide a solution to these problems by compensating for the distorted currents drawn by non-linear loads. Optimal methods for controlling these active filters have been determined by computer simulation and experimental implementation. Methods used for isolating the harmonic content of an unbalanced three phase load current were compared by computer simulations. A technique based on the Fast Fourier Transform (FFT) was developed as part of this work and shown to perform favourably. Notch Filtering, Sinusoidal Subtraction, Instantaneous Reactive Power Theory, Synchronous Reference Frame and Fast Fourier Transform methods were simulated. The methods shown to be suitable for compensation of three phase unbalanced loads were implemented in a Digital Signal Processor to evaluate true performance. These methods were Notch Filtering, Sinusoidal Subtraction, Fast Fourier Transform, and a High Pass Filter based method. A completely digital hysteresis current controller for a three phase active filter inverter has been developed and implemented with a Field Programmable Gate Array. This controller interfaces directly to a digital signal processor and is resistant to electromagnetic interference. Results from the experimental hardware verified that the active filter model used for simulation is accurate, and may be used for further development of harmonic isolation methods. A technique using notch filtering gives the best performance for steady loads, with the FFT based technique giving the most flexible operation for a range of load current characteristics. Novel use of the FFT based harmonic isolation technique allows selective cancellation of individual harmonics, with particular application to multiple shunt filters connected in parallel.

Αρχικά εξετάζεται η φυσιολογία της καρδιάς και η λειτουργία της ως αντλίας, οι διάφορες φάσεις του καρδιακού κύκλου και τα αίτια των φαινομένων που παρατηρούνται σε αυτόν. Στη συνέχεια μελετώνται οι καρδιακοί ήχοι, η προέλευσή τους και μέθοδοι παρατήρησής τους, το ηλεκτροκαρδιογράφημα, τα χαρακτηριστικά του και οι εφαρμογές του. Ακολούθως εξετάζεται η δειγματοληψία, επεξεργασία και ανάλυση βιολογικών σημάτων. Πιο συγκεκριμένα, μελετώνται οι μέθοδοι και οι αλγόριθμοι με τους οποίους λαμβάνονται τα ηλεκτροκαρδιογραφικά και φωνοκαρδιογραφικά σήματα, απομακρύνεται ο θόρυβος που προστίθεται από διάφορες πηγές και ανιχνεύονται συγκεκριμένα χαρακτηριστικά των κυματομορφών των σημάτων και ιδιότητές τους, για να αναλυθούν και να γίνει διάγνωση δυσλειτουργιών και παθολογιών της καρδιάς. Τέλος, αναπτύσσεται κώδικας συνεχούς δειγματοληψίας τεσσάρων καναλιών και καταγραφής των δεδομένων δειγματοληψίας σε μια κάρτα μνήμης πολυμέσων, προγραμματίζοντας τον μικροεπεξεργαστή MSP430F169 της Texas Instruments. / In the first part, the physiology of the heart and its functionality as a pump are studied, considering the different phases of a heart cycle and the phenomena observed within. Furthermore, heart sounds, their origin and methods of detection are examined, alongwith the electrocardiogram, its characteristics and applications. Secondly, sampling, processing and analysing methods of biosignals are described. In particular,techniques for ECG and PCG signal acquisition, noise reduction and detection of unique properties and characteristics are studied for analysis and diagnosis of heart disorders and pathologies. Finally, a code for continuous sampling of four channels and storing the sampled data into a multimedia memory card is developed, programming the MSP430F169 microcontroller, by Texas Instruments.

Υλοποίηση ενός επεξεργαστή διακριτών ηχητικών σημάτων, με χρήση μεθόδων STFT. Ο επεξεργαστής έχει την ικανότητα να διαχωρίζει τα μεταβατικά φαινόμενα και εφαρμόζει διαφορετικού τύπου επεξεργασία σε αυτά, από ότι στον υπόλοιπο όγκο του σήματος εισόδου. / Implementation of a processor of discrete audio signals, using STFT (overlap-add) methods. This processor can successfully detect transients in the input signal and apply a separate filter to them.

Στην παρούσα διπλωματική εργασία γίνεται μελέτη της ανάλυσης και σύνθεσης ήχου με τη βοήθεια ημιτονοειδών. Ειδικότερα, εξετάζονται οι παράμετροι της ανάλυσης και σύνθεσης και πως αυτες επηρεάζουν την τελική ανακατασκευή του σήματος. Στη συνέχεια γίνεται διερεύνηση της ανάλυσης και σύνθεσης μόνο στις χαμηλές συχνότητες. Με βάση ένα περιορισμένο εύρος ζώνης, γίνεται ανίχνευση των τονικών υψών. Αναπτύσσονται τρεις μέθοδοι κατηγοριοποίησης τους και στη συνέχεια γίνεται μία αξιολόγηση των μεθόδων αυτών μέσω των μέτρων NMR και PEAQ. / In this degree thesis sound analysis and synthesis using sinusoidals is studied. More specifically, parameters of analysis and synthesis are examined and how they affect the final reconstruction of a signal. Further research is conducted for analysis and synthesis at low sound frequencies. Based on a limited bandwidth, pitch detection is taking place on the input signal. Three methods of categorizing frequencies are developed and they are evaluated using the metrics of NMR (Noise to Mask Ratio) and PEAQ (Perceptual Evaluation of Audio Quality).

Thesis (MScEng (Electrical and Electronic Engineering))--University of Stellenbosch, 2011. / ENGLISH ABSTRACT: Keywords: digital signal processing, embedded systems, telecommunications, satellite technology. The rapid growth of the telecommunications industry is a worldwide phenomenon with people and computers generating and transmitting more and more information daily. Despite this growth, there are still areas in South Africa which lack terrestrial communications coverage. People inhabit these rural areas and their essential communication needs are not met. Satellite based communication coverage can provide a valuable service in these circumstances. In this thesis, the design of a satellite-based communications payload which makes use of software de ned radio techniques is presented in terms of the Open Systems Interconnect layer structure. A robust hardware platform using a space-quali ed on-board computer, a Xilinx Virtex-5 Field Programmable Gate Array (FPGA) and a Freescale digital signal processor (DSP) is designed, implemented and thoroughly tested. A device driver is designed for hardware and rmware components. A prototype ground station is also designed and constructed using a low-power PC, a Xilinx Spartan-3E FPGA, a Freescale DSP and radio frequency hardware. A wide range of testing methodologies were successfully utilised to deploy a functional system which is critically evaluated in the last chapter. / AFRIKAANSE OPSOMMING: Sleutelwoorde: syferseinverwerking, toegewyde stelsels, telekommunikasie, satelliettegnologie. Die vinnig groeiende telekommunikasieindustrie is 'n wêreldwye verskynsel waarin mense en rekenaars daagliks meer en meer data genereer. Ten spyte van die groei, is daar nog steeds gebiede in Suid-Afrika wat aan 'n gebrek van aardse kommunikasiedekking lei. Mense bewoon dié areas maar daar word nie aan hul noodsaaklike kommunikasiebehoeftes voldoen nie. Satelliet-gebaseerde kommunikasiedekking kan 'n waardevolle diens in hierdie omstandighede wees. Hierdie tesis beskryf die ontwerp van 'n ruimtegebaseerde kommunikasieloonvrag wat gebruik maak van sagteware-gede nieerde radiotegnieke aangebied in terme van die Open Systems Interconnect laagstruktuur. 'n Robuuste apparatuurplatform wat gebruik maak van 'n ruimte-gekwali seerde rekenaar, 'n Xilinx Virtex-5 Veldprogrameerbare Hek-Skikking (VPHS) en 'n Freescale syferseinverwerker is ontwerp, geïmplementeer en deeglik getoets. 'n Toestelbestuurder moes ontwerp word vir die apparatuur- en fermatuur-komponente. 'n Prototipe grondstasie is ook ontwerp en gebou met behulp van' n lae-krag PC, 'n Xilinx Spartan-3E VPHS, 'n Freescale seinverwerker en radiofrekwensie apparatuur. 'n Wye verskeidenheid van toetsmetodes is suksesvol benut om 'n funksionele stelsel te ontwikkel wat krities geëvalueer word in die laaste hoofstuk.

Les circuits logiques "Rapid Single-Flux-Quantum" (RSFQ) à base de jonctions Josephson supraconductrices sont utilisés pour générer, traiter et transmettre des impulsions ultra-courtes dont l'aire quantifiée est celle du quantum de flux magnétique h/2e et correspond à 2.07 mV.ps. De tels circuits sont utilisés pour traiter le signal à très haute fréquence avec des fréquences d'horloge dans la gamme 10-120 GHz et une puissance consommée environ 100 à 1000 fois plus faible (incluant le coût énergétique du refroidissement à 4,2 K) que celle des meilleurs circuits semi-conducteurs équivalents. La logique RSFQ est une alternative intéressante pour les super-ordinateurs et offre des performances inégalées pour traiter les signaux micro-ondes à la volée. Une fois les signaux numérisés et traités à température cryogénique, le défi principal est de transférer à température ambiante les signaux numériques de faible tension (dans la gamme 200-1000µV) à des débits de 1 à 10 Gbps par voie, tout en limitant la charge thermique sur le système de réfrigération cryogénique, afin de construire un système performant à très haut débit numérique. Une solution à ce verrou est de transmettre les signaux par un système d'émission-réception sans fil avec la bande passante suffisante. Ce travail examine différents systèmes de transmission sans fil à courte distance, correspondant à la configuration physique entre les étages à températures cryogénique et ambiante, pour des taux de transmission de quelques Gbps. Il s'est construit sur quatre points cruciaux à résoudre :• le choix et l'étude du codage numérique approprié pour être utilisé comme support de transmission en bande de base des signaux sans utiliser de modulation analogique, comme les codages Polar Return-to-Zero et Manchester ;• l'étude et la sélection d'antennes ultra large bande avec une attention particulière portée sur les antennes Vivaldi antipodales et les antennes monopôles pour satisfaire aux contraintes cryogéniques ;• l'étude du taux d'erreur du système de transmission. Deux méthodes ont été développées pour récupérer les signaux numériques et minimiser le taux d'erreur ;• la comparaison entre simulations et mesures afin d'évaluer la performance du système global. / Rapid Single-Flux-Quantum (RSFQ) logic circuits based on superconducting Josephson junctions are using to generate, process and transmit very short quantized pulses whose area is the quantum of magnetic flux h/2e and corresponds to 2.07 mV.ps. Such circuits are used to process signals at very high speed with clock frequencies in the 10-120 GHz range and a power consumption about 100 to 1000 times lower that their best available semiconductor counterparts (including the cost of cooling down to 4,2K). RSFQ logic is an interesting alternative for supercomputers and offers unsurpassed performances for processing microwave signals on the fly. Once digital signals are processed at cryogenic temperature the key challenge is to transfer at room temperature the low-voltage output digital signals (about 200-1000µV) at high rates of about 1-10Gbps per channel, by limiting the thermal burden on the cryogenic system, in order to build high performance high throughput systems.A solution is to transmit the signals with a wireless emitting-receiving antenna set with a suitable bandwidth. This work examines several wireless baseband transmission systems in a short distance configuration, associated to the distance between the cryogenic and room temperature stages, for data rates in the range of a few Gbps. It elaborates on four crucial issues :• the choice and study of the proper line codes to be used for baseband transmission of digital signals without the need for analogue modulations, such as Polar Return-to-Zero and Manchester encodings ;• the study and selection of ultra-wide bandwidth antennas with a focus on small size Antipodal Vivaldi Antennas and monopole antennas to meet cryogenic constraints ;• the study of the Bit Error Rate (BER) of the transmitting system. Two methods were developed to recover the digital output signals and minimize the BER.• the comparison between simulations and measurements to assess the performance of the overall system.

Este trabalho envolve o emprego de recentes tecnologias ligadas a promissora área de Inteligência Computacional e a tradicional área de Processamento de Sinais Digitais. Tem por objetivo o desenvolvimento de uma aplicação especifica na área de Processamento de Voz: o reconhecimento de locutor. Inúmeras aplicações, ligadas principalmente a segurança e controle, são possíveis a partir do domínio da tecnologia de reconhecimento de locutor, tanto no que diz respeito a identificação quanto a verificação de diferentes locutores. O processo de reconhecimento de locutor pode ser dividido em duas grandes fases: extração das características básicas do sinal de voz e classificação. Na fase de extração, procurou-se aplicar os mais recentes avanços na área de Processamento Digital de Sinais ao problema proposto. Neste contexto, foram utilizadas a frequência fundamental e as frequências formantes como parâmetros que identificam o locutor. O primeiro foi obtido através do use da autocorrelação e o segundo foi obtido através da transformada de Fourier. Estes parâmetros foram extraídos na porção da fala onde o trato vocal apresenta uma coarticulação entre dois sons vocálicos. Esta abordagem visa extrair as características desta mudança do aparato vocal. Existem dois tipos de reconhecimento de locutor: identificação (busca-se reconhecer o locutor em uma população) e verificação (busca-se verificar se a identidade alegada é verdadeira). O processo de reconhecimento de locutor é dividido em duas grandes fases: extração das características (envolve aquisição, pré-processamento e extração dos parâmetros característicos do sinal) e classificação (envolve a classificação do sinal amostrado na identificação/verificação do locutor ou não). São apresentadas diversas técnicas para representação do sinal, como analise espectral, medidas de energia, autocorrelação, LPC (Linear Predictive Coding), entre outras. Também são abordadas técnicas para extração de características do sinal, como a frequência fundamental e as frequências formantes. Na fase de classificação, pode-se utilizar diversos métodos convencionais: Cadeias de Markov, Distância Euclidiana, entre outros. Além destes, existem as Redes Neurais Artificiais (RNAs) que são consideradas poderosos classificadores. As RNAs já vêm sendo utilizadas em problemas que envolvem classificações de sinais de voz. Neste trabalho serão estudados os modelos mais utilizados para o problema de reconhecimento de locutor. Assim, o tema principal da Dissertação de Mestrado deste autor é a implementação de um sistema de reconhecimento de locutor utilizando Redes Neurais Artificiais para classificação do locutor. Neste trabalho tamb6m é apresentada uma abordagem para a implementação de um sistema de reconhecimento de locutor utilizando as técnicas convencionais para o processo de classificação do locutor. As técnicas utilizadas são Dynamic Time Warping (DTW) e Vector Quantization (VQ). / This work deals with the application of recent technologies related to the promising research domain of Intelligent Computing (IC) and to the traditional Digital Signal Processing area. This work aims to apply both technologies in a Voice Processing specific application which is the speaker recognition task. Many security control applications can be supported by speaker recognition technology, both in identification and verification of different speakers. The speaker recognition process can be divided into two main phases: basic characteristics extraction from the voice signal and classification. In the extraction phase, one proposed goal was the application of recent advances in DSP theory to the problem approached in this work. In this context, the fundamental frequency and the formant frequencies were employed as parameters to identify the speaker. The first one was obtained through the use of autocorrelation and the second ones were obtained through Fourier transform. These parameters were extracted from the portion of speech where the vocal tract presents a coarticulation between two voiced sounds. This approach is used to extract the characteristics of this apparatus vocal changing. In this work, the Multi-Layer Perceptron (MLP) ANN architecture was investigated in conjunction with the backpropagation learning algorithm. In this sense, some main characteristics extracted from the signal (voice) were used as input parameters to the ANN used. The output of MLP, trained previously with the speakers features, returns the authenticity of that signal. Tests were performed with 10 different male speakers, whose age were in the range from 18 to 24 years. The results are very promising. In this work it is also presented an approach to implement a speaker recognition system by applying conventional methods to the speaker classification process. The methods used are Dynamic Time Warping (DTW) and Vector Quantization (VQ).

Este trabalho propõe um algoritmo rápido para segmentação de partes móveis em vídeo, tendo como base a detecção de volumes fechados no espaço tridimensional. O vídeo de entrada é pré-processado com um algoritmo de detecção de bordas baseado em linhas de nível para produzir os objetos. Os objetos detectados são agrupados utilizando uma combinação dos métodos de mean shift clustering e meta-agrupamento. Para diminuir o tempo de computação, somente alguns objetos e quadros são utilizados no agrupamento. Uma vez que a forma de detecção garante que os objetos persistem com o mesmo rótulo em múltiplos quadros, a seleção de quadros impacta pouco no resultado final. Dependendo da aplicação desejada os grupos podem ser refinados em uma etapa de pós-processamento. / This work presents a very fast algorithm to segmentation of moving parts in a video, based on detection of surfaces of the scene with closed contours. The input video is preprocessed with an edge detection algorithm based on level lines to produce the objects. The detected objects are clustered using a combination of mean shift clustering and ensemble clustering. In order decrease even more the computation time required, two methods can be used combined: object filtering by size and selecting only a few frames of the video. Since the detected objects are coherent in time, frame skipping does not affect the final result. Depending on the application the detected clusters can be refined using post processing steps.

This research focuses on a novel integrated approach for computing and representing complex numbers as a single entity without the use of any dedicated multiplier for calculating the fast Fourier transform algorithm (FFT), using the Distributed Arithmetic (DA) technique and Complex Binary Number Systems (CBNS). The FFT algorithm is one of the most used and implemented technique employed in many Digital Signal Processing (DSP) applications in the field of science, engineering, and mathematics. The DA approach is a technique that is used to compute the inner dot product between two vectors without the use of any dedicated multipliers. These dedicated multipliers are fast but they consume a large amount of hardware and are quite costly. The DA multiplier process is accomplished by shifting and adding only without the need of any dedicated multiplier. In today's technology, complex numbers are computed using the divide and conquer approach in which complex numbers are divided into two parts: the real and imaginary. The CBNS technique however, allows for each complex addition and multiplication to be computed in one single step instead of two. With the combined DA-CBNS approach for computing the FFT algorithm, those dedicated multipliers are being replaced with a DA system that utilize a Rom-based memory for storing the twiddle factor 'wn' value and the complex arithmetic operations being represented as a single entity, not two, with the CBNS approach. This architectural design was implemented by coding in a very high speed integrated circuit (VHSIC) hardware description language (VHDL) using Xilinx ISE design suite software program version 14.2. This computer aided tool allows for the design to be synthesized to a logic gate level in order to be further implemented onto a Field Programmable Gate Array (FPGA) device. The VHDL code used to build this architecture was downloaded on a Nexys 4 DDR Artix-7 FPGA board for further testing and analysis. This novel technique resulted in the use of no dedicated multipliers and required half the amount of complex arithmetic computations needed for calculating an FFT structure compared with its current traditional approach. Finally, the results showed that for the proposed architecture design, for a 32 bit, 8-point DA-CBNS FFT structure, the results showed a 32% area reduction, 41% power reduction, 59% reduction in run-time, 42% reduction in logic gate cost, and 66% increase in speed. For a 28 bit, 16-point DA-CBNS FFT structure, its area size, power consumption, run-time, and logic gate, were also found to be reduced at approximately 30%, 37%, 60%, and 39%, respectively, with an increase of speed of approximately 67% when compared to the traditional approach that employs dedicated multipliers and computes its complex arithmetic as two separate entities: the real and imaginary.

Este trabalho envolve o emprego de recentes tecnologias ligadas a promissora área de Inteligência Computacional e a tradicional área de Processamento de Sinais Digitais. Tem por objetivo o desenvolvimento de uma aplicação especifica na área de Processamento de Voz: o reconhecimento de locutor. Inúmeras aplicações, ligadas principalmente a segurança e controle, são possíveis a partir do domínio da tecnologia de reconhecimento de locutor, tanto no que diz respeito a identificação quanto a verificação de diferentes locutores. O processo de reconhecimento de locutor pode ser dividido em duas grandes fases: extração das características básicas do sinal de voz e classificação. Na fase de extração, procurou-se aplicar os mais recentes avanços na área de Processamento Digital de Sinais ao problema proposto. Neste contexto, foram utilizadas a frequência fundamental e as frequências formantes como parâmetros que identificam o locutor. O primeiro foi obtido através do use da autocorrelação e o segundo foi obtido através da transformada de Fourier. Estes parâmetros foram extraídos na porção da fala onde o trato vocal apresenta uma coarticulação entre dois sons vocálicos. Esta abordagem visa extrair as características desta mudança do aparato vocal. Existem dois tipos de reconhecimento de locutor: identificação (busca-se reconhecer o locutor em uma população) e verificação (busca-se verificar se a identidade alegada é verdadeira). O processo de reconhecimento de locutor é dividido em duas grandes fases: extração das características (envolve aquisição, pré-processamento e extração dos parâmetros característicos do sinal) e classificação (envolve a classificação do sinal amostrado na identificação/verificação do locutor ou não). São apresentadas diversas técnicas para representação do sinal, como analise espectral, medidas de energia, autocorrelação, LPC (Linear Predictive Coding), entre outras. Também são abordadas técnicas para extração de características do sinal, como a frequência fundamental e as frequências formantes. Na fase de classificação, pode-se utilizar diversos métodos convencionais: Cadeias de Markov, Distância Euclidiana, entre outros. Além destes, existem as Redes Neurais Artificiais (RNAs) que são consideradas poderosos classificadores. As RNAs já vêm sendo utilizadas em problemas que envolvem classificações de sinais de voz. Neste trabalho serão estudados os modelos mais utilizados para o problema de reconhecimento de locutor. Assim, o tema principal da Dissertação de Mestrado deste autor é a implementação de um sistema de reconhecimento de locutor utilizando Redes Neurais Artificiais para classificação do locutor. Neste trabalho tamb6m é apresentada uma abordagem para a implementação de um sistema de reconhecimento de locutor utilizando as técnicas convencionais para o processo de classificação do locutor. As técnicas utilizadas são Dynamic Time Warping (DTW) e Vector Quantization (VQ). / This work deals with the application of recent technologies related to the promising research domain of Intelligent Computing (IC) and to the traditional Digital Signal Processing area. This work aims to apply both technologies in a Voice Processing specific application which is the speaker recognition task. Many security control applications can be supported by speaker recognition technology, both in identification and verification of different speakers. The speaker recognition process can be divided into two main phases: basic characteristics extraction from the voice signal and classification. In the extraction phase, one proposed goal was the application of recent advances in DSP theory to the problem approached in this work. In this context, the fundamental frequency and the formant frequencies were employed as parameters to identify the speaker. The first one was obtained through the use of autocorrelation and the second ones were obtained through Fourier transform. These parameters were extracted from the portion of speech where the vocal tract presents a coarticulation between two voiced sounds. This approach is used to extract the characteristics of this apparatus vocal changing. In this work, the Multi-Layer Perceptron (MLP) ANN architecture was investigated in conjunction with the backpropagation learning algorithm. In this sense, some main characteristics extracted from the signal (voice) were used as input parameters to the ANN used. The output of MLP, trained previously with the speakers features, returns the authenticity of that signal. Tests were performed with 10 different male speakers, whose age were in the range from 18 to 24 years. The results are very promising. In this work it is also presented an approach to implement a speaker recognition system by applying conventional methods to the speaker classification process. The methods used are Dynamic Time Warping (DTW) and Vector Quantization (VQ).