The overall aim of this thesis on congenital clubfoot was to estimate the incidence with a national perspective, analyse residual deformities and their management, and evaluate patient reported long-term quality of life and foot function. Paper I was a prospective, nationwide sampling of 280 children with congenital clubfoot during 1995-96. The average incidence was 1.4‰. There was regional heterogeneity but no seasonal variation. Paper II evaluated ultrasonography on 54 newborn, prospectively followed up to 12 months of age. Significant increase of medial malleolus to navicular distance (MM-N-distance) and of soft tissue thickness with increasing age was seen and with acceptable reliability. Paper III assessed 35 children (47 feet) after previous posterior release, mean age of 4.5 years, and the MM-N-distance was shorter in unilateral clubfeet (21 patients) than in contralateral normal feet. No association between navicular position and forefoot adduction (FFA) was determined. Smaller FFA yielded better subjective and functional outcome. Paper IV reviewed distraction treatment with Ilizarov External Fixator in seven patients (10 feet), 6-15 years of age, with relapsed deformities. All patients, except one, reported satisfaction with the overall result but less stiffness was experienced in only 4/10 feet. Paper V evaluated self-estimated outcome in 83 patients (63 males, 20 females), mean age of 64 years, through SF-36 and EQ-5D, and through AAOS foot and ankle score. Age and gender adjusted norm groups were used. Female patients scored worse than male patients did. Both males and females reported negative influence on foot and ankle function. Conclusion: The incidence of congenital clubfoot in Sweden is higher than in previous Scandinavian studies. Ultrasonography is reliable for describing pathoanatomy of the talo-navicular joint in clubfeet and can detect "spurious" (false) correction. Distraction treatment with the Ilizarov External Fixator yields subjective improvement but stiffness remains. Long-term influence on daily life activities is limited to foot and ankle function for both genders, but only female patients report negative influence on physical aspects of quality of life.

Today&amp / #8217 / s analog modulators and upconverters are inadequate to synthesize and modulate signals with high &amp / #8216 / Spurious Free Dynamic Range&amp / #8217 / (SFDR). Thus, the main objective of this thesis is to design and implement a &amp / #8216 / Digital Signal Synthesizer&amp / #8217 / (DSS) that is capable of synthesizing signals between 50-100 MHz with 60dB SFDR and to modulate them variable symbol rates and modulation techniques with very high phase/frequency resolution and switching speed while keeping the amplitude modulation occurring during a modulated symbol duration as small as possible. In this thesis, digital words of the desired signals are first synthesized in a &amp / #8216 / Field Programmable Gate Array&amp / #8217 / (FPGA) using &amp / #8216 / Direct Digital Synthesizer&amp / #8217 / (DDS) fundamentals and then converted to analog signals with a high speed &amp / #8216 / Digital to Analog Converter&amp / #8217 / (DAC). In order to attain the analog requirements, the system variables such as DAC analog performance, nonlinearities, sample and hold affects, DDS parameters, system clock, bandwidth requirements of analog filters and how they effect the output performance are studied. FPGA blocks that are capable of modulating and synthesizing desired signals are designed and programmed on a FPGA. Finally, single tone and modulated signals are synthesized with this DSS implementation and measured in order to verify this system&amp / #8217 / s performance and capabilities.

The random fluctuations of signal phase of an oscillator limit the precision of time and frequency measurements. The noise and long-term stability of the system’s oscillator or clock is of major importance in applications such as optical and wireless communications, high-speed digital electronics, radar, and astronomy. The Optoelectronic Oscillator (OE Oscillator), a new class of time delay oscillator with promise as a low-phase noise source of microwave carriers, was introduced by Steve Yao and Lute Malek in 1996. The OE Oscillator combines into a closed loop an RF photonic link and an RF chain. The RF photonic link consists of a laser, electro-optic modulator, optical fibre delay line, and a photo-receiver that together provide an RF delay. An RF chain consists of one or more amplifiers and a RF resonator that together provide the sustaining amplification and the frequency selectivity necessary for single mode oscillation of the loop. The low loss of optical fibres enables the attainment of delays that correspond to optical fibre lengths of several kilometers. It is the long delay, unattainable in an all electronic implementations that is responsible for the superior phase noise performance of an OE Oscillator. In this thesis the fundamental principles of operation of an OE Oscillator are described and the principal sources of in-loop phase fluctuations that are responsible for phase-noise identified. This lays the ground for an exposition of the mechanism that describes the perturbation of a time delay oscillator by injection into the loop of a carrier that is detuned in frequency from the natural frequency of the oscillator. For sufficiently small detuning the oscillator can become phase locked to the injected carrier. The model presented in the thesis generalises the traditional Yao-Maleki and Leeson model to include all the important features that describe the injection locking dynamics of an OE Oscillator. In particular the common assumptions of single mode oscillation and weak injection are removed. This is important to correctly predict the effect of injection locking on the spurious peaks in the phase noise spectrum corresponding to the side-modes of a time delay oscillator. Simulation results are presented in order to validate the dynamics of the oscillator under injection and analytic results on the lock-in range and phase noise spectrum. A 10 GHz OE Oscillator with a single 5km delay line is used as an example in the simulation illustration.

L’apprentissage machine a parcouru beaucoup de chemin avec des succès marquants ces dernières années. Pourtant, les réseaux de neurones font encore des erreurs surprenantes en présence de corrélations factices. Le réseau basera sa décision sur des caractéristiques non pertinentes dans les données qui se corréleront avec son objectif lors de l'apprentissage. En revanche, les humains sont moins enclins à commettre de telles erreurs, car ils peuvent exploiter des sources d'information plus diverses pour leurs décisions. Bien que les représentations factorisées (démêlée) soient souvent présentées comme la solution pour une bonne généralisation, nous soutenons qu'un ensemble d'experts devrait apprendre aussi librement que possible avec des contraintes minimales, puis rééquilibrer leur sortie proportionnellement à leur similarité. Dans cette thèse, nous proposons une nouvelle approche : en nous inspirant de la façon dont les neurones biologiques sont connectés latéralement, nous introduisons un mécanisme pour rééquilibrer la décision des paires de neurones qui répondent de manière similaire aux entrées, favorisant la diversité de l'information par rapport à leur similarité. Nous démontrons par un cas de test extrême que l’inhibition locale peut avoir un impact positif sur la robustesse des décisions d’un réseau de neurones. / Machine learning has made remarkable strides in recent years, yet neural networks still make surprising mistakes in the presence of spurious correlations. The network will base its decision on irrelevant features in the data that happen to correlate with its objective during training. In contrast, humans are less prone to making such errors since we can leverage more diverse sources of information to make decisions. While maximally factorized or disentangled representations is often cited as the solution for generalization, we argue that an ensemble of experts should learn as freely as possible with minimal constraints and then rebalance their output proportional to their similarity. In this thesis, we propose a novel approach to do this: inspired by the lateral connections between biological neurons, we introduce a mechanism that rebalances the output of pairs of neurons that respond similarly to inputs, promoting information diversity over redundancy. We demonstrate by subjective the network to an extreme test case that local inhibition can have a positive impact on an ANN’s decision robustness.

No description available.

Given the ever increasing reliance of today’s society on ubiquitous wireless access, the paradigm of dynamic spectrum access (DSA) as been proposed and implemented for utilizing the limited wireless spectrum more efficiently. Orthogonal frequency division multiplexing (OFDM) is growing in popularity for adoption into wireless services employing DSA frame- work, due to its high bandwidth efficiency and resiliency to multipath fading. While these advantages have been proven for many wireless applications, including LTE-Advanced and numerous IEEE wireless standards, one potential drawback of OFDM or its non-contiguous variant, NC-OFDM, is that it exhibits high peak-to-average power ratios (PAPR), which can induce in-band and out-of-band (OOB) distortions when the peaks of the waveform enter the compression region of the transmitter power amplifier (PA). Such OOB emissions can interfere with existing neighboring transmissions, and thereby severely deteriorate the reliability of the DSA network. A performance-enhancing digital pre-distortion (DPD) technique compensating for PA and in-phase/quadrature (I/Q) modulator distortions is proposed in this dissertation. Al- though substantial research efforts into designing DPD schemes have already been presented in the open literature, there still exists numerous opportunities to further improve upon the performance of OOB suppression for NC-OFDM transmission in the presence of RF front-end impairments. A set of orthogonal polynomial basis functions is proposed in this dissertation together with a simplified joint DPD structure. A performance analysis is presented to show that the OOB emissions is reduced to approximately 50 dBc with proposed algorithms employed during NC-OFDM transmission. Furthermore, a novel and intuitive DPD solution that can minimize the power regrowth at any pre-specified frequency in the spurious domain is proposed in this dissertation. Conventional DPD methods have been proven to be able to effectively reduce the OOB emissions that fall on top of adjacent channels. However more spectral emissions in more distant frequency ranges are generated by employing such DPD solutions, which are potentially in violation of the spurious emission limit. At the same time, the emissions in adjacent channel must be kept under the OOB limit. To the best of the author’s knowledge, there has not been extensive research conducted on this topic. Mathematical derivation procedures of the proposed algorithm are provided for both memoryless nonlinear model and memory-based nonlinear model. Simulation results show that the proposed method is able to provide a good balance of OOB emissions and emissions in the far out spurious domain, by reducing the spurious emissions by 4-5 dB while maintaining the adjacent channel leakage ratio (ACLR) improvement by at least 10 dB, comparing to the PA output spectrum without any DPD.

Dans ce travail, on s'intéresse au développement d'une méthode hybride qui couple spatialement les schémas d'intégration temporelle explicite et implicite. Afin de répondre aux contraintes du solveur industriel FLUSEPA, les schémas explicite Heun et implicite Crank-Nicolson ont été hybridés via un paramètre de transition : l'approche mise en place est appelée schéma AION. Cette dernière est étudiée en détails avec une attention particulière sur son comportement spectral et sa capacité à maintenir l'ordre de précision. On montre que le traitement hybride a d'intéressants comportements dissipatif et dispersif tout en empêchant la réflexion d'ondes parasites et en maintenant la précision attendue. De plus, l'approche hybride est validée sur plusieurs cas académiques à la fois pour les flux convectifs et pour les flux diffusifs. Et comme espéré, la méthode est plus intéressante en terme de temps de calcul que les méthodes standards d'intégration temporelle. Pour l'extension de cette approche à la méthode temporelle adaptative présente dans FLUSEPA, il a été nécessaire d'améliorer le traitement qui permet à la méthode d’être conservative tout en obtenant des propriétés spectrales acceptables. Finalement l'approche hybride a été aussi étendue pour la modélisation RANS/LES de la turbulence avec des temps de calcul intéressants tout en capturant la physique de l'écoulement / This work deals with the design of a hybrid time integrator that couples spatially explicit and implicit time integrators. In order to cope with the industrial solver of Ariane Group called FLUSEPA, the explicit scheme of Heun and the implicit scheme of Crank-Nicolson are hybridized using the transition parameter : the whole technique is called AION time integration. The latter is studied into details with special focus on spectral behaviour and on its ability to keep the accuracy. It is shown that the hybrid technique has interesting dissipation and dispersion properties while maintaining precision and avoiding spurious waves. Moreover, this hybrid approach is validated on several academic test cases for both convective and diffusive fluxes. And as expected the method is more interesting in term of computational time than standard time integrators. For the extension of this hybrid approach to the temporal adaptive method implemented in FLUSEPA, it was necessary to improve some treatments in order to maintain conservation and acceptable spectral properties. Finally the hybrid time integration was also applied to a RANS/LES turbulent test case with interesting computational time while capturing the flow physics.

在線性迴歸分析中，資料的不適當，常導致研究者選擇了不當的模式，為避免此缺失，在分析資料前須先做好診斷工作。本文中將從貝氏觀點提出一些不同的診斷方法以供參考。首先推導出均數移動參數a=(a₁,…,a_k)'的事後分配，並利用a'a/k的事後均數診斷出不當資料點。接著，考慮在個別模式下以β事後分配之總變異及廣義變異為標準，診斷出離群值及具有潛在影響力之觀測值。最後，分別利用(i)β的事後分配(ii)σ²的事後分配(iii)(β,σ²)的聯合事後分配，推導出對應的對稱均方差以做為診斷標準。 / In this thesis, some different diagnostic methodologies for outliers and influential observations from Bayesian point of view are proposed. We firstly derive the marginal posterior distribution of the mean-shift parameter a=(a₁,a_k)¹, then use the posterior mean of a¹a/k to detect the spurious data items. Secondly, we use the posterior total variance and generalized variance of β as diagnostic criterions for outliers and influential observations. Finally, we utilize (i) the posterior distribution of β, (ii) the posterior distribution of σ², and (iii) the joint posterior distribution of β, σ² to find their corresponding symmetric mean square differences , which can be used as diagnostic criterions.

Une nouvelle méthode de test pour les convertisseurs ADC et DAC embarqués dans un système complexe a été développée en prenant en compte les nouvelles contraintes affectant le test. Ces contraintes, dues aux tendances de design de systèmes, sont un nombre réduit de point d'accès aux entrées/sorties des blocs analogiques du système et une augmentation galopante du nombre et des performances des convertisseurs intégrés. La méthode proposée consiste à connecter les convertisseurs DAC et ADC dans le domaine analogique pour n'avoir besoin que d'instruments de test numériques pour générer et capturer les signaux de test. Un algorithme de traitement du signal a été développé pour discriminer les erreurs des DACs et ADCs. Cet algorithme a été validé par simulation et par expérimentation sur des produits commercialisés par NXP. La dernière partie de la thèse a consisté à développer de nouvelles applications pour l'algorithme.

Dans ce travail, on s'intéresse aux méthodes numériques d'ordre élevé pour des écoulements instationnaires autour de géométries complexes. On commence par analyser l'approche hybride pour la méthode industrielle des Volumes Finis à l'ordre faible. Cela consiste à calculer en même temps sur des maillages structurés et non structurés avec des schémas numériques dédiés. Les maillages structurés et non structurés sont ensuite couplés par un raccord non conforme. Ce dernier est analysé en détails avec une attention particulière pour des écoulements instationnaires. On montre qu'un traitement dédié à l'interface empêche la réflexion d'ondes parasites. De plus, l'approche hybride est validée sur plusieurs cas académiques à la fois pour les flux convectifs et pour les flux diffusifs. L'extension de cette approche hybride à l'ordre élevé est limitée par l'efficacité des schémas non structurés d'ordre élevé en terme de temps de calcul. C'est pourquoi une nouvelle approche est explorée : la méthode des différences spectrales. Un nouveau cadre est spécialement développé pour réaliser l'analyse spectrale des méthodes spectrales discontinues. La méthode des différences spectrales semble être une alternative viable en terme de temps de calcul et de nombre de points par longueur d'onde nécessaires à une application donnée pour capturer la physique de l'écoulement. / This work deals with high-order numerical methods for unsteady flows around complex geometries. In order to cope with the low-order industrial Finite Volume Method, the proposed technique consists in computing on structured and unstructured zones with their associated schemes: this is called a hybrid approach. Structured and unstructured meshes are then coupled by a nonconforming grid interface. The latter is analyzed in details with special focus on unsteady flows. It is shown that a dedicated treatment at the interface avoids the reflection of spurious waves. Moreover, this hybrid approach is validated on several academic test cases for both convective and diffusive fluxes. The extension of this hybrid approach to high-order schemes is limited by the efficiency of unstructured high-order schemes in terms of computational time. This is why a new approach is explored: The Spectral Difference Method. A new framework is especially developed to perform the spectral analysis of Spectral Discontinuous Methods. The Spectral Difference Method seems to be a viable alternative in terms of computational time and number of points per wavelength needed for a given application to capture the flow physics.