Joono Cheong

Motion synchronization in virtual environments with shared haptics and large time delays

This paper deals with a synchronization scheme to achieve a high level of consistency in peer-to-peer based virtual environments (VEs) for shared haptics with large time delays. We mainly focus on the VEs with dynamic objects that show physically realistic behaviors. The synchronization scheme utilizes an advanced feedback controller to compensate for the state error between geographically separated sites. Using feedback control theory of time delay systems, the controller is shown to result in closed loop stability and robust to perturbations in the time delay. The maximum allowable delay is determined using stability analysis, which indicates the allowable bound on the communication delay for the system.

PID composite controller and its tuning for flexible link robots

This paper proposes a PID-type composite controller for controlling flexible arms modeled by the singular perturbation approach, and investigates a tuning method based on the proposed controller structure. For the slow sub-controller, a PD plus disturbance observer is used, which eventually takes on PID form, and for the fast sub-controller, modal feedback PID control is utilized. The effects of design parameters of the controller on the closed loop response are investigated. Through simulation and experiments, the adequacy and performance of the proposed method are verified.

Inverse kinematics of multilink flexible robots for high-speed applications

A straightforward inverse kinematic algorithm for multilink flexible robots is proposed to improve the control performance. The inclusion of a dynamic constraint maximizes the performance of feedback controllers in high-speed applications. To obtain a numerically feasible solution, the singular perturbation approach is employed, which decomposes the inverse kinematics into an averaged part (slow part) and a parasitic part (fast part). The solution of the averaged part is considered the desired inverse kinematics, while the parasitic part is intentionally removed. The parameter expansion is carried out to obtain the solution sequentially. The implicit expansion method, which is a refined version of the expansion method, reduces computing time considerably. The formula in discrete time offers efficiency in computer applications. In addition, a requirement on differentiability of the desired task trajectory is derived.

Bandwidth modulation of rigid subsystem for the class of flexible robots

Focuses on the design of rigid part motion control and the selection of bandwidth of the rigid subsystem. Based on the passivity approach and disturbance observer, we investigate the relationship between macro joint tracking performance and vibration suppression capability using the joint motion bandwidth parameter. For the flexibility isolated system, the independent sub-controllers for the rigid and flexible part are proposed just like the singular perturbation approach. The validity of the proposed method is verified by experiments.

Joint tracking controller for multi-link flexible robot using disturbance observer and parameter adaptation scheme

An improved composite controller of singular perturbation approach is designed for controlling a multi-link flexible robot with uncertainties. We adopt the standard form of a singular perturbation approach for modeling. To reduce the coupling effect from flexibility, the bandwidth of a slow subsystem is modulated by considering the fundamental frequency. The disturbance observer provides a means for defining the bandwidth of a slow subsystem as well as compensating disturbances. At the same time, uncertainties in the fast subsystem are updated to enhance the capability for vibration suppression in conjunction with PID (Proportional-integrative derivative) modal feedback. We draw conditions for Lyapunov stability of the modal feedback and adaptive scheme. A numerical simulation will support the validity of our research results.

Synchronization control for physics-based collaborative virtual environments with shared haptics

In this paper, we propose a synchronization scheme to achieve a high level of consistency in peer-to-peer-based shared virtual environments (SVEs), as well as to display natural and realistic motions of virtual objects, in collaborative haptic tasks. The synchronization scheme utilizes an advanced feedback controller to compensate for the state error between geographically separated sites with a significant amount of time delay. It is designed using the mathematical model of a two-user SVE manipulating a freely moving object represented as a mass with damping resistance, with a haptic interface. Thanks to feedback control theory of time delay systems, the controller is shown to result in closed-loop stability and is be robust to perturbations in the time delay. Together with the synchronization control, a recovery filter is also designed and integrated so as to preserve the natural behavior of the synchronized object, which is, otherwise, affected by the feedback control action. In addition to verifying the theoretical results, two experiments using real Internet and local area network communications are carried out. These tests clearly support the validity of the analyses and demonstrate the applicability of the synchronization scheme.

Inverse kinematics of multi-link flexible robots for high speed applications

A straightforward inverse kinematic algorithm for multilink flexible robots is proposed to improve the control per- formance. The inclusion of a dynamic constraint maximizes the performance of feedback controllers in high-speed applications. To obtain a numerically feasible solution, the singular perturbation approach is employed, which decomposes the inverse kinematics into an averaged part (slow part) and a parasitic part (fast part). The solution of the averaged part is considered the desired inverse kinematics, while the parasitic part is intentionally removed. The parameter expansion is carried out to obtain the solution sequen- tially. The implicit expansion method, which is a refined version of the expansion method, reduces computing time considerably. The formula in discrete time offers efficiency in computer applications. In addition, a requirement on differentiability of the desired task trajectory is derived.

Two-Step Controller for 3-D Flexible Link Manipulators: Bandwidth Modulation and Modal Feedback Approach

For joint tracking control of flexible robots, this paper presents a two-step design of controller: the bandwidth modulation with modal feedback approach. First, we focus on the design of rigid parts motion controller considering the bandwidth of the rigid subsystem. We investigate the relationship between macro joint tracking performance and vibration suppression capability using the bandwidth parameter. After adjusting the bandwidth of rigid motion, the composite control, which is the second step consisting of rigid and flexible sub-controllers, is applied like singular perturbation approach. As the flexible sub-controller, we propose a direct modal feedback controller that is very simple, but effective to suppress the vibration. The validity and effectiveness of the proposed method are verified by experiments using a POSTECH 3-D flexible robot.

System mode approach for analysis of horizontal vibration of 3-D two-link flexible manipulators

This paper deals with the system mode analysis of horizontal vibration for 3-D two-link flexible manipulators. For the analysis, we formulate and solve a set of partial differential equations which represent vibration mixed with bending and torsional moment. The inclusion of torsional vibration complicates the analysis, but the results are more precise and realistic. We obtain a number of geometrical and dynamical boundary conditions depending on manipulator configuration. There are two possible boundary conditions at the rotary joint: clamped and pinned with spring condition. We perform a examination and comparison between the two joint conditions. Numerical and experimental tests show the validity and effectiveness of the proposed analysis and modelling.

Accessibility and identifiability of horizontal vibration in 3-D two-link flexible robots: system mode approach

Abstract The accessibility of horizontal vibration in a 3-D two-link flexible robot shows configuration-dependent nature (International Journal of Robotics Research 16 (1997) 567). This paper deals with physical meaning of the accessibility issue in conjunction with system mode approach. The identifiability which is dual to the accessibility is also discussed. The analysis of horizontal vibration based on system mode approach takes an important role in examining the vibration accessibility. The ensuing Lagrangian dynamic formulation enables the formal definition of rigid–flexible coupled dynamic terms which show clear physical meaning. Both theoretical and numerical studies are presented to elucidate the meaning of the accessibility and the identifiability of horizontal vibration. In addition, the experimental results support the theoretical results.