Jung-Su Kim

Design of a supervisory control system for multiple robotic systems

This paper presents a design experience of a supervisory control system for coordination of multiple robotic devices. To effectively program job commands, a Petri net-type graphical robot language(PGRL) is proposed, where some functions for coordination among tasks such as concurrency and synchronization, can be easily programmed. Each task of PGRL is described by employing formal model languages, which are composed of three modules, sensory, data handling, and action module. It is expected that by using our proposed PGRL and formal model languages, one can efficiently describe a job or task, and hence can easily operate a complex real-time concurrent system. The proposed control system has been implemented by using VME-based 32-bit microprocessor boards and a real-time multitasking operating system (VxWorks), and is shown to successfully work for robotic jobs.

Output Regulation Problem and Solution for LTV Minimum Phase Systems with Time-varying Exosystem

Output regulation problem is to design a dynamic feedback controller that rejects/follows the unknown disturbance/reference asymptotically assuming that the disturbance/reference is generated by a known model (called exosystem). Although the model is known, the disturbance/reference is still unknown because the initial condition of the exosystem is unknown. It has been studied very actively since the seminal work by Francis in 1977, and up to now, we have quite complete solution in the conventional sense. However, most results assume that the exosystem is time-invariant. Although nonlinear exosystem has also been considered, the time-varying exosystem has not been incorporated because of the Poisson stability required to the nonlinear exosystem. In this paper, we extend output regulation theory into the case of time-varying linear systems with time-varying linear exosystems. This is enabled by considering differential regulator equation rather than the classical (algebraic) regulator equation. While the differential regulator equation was also employed by Zhang and Serrani in 2005, who solved the output regulation problem with the periodic exosystem, we use it for deriving an explicit solution of the class of minimum-phase systems having general time-varying exosystem. As an application, we solve the track following problem for an optical disc drive system whose spindle motor is accelerating or decelerating. This case leads to the disturbance having time-varying frequency, which turns into the time-varying exosystem. It seems that this result can be applied to robotics/mechatronic systems control

Region-based Q-learning for intelligent robot systems

It is desirable for autonomous robot systems to possess the ability to behave in a smooth and continuous fashion when interacting with an unknown environment. Since Q-learning requires a lot of memory and time to optimize a series of actions in a continuous state space, it may not be easy to apply the method to such a real environment. In this paper, for a continuous state space application, we propose a new method of Q-learning that incorporates a region-based reward assignment being used to solve a structural credit assignment problem and a triangular type Q-value model. Our learning method can estimate a current Q-value by a relationship with its neighboring states and has the ability to learn its actions similar to that of Q-learning. Thus, our method can enable robots to move smoothly in a real environment. To show the validity of our method, navigation comparison with Q-learning is given and visual tracking simulation results involving a 2-DOF SCARA robot are also presented.

Probing the curvature of a convex object via discrete active touch

In this paper we propose a method of probing the local curvature of a three-dimensional (3D) convex object using the force and torque information which is obtained from the discrete active touch with a robot finger. A scheme for estimating two-dimensional (2D) curvature of a convex object is introduced and the way of extracting 3D curvatures from 2D curvatures is presented. To verify the effectiveness of the proposed method, numerical simulations have been performed and experiments have also been carried out with an experimental system consisting of a finger.

Synchronization of Linear Time-Varying Multi-Agent Systems with Heterogeneous Time-Varying Disturbances Using Integral Controller

This paper presents synchronization of LTV (Linear Time-Varying) MAS (Multi-Agent Systems) with heterogeneous time-varying disturbances under a fixed, connected, and undirected communication network. All the agents can collect only relative state information from their neighborhoods. To achieve synchronization of the MAS, an integral control scheme is proposed based on relative state information between agents.

Disturbance Observer based Boundary Tracking for Environment Monitoring

This paper presents a boundary tracking control of an agent. To this end, it is shown that the boundary tracking problem can be reformulated into a robust control of uncertain double integrator first. Then, a disturbance observer (DOB) based control is proposed solving the robust control problem. Unlike the existing results in the literature, the proposed DOB based control requires only the local position measurement of the boundary (not the gradient information). The performance of the proposed control is demonstrated for two cases: the measurement of the boundary is given in a continuous or discrete manner. Finally, it is shown that the proposed control can be used for environmental monitoring as well by showing that the agent follows a level curve of real environmental monitoring data.

Design of Q-filters for Disturbance Observers via BMI Approach

The type-k disturbance observer, also known as high order disturbance observer, is a robust output feedback controller which can reject polynomial-in-time disturbances exactly and at the same time unmodeled disturbances approximately. To ensure the robust stability of the closed-loop system, the key component of the controller called Q-filter should be designed appropriately. In this paper, we formulate the design problem in terms of a set of bilinear matrix inequalities so that the coefficients of Q-filters are chosen simultaneously. The solution from the proposed method can be less conservative compared to the recursive design procedure developed recently.

Practical distributed coordinated tracking of second order heterogeneous linear agents: Relative measurement feedback

This paper considers the coordinated tracking problem for second order linear agents. Each agent is to follow the trajectory of a virtual leader which is a double integrator. It is assumed that the agents may have different dynamics from each other and may have external disturbances. Moreover, we consider the case where the systems are allowed to exchange only outputs rather than all states. A dynamic output feedback controller is presented to guarantee that the trajectories of agents practically synchronize to that of virtual leader.