• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.43.6-43
• /
• 2001

In this paper, we describe an actor-critic method as a kind of temporal difference (TD) algorithms. The value function is regarded as a current estimator, in which two value functions have different inputs: one is an actual experience; the other is a simulated experience obtained through a predictive model. Thus, the parameter´s updating for the actor and critic parts is based on actual and simulated experiences, where the critic is constructed by a radial-basis function neural network (RBFNN) and the actor is composed of a kinematic-based controller. As an example application of the present method, a tracking control problem for the position coordinates and azimuth of a nonholonomic mobile robot is considered. The effectiveness is illustrated by a simulation.

• Proceedings of the KIEE Conference
• /
• 2005.07d
• /
• pp.2933-2935
• /
• 2005

The IRPO(Intensive Randomized Policy Optimizer) algorithm is a recently developed tool in the area of reinforcement leaming. And it has been shown to be very successful in several application problems. To compare with a general RL method, IRPO has some difference in that policy utilizes the entire history of agent -environment interaction. The policy is derived from the history directly, not through any kind of a model of the environment. In this paper, we consider a robot-control problem utilizing a IRPO algorithm. We also developed a MATLAH-based animation program, by which the effectiveness of the training algorithms were observed.

• The Journal of Korea Robotics Society
• /
• v.19 no.1
• /
• pp.92-97
• /
• 2024

This paper provides a study of distributional perspective on reinforcement learning for application in mobile robot navigation. Mapless navigation algorithms based on deep reinforcement learning are proven to promising performance and high applicability. The trial-and-error simulations in virtual environments are encouraged to implement autonomous navigation due to expensive real-life interactions. Nevertheless, applying the deep reinforcement learning model in real tasks is challenging due to dissimilar data collection between virtual simulation and the physical world, leading to high-risk manners and high collision rate. In this paper, we present distributional reinforcement learning architecture for mapless navigation of mobile robot that adapt the uncertainty of environmental change. The experimental results indicate the superior performance of distributional soft actor critic compared to conventional methods.

• (The) Research of the performance art and culture
• /
• no.22
• /
• pp.371-411
• /
• 2011

In the twenty first century of rapid cultural change performing art requires new mode of expression based on imaginative power and creativity as well as establishing its own identity. The modern technological environment support this with advanced technology and bring about the expansion of reason from new experience. The introduction of digital media on artistic expression in particular, expands the physical ability of human body which is the main subject of performing art. A virtual body from digital technology is freed from physical boundaries and goes over space and time. It also suggests the possibility of new mode of communication with audience. This study aims at examining the subject of performing art and its digitalized movement focusing on EveR, the world's first professional robot actor. The robot actor which came on stage according to the new expression medium, a digital body, stands in need not only of technological value but also of cultural and artistic application for expression in art. In this endeavor to meet the demand, this study examines the development process and function of 'EveR' the robot actor. Also it searches into the performance of Ever which replaced human being as well as the historical significance of the title:the world's first. To be more specific, there is a example research on two performances:a pansori play "EveR is simply stunning(2009)" and children's play "The Robot Princess and Seven Dwarfs(2009)." Through this example research, it is enabled to anticipate the influence of robot actors on performing arts and to search for the better way of them to evolve. Furthermore, it aims at finding ways to create high value through promoting robot actors to be familiar to the public as well as supporting them to become active cultural contents. The performance with robotic technology is one of the artistic experiment that may cause the change of the future of performing art by actualizing technological imagination together with human body and machinery. As a consequence, it is expected that the meeting of performing art and robotic technology gives positive influence on activating performing art as one of the integrated cultural phenomenon which satisfies the taste of modern era. Moreover, this study may also be the beginning of the expansion of performing art to stretch to diverse field.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.2126-2131
• /
• 2005

Interactions of a humanoid with a human are important, when the humanoid is requested to provide people with human-friendly services in unknown or uncertain environment. Such interactions may require more complicated and human-like behaviors from the humanoid. In this work the arm motions of a human are discussed as the early stage of human motion imitation by a humanoid. A motion capture system is used to obtain human-friendly arm motions as references. However the captured motions may not be applied directly to the humanoid, since the differences in geometric or dynamics aspects as length, mass, degrees of freedom, and kinematics and dynamics capabilities exist between the humanoid and the human. To overcome this difficulty a method to adapt captured motions to a humanoid is developed. The geometric difference in the arm length is resolved by scaling the arm length of the humanoid with a constant. Using the scaled geometry of the humanoid the imitation of actor's arm motions is achieved by solving an inverse kinematics problem formulated using optimization. The errors between the captured trajectories of actor arms and the approximated trajectories of humanoid arms are minimized. Such dynamics capabilities of the joint motors as limits of joint position, velocity and acceleration are also imposed on the optimization problem. Two motions of one hand waiving and performing a statement in sign language are imitated by a humanoid through dynamics simulation.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.4
• /
• pp.505-510
• /
• 2005

The RPO(randomized policy optimizer) algorithm, which utilizes probabilistic policy for the action selection, is a recently developed tool in the area of reinforcement learning, and has been shown to be very successful in several application problems. In this paper, we propose a modified RPO algorithm, whose critic network is adapted via RLS(Recursive Least Square) algorithm. In order to illustrate the applicability of the modified RPO method, we applied the modified algorithm to Kimura's robot and observed very good performance. We also developed a MATLAB-based animation program, by which the effectiveness of the training algorithms on the acceleration or the robot movement were observed.

• The KIPS Transactions:PartD
• /
• v.11D no.2
• /
• pp.461-470
• /
• 2004

As the semiconductor industrials change 200㎜-sized semiconductor wafer production process to 300㎜-sized one, it requires to develop the software for monitoring and simulating the robot which transfers a 300㎜-sized semiconductor wafer. Because such a software don't run at standalone but communicate MCS(Material Control System) and Its subsystem a robot, its architecture is very complex. Therefore, in order to develop such a software systematically, we must utilize an object-oriented development methodology. UML. This paper presents an UML process application developing the software for monitoring and simulating the robot which transfers a semiconductor wafer on the production process.

• The Journal of Korea Robotics Society
• /
• v.3 no.2
• /
• pp.146-153
• /
• 2008

본 논문에서는 인간 운동 제어 이론과 기계학습을 기반으로 하여 로봇의 접촉 작업 수행을 위한 새로운 운동 학습 전략을 제시하였다. 성공적인 접촉 작업 수행을 위한 본 연구의 전략은 강화학습 기법을 통하여 최적의 작업 수행을 위한 임피던스 매개 변수를 찾는 것이다. 본 연구에서는 최적의 임피던스 매개 변수를 결정하기 위하여 Recursive Least-Square (RLS) 필터 기반 episodic Natural Actor-Critic 알고리즘이 적용되었다. 본 논문에서는 제안한 전략의 효용성을 증명하기 위해 동역학 시뮬레이션을 수행하였고, 그 결과를 통하여 접촉작업에서의 작업 최적화 및 환경이 가지는 불확실성에 대한 적응성을 보여 주었다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 1998.06b
• /
• pp.16.1-164
• /
• 1998

This paper describes Facial Expression Explorer to search for the components of a facial expression and to map the expression to other expressionless figures like a robot, frog, teapot, rabbit and others. In general, it is a time-consuming and laborious job to create a facial expression manually, especially when the facial expression must personify a well-known public figure or an actor. In order to extract a blending ratio from facial images automatically, the Facial Expression Explorer uses Networked Genetic Algorithm(NGA) which is a fast method for the convergence by GA. The blending ratio is often used to create facial expressions through shape blending methods by animators. With the Facial Expression Explorer a realistic facial expression can be modeled more efficiently.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.23 no.6
• /
• pp.565-570
• /
• 2013

The mirror neuron regions which are distributed in cortical area handled a functionality of intention recognition on the basis of imitative learning of an observed action which is acquired from visual-information of a goal-directed action. In this paper an automated intention recognition system is proposed by applying computational model of mirror neuron system to the human-robot interaction system. The computational model of mirror neuron system is designed by using dynamic neural networks which have model input which includes sequential feature vector set from the behaviors from the target object and actor and produce results as a form of motor data which can be used to perform the corresponding intentional action through the imitative learning and estimation procedures of the proposed computational model. The intention recognition framework is designed by a system which has a model input from KINECT sensor and has a model output by calculating the corresponding motor data within a virtual robot simulation environment on the basis of intention-related scenario with the limited experimental space and specified target object.