Font Size: a A A

Research On Hamilton For Robot Binocular Visual Servoing Control And The Observer Desinging

Posted on:2013-07-01Degree:DoctorType:Dissertation
Country:ChinaCandidate:B YangFull Text:PDF
GTID:1228330392454777Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
The performance of the robot visual servo is an important aspect on the intelligencelevel of the autonomous robot. For the robot visual servo system, the controller designingwill become complex and the stability analysis will become difficult, due to the existenceof the uncertainty, the disturbance and the unmeasurable state in the robotic system.Meanwhile, generalized Hamiltonian systems, as an important class of nonlinear systems,owing to its clear structure and physical meaning, and that the Hamiltonian function isconsidered as the total energy of the systems, demonstrates salient superiority in stabilityanalysis and control design. The visual servo control via Hamiltonian function method forcamera-in-hand robotic systems and the observer design is considered. The uncertaintyand the disturbance be taken into consideration two classes of generalized Hamiltonianrealization is proposed which for the binocular stereo visual servoing system respectively.Based on the Immersion and invariance (I&I) technique, a speed observer is design forrobotics system and visual servoing system, respectively. The detail works are as follows:Firstly, an visual servo control via Hamiltonian function method for camera-in-handbinocular robotic systems is considered. We introduce a binocular stereo vision model.The concept of “imaging sensitivity” is proposed. It is to compare of the traditionalmonocular model and the binocular model. Based the compare, we can find the binocularmodel can overcome the depth of the visual blind spot of the monocular model. For thebinocular vision visual servo system, generalized Hamiltonian realization is given whichconsider the uncertainty is or isn’t exist in robotics system, respectively. We find aninternal proposition through analysis the internal structure of robotics system. Theuncertainty visual servo robotic system is reformulated as a generalized Hamiltoniansystem based the internal proposition and pre-state feedback control. Meanwhile, based onHamiltonian function method, a visual servo controller is constructed. For the generalizedHamiltonian system, the stability analysis is given.Secondly, for a class of disturbance visual servo robotic system, a visual servocontroller will be designed based Hamiltonian function method. The disturbance visual servo robotic system is reformulated as a generalized Hamiltonian system via afactorization of the mass matrix. For the generalized Hamiltonian system, an L2disturbance attenuation control is proposed, and we design a visual servo controller. Then,the asymptotically stable analysis of the closed-loop system is given.Thirdly, for a class of robotic system is not equipped with the angular velocitysensors, a speed observer is design based on the Immersion and invariance (I&I) technique.Based this technique, we design a target dynamical system and a manifold. A Partialdifferential equation (PDE) is solved to make the manifold is attractive and invariant. Wewill design a full-order state observer to estimate unknown states. The dynamics of theobserver error is obtained. For the dynamics system a generalized Hamiltonian realizationis given, and ensure that the closed-loop system is asymptotically stable.Finally, we design a reduced-order speed observer and a visual servo controller for abinocular visual servo system based on the Immersion and invariance (I&I) technique andHamiltonian method. A generalized Hamiltonian realization is given that the observererror system is transformed into a port-controlled Hamiltonian (PCH) system. For thePCH systems that the joint speeds of a robot been observed, an image-based controller isdesigned to satisfy the visual servoing performance. The stabilization analysis of theclosed-loop system is given.
Keywords/Search Tags:robot, generalized hamiltonian systems, visual servoing, immersion andinvariance, observer, binocular stereo vision model, L2disturbanceattenuation
PDF Full Text Request
Related items