Whether it is a car, a mobile phone, or a computer, we are noticing how automation and production with robots plays an important role in the industry of our modern world. We find it in factories, manufacturing products, automotive cruise control, construction equipment, autopilot on airplanes, and countless other industrial applications.         Automation technology can vary greatly depending on the field of application. On one end, we have systems that are operated by the user and rely fully on human ability. Examples of these are heavy-mobile equipment, remote controlled systems, helicopters, and many more. On the other end, we have autonomous systems that are able to make algorithmic decisions independently of the user.         Society has always envisioned robots with the full capabilities of humans. However, we should envision applications that will help us increase productivity and improve our quality of life through human-robot collaboration. The questions we should be asking are: “What tasks should be automated?'', and “How can we combine the best of both humans and automation?”. This thinking leads to the idea of developing systems with some level of autonomy, where the intelligence is shared between the user and the system. Reasonably, the computerized intelligence and decision making would be designed according to mathematical algorithms and control rules.         This thesis considers these topics and shows the importance of fundamental mathematics and control design to develop automated systems that can execute desired tasks. All of this work is based on some of the most modern concepts in the subjects of robotics and control, which are synthesized by a method known as the Virtual Holonomic Constraints Approach. This method has been useful to tackle some of the most complex problems of nonlinear control, and has enabled the possibility to approach challenging academic and industrial problems. This thesis shows concepts of system modeling, control design, motion analysis, motion planning, and many other interesting subjects, which can be treated effectively through analytical methods. The use of mathematical approaches allows performing computer simulations that also lead to direct practical implementations.

Model-based control strategies for robot manipulators can present numerous performance advantages when an accurate model of the system dynamics is available. In practice, obtaining such a model is a challenging task which involves modeling such physical processes as friction, which may not be well understood and difficult to model. Furthermore, uncertainties in the physical parameters of a system may be introduced from significant discrepancies between the manufacturer data and the actual system. Traditionally, adaptive and robust control strategies have been developed to deal with parametric uncertainty in the dynamic model, but often require knowledge of the structure of the dynamics. Recent approaches to model-based manipulator control involve data-driven learning of the inverse dynamics relationship, eliminating the need for any a-priori knowledge of the system model. Locally Weighted Projection Regression (LWPR) has been proposed for learning the inverse dynamics function of a manipulator. Due to its use of simple local, linear models, LWPR is suitable for online and incremental learning. Although global regression techniques such as Gaussian Process Regression (GPR) have been shown to outperform LWPR in terms of accuracy, due to its heavy computational requirements, GPR has been applied mainly to offline learning of inverse dynamics. More recent efforts in making GPR computationally tractable for real-time control have resulted in several approximations which operate on a select subset, or sparse representation of the entire training data set. Despite the significant advancements that have been made in the area of learning control, there has not been much work in recent years to evaluate these newer regression techniques against traditional model-based control strategies such as adaptive control. Hence, the first portion of this thesis provides a comparison between a fixed model-based control strategy, an adaptive controller and the LWPR-based learning controller. Simulations are carried out in order to evaluate the position and orientation tracking performance of each controller under varied end effector loading, velocities and inaccuracies in the known dynamic parameters. Both the adaptive controller and LWPR controller are shown to have comparable performance in the presence of parametric uncertainty. However, it is shown that the learning controller is unable to generalize well outside of the regions in which it has been trained. Hence, achieving good performance requires significant amounts of training in the anticipated region of operation. In addition to poor generalization performance, most learning controllers commence learning entirely from `scratch,' making no use of any a-priori knowledge which may be available from the well-known rigid body dynamics (RBD) formulation. The second portion of this thesis develops two techniques for online, incremental learning algorithms which incorporate prior knowledge to improve generalization performance. First, prior knowledge is incorporated into the LWPR framework by initializing the local linear models with a first order approximation of the prior information. Second, prior knowledge is incorporated into the mean function of Sparse Online Gaussian Processes (SOGP) and Sparse Pseudo-input Gaussian Processes (SPGP), and a modified version of the algorithm is proposed to allow for online, incremental updates. It is shown that the proposed approaches allow the system to operate well even without any initial training data, and further performance improvement can be achieved with additional online training. Furthermore, it is also shown that even partial knowledge of the system dynamics, for example, only the gravity loading vector, can be used effectively to initialize the learning.

Typescript (Photocopy) / Includes paper co-authored by the author as attachment / Bibliography: leaves 133-140 / vii, 140, [12] leaves : ill ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Thesis (Ph.D.)--University of Adelaide, Dept. of Applied Mathematics, 1988

Thesis (Ph.D.(Mechanical Engineering))--University of Pretoria, 2003. / Summaries in Afrikaans and English. Includes bibliographical references.

Thesis (M.Eng.(Mechanical Engineering))--University of Pretoria, 1999. / Summaries in Afrikaans and English. Includes bibliographical references.

Thesis (M. Eng.)--Carleton University, 1994. / Also available in electronic format on the Internet.

Thesis (M.S.)--Mississippi State University. Department of Electrical and Computer Engineering. / Title from title screen. Includes bibliographical references.

Ο στόχος της εργασίας αυτής είναι η κατασκευή και ο έλεγχος ενός λαπαροσκοπικού εργαλείου με πλεονάζοντες βαθμούς ελευθερίας (ΒΕ). Ο βραχίονας αποτελείται από συνδεδεμένα τμήματα ή «σπονδύλους» που ενεργοποιούνται από καλώδια Ni-Ti, γνωστά και ως Shape Memory Alloys, που λειτουργούν ως δυαδικοί ενεργοποιητές με 2 καταστάσεις. Έτσι, εξασφαλίζεται η επαναληψιμότητα στις κινήσεις του βραχίονα και δεν απαιτούνται αισθητήρες για τη θέση των αρθρώσεων. Κάθε σπόνδυλος αποτελείται από 3 ενεργούς πρισματικούς ενεργοποιητές (SMA) τοποθετημένους έτσι ώστε να σχηματίζουν ένα τρίποδο, το οποίο παρέχει 3 ΒΕ σε κάθε άρθρωση (δύο περιστροφικούς και ένα μεταφορικό). Η ανεξάρτητη ενεργοποίηση κάθε άρθρωσης γίνεται από μικροελεγκτές που επικοινωνούν με το πρωτόκολλο I2C και είναι τοποθετημένοι στο εσωτερικό κάθε συνδέσμου. Θα παρουσιαστούν διάφορα ζητήματα σχεδιασμού, η κινηματική και η προσομοίωση του εργαλείου, καθώς και ένα αρχικό πείραμα. / The subject of this work is the development of a prototype hyper-redundant laparoscopic tool. The manipulator consists of cascaded modules which are powered by Shape Memory Alloy wires (NiTi), acting as binary actuators with two stable states. As a result, the repeatability of the manipulator’s movement is ensured, alleviating the need for sensing of the manipulator’s joint-positions. Each module is composed of three active prismatic actuators in a tripod configuration providing a 3-DOF (two rotational and one translational) maneuverability for each joint. I2C-networked microcontrollers activate the individual tendon in each joint. Certain design aspects as well as the kinematics of the binary manipulator are presented followed by simulation and experimental studies on the laparoscopic tool prototype.

CAPES; CNPq / Neste trabalho é apresentado a modelagem e controle de um manipulador robótico com características flexíveis tanto nas juntas como nos elos. A modelagem matemática é obtida através da formulação de Lagrange. Na estratégia de controle são considerados como atuadores os motores CC (Corrente Contínua) e atuadores classificados como materiais inteligentes como a Liga de Memória de Forma (LMF) e freio Magneto-Reológico (MR-Break). O controle SDRE (State Dependent Ricatti Equation) é utilizado como a lei de controle com base nas técnicas de controle ótimo para sistemas dinâmicos não lineares. Para demonstrar a eficiência e estabilidade do controle são apresentadas simulações numéricas considerando duas configurações distintas: a primeira demonstra o controle do manipulador robótico com dois elos rígidos, duas juntas flexíveis e utilizando controle de torque no elo do manipulador através de atuador MR. A segunda configuração trata do controle de um sistema com três elos, sendo um deles flexível, dois rígido e duas das juntas flexíveis, utilizam-se como atuadores os motores CC dos elos e um atuador LMF acoplado na viga flexível. Com o objetivo de validação o controle de posição através de motor CC e a influencia do eixo flexível na dinâmica do sistema um protótipo de manipulador com dois elos rígidos e duas juntas flexíveis foi confeccionado. Resultados numéricos e experimentais demonstraram a eficiência do controle de posicionamento dos elos através do controle SDRE e a influencia do eixo flexível na dinâmica do manipulador robótico. / This work presents the modeling and control of robotic manipulators with flexible characteristics, both joints as in the links. The mechanical model is obtained by Lagrangian formulation. Control strategy is considered as actuators DC motors (Direct Current) and actuators classified as smart materials such as shape memory alloy (SMA) and Magnetorheological brake (MR). The law of implemented control is based on optimal control techniques for dynamic systems, this control is given the name of SDRE control (State Dependent Ricatti Equation). To demonstrate the efficiency and control the stability of the numerical simulations are presented in three systems: the first demonstrates control of a robotic manipulator with two hard links, two flexible joints and adding torque control in the manipulator of the link through actuator MR. The second case concerns the control of a system with three links, one being flexible, rigid and two of two flexible joints, is used as actuators DC motors of the links and a coupled SMA actuator to the flexible beam. Aiming to validating the control position by DC motor and the influence of the flexible shaft in system dynamics a manipulator prototype with two rigid links and two flexible joints was made. Numerical and experimental results demonstrated the links positioning control efficiency through SDRE control and the influence of the flexible shaft in the dynamics of robotic manipulator.

Orientador: Germano Francisco Biondi / Banca: Oswaldo Durival Rossi Junior / Banca: José Paes de Almeida Nogueira Pinto / Resumo: O presente trabalho teve como objetivo constatar "in loco" a situação estrutural e higiênico sanitária de 44 estabelecimentos alimentícios (bares e lanchonetes, mini-mercados, padarias, restaurantes) da cidade de Botucatu, levantando, através dos dados obtidos, quais são os setores que oferecem maior risco à saúde pública. Foi aplicado um questionário chamado de diagnóstico tecnológico em cada um deles, o qual continha 144 questões divididas em cinco itens, sendo eles: Item 1 - Edificação e Instalações, Item 2 - Equipamentos, Móveis e Utensílios, Item 3 - Funcionários, Item 4 - Fluxo de Produção, Item 5 - Sistema da Garantia da Qualidade. Os dados obtidos permitiram que se cumprisse o objetivo de responder as seguintes questões: a) Qual setor (Bares e Lanchonetes, Mini-Mercados, Restaurantes e Padarias) obteve pior desempenho comparativamente aos outros em cada questão? b)Qual dos itens obteve o pior resultado geral? c) Qual o setor que apresentou maior número de não conformidades? d) Qual item é mais problemático em cada setor? Os dados analisados demonstraram que: houve diferença estatística (p<0,05) entre o desempenho de cada setor por questão, em 14 oportunidades, sendo o Setor Padarias o de pior resultado; quanto aos itens, o de pior desempenho geral foi o Item 5 - Sistema da Garantia da Qualidade, com 61% de não conformidade; o setor que apresentou maior número de não conformidades foi o Setor Padarias, com 48 em 144 possibilidades (33%); destacaram-se negativamente perante os itens: o Setor Padarias no Item 1, o Setor Restaurantes no Item 2, o setor Mini-Mercados no Item 3 e o Setor Bares e Lanchonetes nos Itens 4 e 5. / Abstract: The present work had as objective to evidence "in loco" the structural situation and hygienic sanitary of 44 feed establishments (bar and snack bars, mini-markets, bakeries, restaurants) of the city of Botucatu, raising, through the obtained data, which are the sectors that offer to greater risk to public health. We administered a questionnaire called the diagnostic technology in each, which contained 144 questions divided into five items, which were: Item 1 - Building and Facilities, Item 2 - Equipment, Furniture and instrument, Item 3 - Staff, Item 4 - Flow Production, Item 5 - System of Quality Assurance. The data obtained have to satisfy the goal of answering the following questions: a) What sector (Bars and Snack Bars, Mini Markets, Restaurants and Bakeries) had worse performance compared to others in every question? b) Which of the items received the worst overall outcome? c) What is the sector that presented greater number of non-conformities? d) What item is most problematic in each sector? The data analyzed showed that there was statistical difference (p<0,05) between the performance of each sector by question, in 14 opportunities, and the Bakeries Sector in the worst results; while for items, the overall performance was worse in the Item 5 - System of Quality Assurance, with 61% of non-compliance; the sector that presented greater number of noncompliance was the Bakeries Sector, with 48 in 144 chances (33%); stood out negatively to the items: the Bakeries Sector in Item 1, the Restaurant Sector in Item 2, the Mini Markets Sector in Item 3 and Bars and Snack Bars Sector in the Items 4 and 5. / Mestre