This thesis includes two topics. The first one is the architecture as well as the circuit implementation of an ultra low-power direct digital frequency synthesizer (DDFS) based on the straight line approximation. The second one is the circuit implementation of the low-power DDFS with an error compensation. The proposed approximation technique replaces the conventional ROM-based phase-to-amplitude conversion circuitry and the linear digital-to-analog converter with a nonlinear digital-to-analog converter (DAC) to realize a simple approximation of the sine function. Thus, the overall power dissipation as well as hardware complexity can be significantly reduced. Besides, by adding the error compensation, the spurious-free dynamic range (SFDR) of the synthesized output signal can be raised drastically.

Recently, multimedia applications are used widely in many embedded and portable systems, such as mobile phones, MP3 player and PDA, which require lower power consumption within high performance constraints. Therefore, power-efficient design becomes a more important objective in Very Large Scale Integration (VLSI) designs. Moreover, the multiplication unit always lies on the critical path and ultimately determines the performance and power consumption of arithmetic computing systems. To achieve high-performance and lengthen the battery lifetime, it is crucial to develop a multiplier with high-speed and low power consumption. In multimedia and digital signal processing (DSP) applications, many low-power approaches have been presented to lessen the power consumption of multipliers by eliminating spurious computations. Moreover, the multiplication operations adopted in these systems usually allow accuracy loss to output data so as to achieve more power savings. Based on these conceptions, this dissertation considers input data characteristics and the arithmetic features of multiplications in various multimedia and DSP applications and presents novel power reduction and truncation techniques to design power-efficient multipliers and high-accuracy fixed-width multipliers. In the design of array and tree multipliers, we first propose a low power pipelined truncated multiplier which dynamically deactivates non-effective circuitry based on input range. Moreover, the proposed multiplier offers a flexible tradeoff between power reduction and product precision. This reconfigurable characteristic is very useful to systems which have different requirement on output precision. Second, a low-power configurable Booth multiplier that supports several multiplication modes and eliminates the redundant computations of sign bits in multipliers as much as possible is developed. This architecture can efficaciously decrease the power consumption of systems which demand computing performance and flexibility simultaneously. Although these two kinds of low power multipliers can achieve significant power savings, the hardware complexity of error compensation circuits and error performance in terms of the mean error and mean-square error are unsuitable for many multimedia systems composed of a large amount of multiply-accumulate operations. To efficiently improve the accuracy with less hardware complexity, we propose new error compensation circuits for fixed-width tree multipliers and fixed-width modified Booth multipliers. In the design of floating-point multipliers, we propose a low power variable-latency floating-point multiplier which is compliant with IEEE 754-1985 and suitable for 3-D graphics and multimedia applications. In the architecture, the significand multiplier is first partitioned into the upper and lower parts. Next, an efficient prediction scheme for the carry bit, sticky bit, and the upper part of significand product is developed. While the correct prediction occurs, the computation of lower part of significand multiplier is shut down and therefore the floating-point multiplication can consume less power and be completed early. In the design of modular multipliers, we propose an efficient modular multiplication algorithm to devise a high performance and low power modular multiplier. The proposed algorithm adopts the quotient pipelining and superfluous-operation elimination technique to discard the data dependency and redundant computational cycles of radix-2 Montgomery¡¦s multiplication algorithm so that the operation speed, power dissipation, and energy consumption of modular multipliers can be significantly improved.

The primary barrier to the production of better machined parts is machine tool error. Present day applications are requiring closer machine part tolerances. The errors in dimensional part accuracy derive from the machine, in this case, a vertical two axis CNC lathe. A two axis vertical lathe can be utilized to produce a variety of parts ranging from cylindrical features to spherical features. A vertical lathe requires a spindle to rotate the work at speeds reaching 3000rpm, while simultaneously requiring the machine tool to be positioned in such a manner to remove material and produce an accurate part. For this to be possible, the machine tool must be precisely controlled in order to produce the correct contours on the part. There are many sources of errors to be considered in the two axis vertical lathe. Each axis of importance contains six degrees of freedom. The machine has linear displacement, angular, spindle thermal drift, straightness, parallelism, orthogonal, machine tool offset and roundness error. These error components must be measured in order to determine the resultant error. The characterization of the machine addresses thermal behavior and geometric errors. This thesis presents the approach of determining the machine tool errors and using these errors to transform the actual tool path closer to the nominal tool path via compensation schemes. One of these schemes uses a laser interferometer in conjunction with a homogenous transformation matrix to construct the compensated path for a circular arc, facing and turning. The other scheme uses a ball bar system to directly construct the compensated tool path for a circular arc. Test parts were created to verify the improvement of the part accuracy using the compensated tool paths.

No description available.

The thesis deals with a development and application of laser interferometer system for the online error calibration of CNC machine tool. The term „online“ means, that laser interferometer is permanently built in the machine tool. Therefore particular error measurements can be done in very short time period.

This thesis, for the first time in the literature, presents the effect of quadrature error compensation on the performance of a fully decoupled MEMS gyroscope and provides experimental data on the sources of quadrature error. Dedicated quadrature error cancellation electrodes operating with only differential DC potentials are designed. Gyroscopes with intentionally placed imperfections are fabricated with SOG based SOI process which provides higher yield and uniformity compared to SOG process. Tests show that the fully closed loop system with quadrature cancellation operates as expected. Gyroscope performance is improved up to 7.8 times for bias instability, 10 times for angle random walk (ARW) and 800 times for output offset with quadrature cancellation. The actual improvement is higher since some sensors cannot be operated without quadrature cancellation and they are not included in improvement calculations. The best obtained performance is bias instability of 0.39

In this thesis, several machining strategies to remove axially symmetric form error with small wavelength by Hydrodynamic Polishing process (abbreviated as HDP) were proposed. Three strategies were proposed progressively in the study so as to remove axially symmetric form error with small wavelength. The first and second tactics were based on a basic algorithm, say, directly solving of a set of simultaneous equations. In the first strategy, a set of simultaneous equations was constructed by relating the total machining action of each dwelling point to the corresponding initial error. Subsequently, a set of dwelling time was obtained by directly solving the simultaneous equations. The second strategy evaluates solutions in a similar way like the first one but more restrictions were concerned in solution evaluation. The third strategy is an optimal based method. A set of dwelling time was obtained by minimizing an objective function with given constraints. A series of computer simulations were conducted to estimate the residual error and examine the validity of the strategies. From the computer simulation, the first and second strategies were confronted with negative-time problem, so that merely limited improving of form precision was obtained. The proposed optimal strategy was shown to have high potential for improving the machining precision by the HDP process. Based on the proposed strategies, a better form precision of the work surface with small wavelength can be obtained.

Precision control of translational motion is vital for many CNC machine tools as the motion of the machinery affects the dimensional tolerance of the manufactured goods. However, the direct measurement along with the accurate motion control of machine usually requires relatively expensive sensors i.e. potentiometers, linear scales, laser interferometers. Hence, this study attempts to develop reference models utilizing low-cost sensors (i.e. rotary encoders) for accurate position estimation. First, an indirect measurement performance is investigated on a Timing Belt driven carriage by a DC Motor with a backlash included Gearbox head. An advanced interpolated technique is proposed to compensate the position errors while using indirect measurement to reduce the total cost. Then, a similar study was realized with a ball screw driven system. Next, a cable drum driven measurement technique is proposed to the machines which have long travel distance like plasma cutters. A test setup is proposed and manufactured to investigate the capstan drive systems. Finally, characteristics of Optical Mouse Sensors are investigated from different point of views and a test setup is proposed and manufactured to evaluate their performances in long terms. Beside all of these parts, motion control algorithms and motion control integrated circuits are designed and manufactured to realize experimental studies in a detailed manner.

Submitted by Maike Costa (maiksebas@gmail.com) on 2017-05-22T13:42:24Z No. of bitstreams: 1 arquivo total.pdf: 4721146 bytes, checksum: 62290d20efdd12371f2308dd6752ec65 (MD5) / Made available in DSpace on 2017-05-22T13:42:24Z (GMT). No. of bitstreams: 1 arquivo total.pdf: 4721146 bytes, checksum: 62290d20efdd12371f2308dd6752ec65 (MD5) Previous issue date: 2016-02-26 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / The computer numerical control machine tools, or CNC machine tools, emerged due to the need of the aviation industry to produce increasingly complex and accurate parts. The retrofitting of machine tools is the modification of a traditional machine for a computer numerical control one. Due to the popularization of computers, electronic components decreased costs, Internet expansion and open source projects, the retrofitting has become increasingly widespread in industry as an option for the purchase of new machine tools with CNC included. The objective of this study was to implement the retrofit on a 3-axis machine from the Precision Engineering Laboratory of the Federal University of Paraíba. For this adaptation, the machine has been modified to use a kit with interface drivers from HobbyCNC and new stepper motors. After the retrofit, it used a 80-XL Laser system for measuring position errors. The tests were performed on each axis, commanding the machine to five positions and measuring the actual positions reached and, then, the same tests were carried out in reverse direction. With the measured results, the positioning errors, hysteresis, repeatability and accuracy calculated by ISO 230-2 standard. It was conducted other tests in compensating the previously measured errors on the Y axis. There was a significant improvement on the Y-axis accuracy after compensation. Also, angular errors tests were carried out, with the use of an electronic level Talyvel 5, along the axis Y. The results of the test were satisfactory and the retrofitting reached its expectations, providing the possibility of further research in the area of numerical control. / As máquinas-ferramentas com controle numérico computadorizado, ou máquinasferramentas CNC, surgiram devido à necessidade da indústria da aviação para produzir peças cada vez mais complexas e precisas. O retrofitting de máquina-ferramenta é a modificação de uma máquina tradicional para uma com controle numérico computadorizado. Devido à popularização dos computadores, diminuição de custos dos componentes eletrônicos, expansão da Internet e projetos de código aberto, o retrofitting tornou-se uma opção cada vez mais comum na indústria como alternativa para a compra de novas máquinas-ferramentas com CNC incluso. O objetivo deste estudo foi a implementação do retrofitting em uma máquina de 3 (três) eixos no Laboratório de Engenharia de Precisão da Universidade Federal da Paraíba. Para essa adaptação, a máquina foi modificada para utilizar um Kit com drivers de interface da HobbyCNC e novos motores de passo. Após o retrofitting, foi utilizado um sistema Laser XL-80 para medição dos erros de posicionamento. Os testes foram realizados, em cada eixo, comandando-se a máquina para 5 posições e medindo-se as posições reais atingidas e, depois, foram realizados os mesmos testes no sentido inverso. Com os resultados medidos, são calculados os erros de posicionamento, histerese, repetitividade e exatidão conforme a norma ISO 230-2. No eixo Y foi realizado outros testes com a compensação dos erros medidos anteriormente. Houve uma melhora significativa na exatidão do eixo Y após a compensação. Também foram realizados testes de erros angulares, com a utilização de um nível eletrônico Talyvel 5, ao longo do eixo Y. Os resultados obtidos no teste foram satisfatórios e o retrofitting atingiu suas expectativas, por proporcionar a possibilidade de realizar mais pesquisas, na área de Comando Numérico Computadorizado.

Technological developments in devices and storages have made large volumes of data collections more accessible than ever. This transformation leads to optimization problems with massive data in both volume and dimension. In response to this trend, the popularity of optimization on high performance computing architectures has increased unprecedentedly. These scalable optimization solvers can achieve high efficiency by splitting computational loads among multiple machines. However, these methods also incur large communication overhead. To solve optimization problems with millions of parameters, communication between machines has been reported to consume up to 80% of the training time. To alleviate this communication bottleneck, many optimization algorithms with data compression techniques have been studied. In practice, they have been reported to significantly save communication costs while exhibiting almost comparable convergence as the full-precision algorithms. To understand this intuition, we develop theory and techniques in this thesis to design communication-efficient optimization algorithms. In the first part, we analyze the convergence of optimization algorithms with direct compression. First, we outline definitions of compression techniques which cover many compressors of practical interest. Then, we provide the unified analysis framework of optimization algorithms with compressors which can be either deterministic or randomized. In particular, we show how the tuning parameters of compressed optimization algorithms must be chosen to guarantee performance. Our results show explicit dependency on compression accuracy and delay effect due to asynchrony of algorithms. This allows us to characterize the trade-off between iteration and communication complexity under gradient compression. In the second part, we study how error compensation schemes can improve the performance of compressed optimization algorithms. Even though convergence guarantees of optimization algorithms with error compensation have been established, there is very limited theoretical support which guarantees improved solution accuracy. We therefore develop theoretical explanations, which show that error compensation guarantees arbitrarily high solution accuracy from compressed information. In particular, error compensation helps remove accumulated compression errors, thus improving solution accuracy especially for ill-conditioned problems. We also provide strong convergence analysis of error compensation on parallel stochastic gradient descent across multiple machines. In particular, the error-compensated algorithms, unlike direct compression, result in significant reduction in the compression error. Applications of the algorithms in this thesis to real-world problems with benchmark data sets validate our theoretical results. / Utvecklandet av kommunikationsteknologi och datalagring har gjort stora mängder av datasamlingar mer tillgängliga än någonsin. Denna förändring leder till numeriska optimeringsproblem med datamängder med stor skala i volym och dimension. Som svar på denna trend har populariteten för högpresterande beräkningsarkitekturer ökat mer än någonsin tidigare. Skalbara optimeringsverktyg kan uppnå hög effektivitet genom att fördela beräkningsbördan mellan ett flertal maskiner. De kommer dock i praktiken med ett pris som utgörs av betydande kommunikationsomkostnader. Detta orsakar ett skifte i flaskhalsen för prestandan från beräkningar till kommunikation. När lösning av verkliga optimeringsproblem sker med ett stort antal parametrar, dominerar kommunikationen mellan maskiner nästan 80% av träningstiden. För att minska kommunikationsbelastningen, har ett flertal kompressionstekniker föreslagits i litteraturen. Även om optimeringsalgoritmer som använder dessa kompressorer rapporteras vara lika konkurrenskraftiga som sina motsvarigheter med full precision, dras de med en förlust av noggrannhet. För att ge en uppfattning om detta, utvecklar vi i denna avhandling teori och tekniker för att designa kommunikations-effektiva optimeringsalgoritmer som endast använder information med låg precision. I den första delen analyserar vi konvergensen hos optimeringsalgoritmer med direkt kompression. Först ger vi en översikt av kompressionstekniker som täcker in många kompressorer av praktiskt intresse. Sedan presenterar vi ett enhetligt analysramverk för optimeringsalgoritmer med kompressorer, som kan vara antingen deterministiska eller randomiserade. I synnerhet visas val av parametrar i komprimerade optimeringsalgoritmer som avgörs av kompressorns parametrar som garanterar konvergens. Våra konvergensgarantier visar beroende av kompressorns noggrannhet och fördröjningseffekter på grund av asynkronicitet hos algoritmer. Detta låter oss karakterisera avvägningen mellan iterations- och kommunikations-komplexitet när kompression används. I den andra delen studerarvi hög prestanda hos felkompenseringsmetoder för komprimerade optimeringsalgoritmer. Även om konvergensgarantier med felkompensering har etablerats finns det väldigt begränsat teoretiskt stöd för konkurrenskraftiga konvergensgarantier med felkompensering. Vi utvecklar därför teoretiska förklaringar, som visar att användande av felkompensering garanterar godtyckligt hög lösningsnoggrannhet från komprimerad information. I synnerhet bidrar felkompensering till att ta bort ackumulerade kompressionsfel och förbättrar därmed lösningsnoggrannheten speciellt för illa konditionerade kvadratiska optimeringsproblem. Vi presenterar också stark konvergensanalys för felkompensering tillämpat på stokastiska gradientmetoder med ett kommunikationsnätverk innehållande ett flertal maskiner. De felkompenserade algoritmerna resulterar, i motsats till direkt kompression, i betydande reducering av kompressionsfelet. Simuleringar av algoritmer i denna avhandling på verkligaproblem med referensdatamängder validerar våra teoretiska resultat. / <p>QC20191120</p>