This body of work presents methods for the optimization, analysis, and control of mixed serial-parallel structures known as SP-Stacks. A SP-Stack is a series of Stewart Platforms (SPs) linked via their top and bottom plates to create a serial chain of parallel mechanisms. SP-Stacks are unique in their bridging of the benefits of parallel architectures (high rigidity, strength, and precision) and serial architectures (reach and manipulability), at the cost of being extremely overactuated. SP-Stacks are also difficult to provide kinematic solutions for, as neither forward nor inverse kinematics of a system are closed form. The first work presented focuses on presenting algorithms and optimization functions pertaining to the kinematic configuration of a SP-Stack. It first presents two methods of fast inverse kinematics (IK) for the SP-Stack which do not take forces into account. The outputs of those more simplistic solvers as used as initial conditions for a Nonlinear Program (NLP) algorithm which optimizes the internal configuration of a SP-Stack such that the end effector (EE) plate remains at the desired location, and the maximal force experienced on any actuator is minimized. The second work presented focuses on hardware testing some of the constituent algorithms and conclusions drawn from the first paper and determining methods of compensating for, in software, detected defects in hardware and hardware measurement systems. This work also demonstrates a different form of force-optimization - compliance control (CC), which is executed on both a single SP responding to external forces, and a 2 SP-Stack responding to regular internal perturbation. Conclusions drawn from these works are useful for stacks of an arbitrary number of SPs, can be extended to other mixed-kinematic systems, and advance the capabilities of these systems to be useful contributors in field robotics. / Master of Science / A stewart platform (SP) is a type of robot which consists of two plates interconnected by six linear actuators in parallel, which allow the robot to either translate or rotate about any axis in space. SPs are limited in their ability to move, as their parallel construction limits their workspace. In order to counteract this, SPs can be stacked on top of one another, creating a SP-Stack. The SP-Stack is capable of using its status as a mixed serial-parallel system to move in a significantly larger area (an advantage derived by the serial component of its architecture) and retain extraordinary rigidity and strength (an advantage from its parallel architecture). As each SP has 6 Degrees of Freedom (6DoF), enabling the previously described free-space motion, a SP-Stack possesses 6n DoF, making it overactuated. An overactuated system has multiple internal configurations which allow for a desired end effector configuration. The body of work presented herein focuses on manipulating the overactuation of SP-Stacks to achieve desirable results such as finding configurations which are most resistant to external loading (optimization of actuator forces) or algorithms which allow SP-Stacks to comply with external loading (compliance control (CC)). The first work presented herein focuses on determining an optimal configuration for a 4 SP-Stack such that the maximum force experienced by any one of its linear actuators is minimized, given a known external force. This work also presents two methods of generating initial configurations for the SP-Stack which are fed into the optimization algorithm which produces the final solution, as well as providing details on the constraints which govern the movement and validity of configurations for the system. The second work presented expands on the work done in the first, moving into hardware testing for verification of algorithms which calculate forces experienced by the linear actuators. The hardware testing showcases some errors that can be introduced by low fidelity hardware, along with methodologies for counteracting those errors. Finally, the second work introduces CC, the ability for a robot to move itself to adapt to incoming forces, and applies it to a physical 2 SP-Stack as a demonstrator.

Inspiration from biological systems suggests that robots should demonstrate same level of capabilities that are embedded in biological systems in performing safe and successful interaction with the humans. The major challenge in physical human robot interaction tasks in anthropic environment is the safe sharing of robot work space such that robot will not cause harm or injury to the human under any operating condition. Embedding human like adaptable compliance characteristics into robot manipulators can provide safe physical human robot interaction in constrained motion tasks. In robotics, this property can be achieved by using active, passive and semi active compliant actuation devices. Traditional methods of active and passive compliance lead to complex control systems and complex mechanical design. In this thesis we present compliant robot manipulator system with semi active compliant device having magneto rheological fluid based actuation mechanism. Human like adaptable compliance is achieved by controlling the properties of the magneto rheological fluid inside joint actuator. This method offers high operational accuracy, intrinsic safety and high absorption to impacts. Safety is assured by mechanism design rather than by conventional approach based on advance control. Control schemes for implementing adaptable compliance are implemented in parallel with the robot motion control that brings much simple interaction control strategy compared to other methods. Here we address two main issues: human robot collision safety and robot motion performance.We present existing human robot collision safety standards and evaluate the proposed actuation mechanism on the basis of static and dynamic collision tests. Static collision safety analysis is based on Yamada’s safety criterion and the adaptable compliance control scheme keeps the robot in the safe region of operation. For the dynamic collision safety analysis, Yamada’s impact force criterion and head injury criterion are employed. Experimental results validate the effectiveness of our solution. In addition, the results with head injury criterion showed the need to investigate human bio-mechanics in more details in order to acquire adequate knowledge for estimating the injury severity index for robots interacting with humans. We analyzed the robot motion performance in several physical human robot interaction tasks. Three interaction scenarios are studied to simulate human robot physical contact in direct and inadvertent contact situations. Respective control disciplines for the joint actuators are designed and implemented with much simplified adaptable compliance control scheme. The series of experimental tests in direct and inadvertent contact situations validate our solution of implementing human like adaptable compliance during robot motion and prove the safe interaction with humans in anthropic domains.

Les travaux présentés dans cette thèse portent sur la commande en couple de manipulateurs redondants.Nous nous intéressons dans ce cadre à deux problématiques. En premier lieu, nous considérons le cas d’imposition d’une contrainte cinématique de point de passage, dite contrainte du RCM, de l’organe terminal (OT) du robot. Nous proposons alors deux approches pour la gestion de cette contrainte. Dans la première approche, la contrainte est garantie dans l’espace nul d’une tâche principale définie en coordonnées de position de l’OT. Cette méthode exploite une définition explicite de la dynamique de l’espace nul et confère un niveau de priorité secondaire à la contrainte. La seconde approche permet de définir la contrainte du RCM comme tâche principale, en lui assignant le niveau de priorité supérieur ou un niveau de priorité défini par le besoin de l’application. Nous proposons pour cela une nouvelle définition de la cinématique du RCM.En second lieu, nous traitons la question des contacts entre le corps du robot et son environnement (ex. l’humain)pendant que l’OT exécute sa tâche « globale ». Nous proposons pour cela une stratégie de compliance appliquée dans l’espace nul du robot afin de préserver la tâche globale lors des contacts. Cette stratégie estdéfinie pour des bras anthropomorphes à 7-DDL, et est formulée en coordonnées de l’angle de bras, paramètre représentant le degré de redondance du robot. Cela permet de définir un intervalle admissible de mouvement de l’angle de bras. Lorsque les limites de cet intervalle sont atteintes, une loi de compliance de type ressort amortisseur oblige le robot à rester dans l’intervalle, malgré les forces externes exercées.Nous évoquons, tout au long de cette thèse, l’application de chirurgie mini-invasive assistée par robot pour illustrer l’utilité de nos contributions. / The work developped in this PhD thesis concerns the control of redundant torque-controlled robots,dealing with two main issues. Firstly, we study the presence of a RCM constraint imposed to the end-effector. We propose two control approaches to guarantee this kinematic constraint. In the first one, the constraint is performed in the null-space of a main task defined in cartesian coordinates(position). An explicit definition of the null-space dynamics is applied on this control approach, and provides a secondary priority order to the RCM constraint. The second approach allows to define the constraint as the main task, obtaining the highest priority level, or in any desired priority level,according to the needs of the application. Therefore, we propose a new kinematic formulation of the RCM constraint.Secondly, we study the physical interaction between the robot’s body and its environment (e.g. human) during the cartesian global task execution. A null-space compliance control strategy is then proposed in order to preserve the global task when the contacts occur. This strategy, defined for anthropomorphic 7-DOF robots, is formulated in swivel angle coordinates, which is a direct representation of the robot’s null-space. A desired feasible range for the swivel angle values is defined by the user, and a spring-damping compliance law is used to constraint the robot to remain within the feasible angle values range, despite the external forces applied to the robot’s body. Robot-assisted minimally invasive surgery has been used throughout this thesis as an example of application, allowing to demonstrate the usefulness of our contributions.

Researchers have offered a variety of solutions for overcoming the old and challenging problem of undesired vibrations. The optimum vibration-control solution that can be a passive, semi-active or active solution, is chosen based on the desired level of vibration-control, the budget and the nature of the vibration source. Mechanical vibration-control systems, which work based on variable stiffness control, are categorized as semi-active solutions. They are advantageous for applications with multiple excitation frequencies, such as seismic applications. The available mechanical variable stiffness systems that are used for vibration-control, however, are slow and usually big, and their slowness and size have limited their application. A new semi-active variable stiffness solution is introduced and developed in this thesis to address these challenges by providing a faster vibration-control system with a feasible size. The new solution proposed in this thesis is a semi-active variable stiffness mount/isolator called the antagonistic Variable Stiffness Mount (VSM), which uses a variable stiffness spring called the Antagonistic Variable stiffness Spring (AVS). The AVS is a kinematically singular prestressable mechanism. Its stiffness can be changed by controlling the prestress of the mechanisms links. The AVS provides additional stiffness for a VSM when such stiffness is needed and remains inactive when it is not needed. The damping of the VSM is constant and an additional constant stiffness in the VSM supports the deadweight. Two cable-mechanisms - kinematically singular cable-driven mechanisms and Prism Tensegrities - are developed as AVSs in this thesis. Their optimal configurations are identified and a general formulation for their prestress stiffness is provided by using the notion of infinitesimal mechanism. The feasibility and practicality of the AVS and VSM are demonstrated through a case study of a typical engine mount by simulation of the mathematical models and by extensive experimental analysis. A VSM with an adjustable design, a piezo-actuation mechanism and a simple on-off controller is fabricated and tested for performance evaluation. The performance is measured based on four criteria: (1) how much the VSM controls the displacement near the resonance, (2) how well the VSM isolates the vibration at high frequencies, (3) how well the VSM controls the motion caused by shock, and (4) how fast the VSM reacts to control the vibration. For this evaluation, first the stiffness of the VSM was characterized through static and dynamic tests. Then performance of the VSM was evaluated and compared with an equivalent passive mount in two main areas of transmissibility and shock absorption. The response time of the VSM is also measured in a realistic scenario.

No description available.