CAREER: Planning and Control for Overconstrained Mechanisms

点击翻译按钮获取中文摘要

AbstractMultiple point contact is common in many manipulation tasks. Examplesinclude arrays for distributed or micro manipulation, vehicles, andtraditional robotic grasping. Although all of these applications havereceived a great deal of attention, no previous works have provided ananalytical approach capable of dealing with the inherent uncertainties inthe contact state--the state that represents whether a given contact issticking, slipping, or is out of contact. These nonsmooth transitionsbetween contact states have a dramatic impact on the dynamics; hence, itis important to systematically mitigate the negative performance thesenonsmooth effects induce. Moreover, these systems are often overactuated,so that each contact interface is independently articulated. This leadsto mechanisms that are nominally kinematically overconstrained--that is,the kinematic relationships between all the point contacts cannot besimultaneously satisfied. Which constraint is broken is sensitive todetails of friction and normal force modeling, so it is necessary toestimate the current contact state and incorporate the contact state intothe motion planning and control. The goal of this research is toproduce hybrid estimators, motion planning algorithms, and controlstrategies that will work in concert to guarantee performance andstability in the face of multiple contact interfaces in an unstructuredenvironment. An overconstrained multiple point manipulator prototype willbe developed and planning and control methods will be applied to thissystem to demonstrate manipulation that is not sensitive to theparticulars of the frictional interfaces.Manufacturing often involves the need to reposition and reorient objectsfor purposes of assembly. To accomplish this, multiple actuators areoften used, and these actuators experience stick and slip contact with theobject due to frictional interactions. The physics of the actuators and,in particular, their interaction with the object are notoriously difficultto model accurately. Hence, there is a strong need for manipulationstrategies that are not sensitive to these low-level details. Moreover,many vehicles, such as the original Mars rover, have a mechanical designthat guarantees that some of the wheels must slip during operation.However, which wheels slip is dependent on unknown environmental conditions. Hence, in this situation as well there is a need to develop motion planning strategies that are guaranteed to work even in the presence of substantial uncertainty arising from the environment. This project will contribute to these needs by developingstrategies that have guaranteed performance in the face of inherentuncertainty. In the short term this project will contribute to macro-scalemanufacturing and vehicle control, and in the long-term will likely impactmicro-scale manufacturing.