• Proceedings of the Korean Society of Computer Information Conference
• /
• 2020.01a
• /
• pp.189-192
• /
• 2020

본 논문은 요양원·요양병원에 입원한 노인들을 위한 실행능력중심 기능성 스마트 게임기기 및 데이터를 관리할 수 있는 시스템을 개발하고자 한다. 시스템은 증강현실 기반의 스마트맵, 동작 감응형 스마트 스틱, 스마트 명령과 연동하는 로봇, 이러한 게임 콘텐츠로 동작되는 모든 기기에서 나오는 데이터를 수집·관리·분석하는 정보 측정 및 분석 시스템으로 동작되는 시스템을 설계하였다. 이 시스템을 통해 한국이나 일본의 노인병원에서 간절히 필요로 하는 요양보호사 없이 노인들끼리 셀프 진행 가능하고 재미있는 기능성 콘텐츠의 수요를 만족시켜 줄 수 있을 것이다. 또한 치매와 같이 완벽한 치료가 불가능한 질병에도 증상악화를 늦추는 효과가 있으리라 예상한다. 향후 시스템을 개발하여 직접 실험을 하면서 다양한 데이터 수집을 통해 빅데이터로 활용하여 보다 전문적이고 효율적인 재활을 진행할 것이다.

• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.3
• /
• pp.141-148
• /
• 2022

Recently, reinforcement learning combined with deep neural network technology has achieved remarkable success in various fields such as board games such as Go and chess, computer games such as Atari and StartCraft, and robot object manipulation tasks. However, such deep reinforcement learning describes states, actions, and policies in vector representation. Therefore, the existing deep reinforcement learning has some limitations in generality and interpretability of the learned policy, and it is difficult to effectively incorporate domain knowledge into policy learning. On the other hand, dNL-RRL, a new relational reinforcement learning framework proposed to solve these problems, uses a kind of vector representation for sensor input data and lower-level motion control as in the existing deep reinforcement learning. However, for states, actions, and learned policies, It uses a relational representation with logic predicates and rules. In this paper, we present dNL-RRL-based policy learning for transportation mobile robots in a manufacturing environment. In particular, this study proposes a effective method to utilize the prior domain knowledge of human experts to improve the efficiency of relational reinforcement learning. Through various experiments, we demonstrate the performance improvement of the relational reinforcement learning by using domain knowledge as proposed in this paper.

• The KIPS Transactions:PartB
• /
• v.9B no.2
• /
• pp.139-146
• /
• 2002

The robot soccer simulation game is a dynamic multi-agent environment. In this paper we suggest a new reinforcement learning approach to each agent's dynamic positioning in such dynamic environment. Reinforcement learning is the machine learning in which an agent learns from indirect, delayed reward an optimal policy to choose sequences of actions that produce the greatest cumulative reward. Therefore the reinforcement learning is different from supervised learning in the sense that there is no presentation of input-output pairs as training examples. Furthermore, model-free reinforcement learning algorithms like Q-learning do not require defining or learning any models of the surrounding environment. Nevertheless these algorithms can learn the optimal policy if the agent can visit every state-action pair infinitely. However, the biggest problem of monolithic reinforcement learning is that its straightforward applications do not successfully scale up to more complex environments due to the intractable large space of states. In order to address this problem, we suggest Adaptive Mediation-based Modular Q-Learning (AMMQL) as an improvement of the existing Modular Q-Learning (MQL). While simple modular Q-learning combines the results from each learning module in a fixed way, AMMQL combines them in a more flexible way by assigning different weight to each module according to its contribution to rewards. Therefore in addition to resolving the problem of large state space effectively, AMMQL can show higher adaptability to environmental changes than pure MQL. In this paper we use the AMMQL algorithn as a learning method for dynamic positioning of the robot soccer agent, and implement a robot soccer agent system called Cogitoniks.

• Journal of Institute of Control, Robotics and Systems
• /
• v.6 no.1
• /
• pp.73-80
• /
• 2000

The stewart platform manipulator proposed by stewart is the parallel manipulator which is composed of several independent actuators connecting the upper plate with the base plate and capable of executing a six degree of freedom motion. The manipulator has a structure of a closed loop form, and provides better load-to-weight ratio and ratio and rigidity than a serial manipulator with an open loop form. Moreover, the manipulator has high positional accuracy because position errors of actuators are not additive. Because of these advantages, this manipulator is widely used in many engineering applications such as a driving simulator, a tool of machining center, a force/torque sensor and so on. When this Stewart platform manipulator is controlled in joint space, it is difficult to design a controller using an analytic method due to nonhnearity and unknown parameters of actuators. Therefore, a PID controller is often used because of easiness in applications. To find the PID control gain, a trial-and-error method is generally used. This method is time-consuming, and does not guarantee a optimal gain. Thus, this paper proposes a GA-PID controller which selects an optimal PID control gain using genetic algorithms. And this proposed controller is evaluated experimentally and shows acceptable performance.

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.23 no.1
• /
• pp.75-82
• /
• 2014

For accurate gain tuning of the lab-manufactured six-axis articulated robot RS2 with less noise, in this study, a program routine using dynamic signal analyzer, which is a realization of a controller design algorithm in the frequency domain, is programmed using LabVIEW$^{(R)}$. The contribution of this paper is the proposal of a simulation technique based on SolidWorks$^{(R)}$ and LabVIEW$^{(R)}$ for the gain tuning of a six-axis articulated robot. To realize the simulation, the LabVIEW$^{(R)}$ program used for experimental gain tuning is incorporated in to SolidWorks$^{(R)}$. A comparison shows that the results of simulation-based gain tuning and experimental gain tuning are almost the same within a 5% error bound. On the basis of the comparison, it can be suggested that the simulation-based technique for gain tuning can be applied instead of experimental gain tuning to a six-axis articulated robot by interlocking SolidWorks$^{(R)}$ and LabVIEW$^{(R)}$.

• Journal of Internet Computing and Services
• /
• v.7 no.3
• /
• pp.31-39
• /
• 2006

Ubiquitous and pervasive environment presents opportunities for a rich set of location aware applications such as car navigation and intelligent robots, interactive virtual games, logistics service, asset tracking etc. Typical indoor location systems require better accuracy than what current outdoor location systems provide, Outdoor location technologies such as GPS have poor indoor performance because of the harsh nature of indoor environments, In this paper we present a novel reducing interference location system that is particularly well suited to support context aware computing. The system, called EEM (Enhance Envelop Method) alms to combine the advantages of real time tracking systems that implement distributed environment with the suitability of infrastructure sensor network.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.1 no.2
• /
• pp.49-55
• /
• 2000

In this paper, we show that our previously proposed second-order iterative learning control method with feedback is more effective and has better convergence performance than the second-order iterative learning control method without feedback, particularly in the case of the existence of initial condition errors. Also the convergence proof of the proposed method is given. And through the simulation result of applying the proposed method to the linear time-varying system, we show that our proposed method has enhanced robustness and stability in case of the existence of initial condition errors.

• Journal of Internet Computing and Services
• /
• v.18 no.4
• /
• pp.43-50
• /
• 2017

Embedded System includes touch, GPS, motion, and acceleration sensor, and can communicate with neighbor devices using wireless communication. Because Arduino with embedded system provides good environment for development and application, developers, engineers, designers, as well as artists, students have a great interest. They utilize Arduino in the robot, home appliances, fashion, culture and so on. In this paper, we design and implement a game using Arduino with embedded system which recognizes the human movement by moving away from one-dimensional game of the existing touch method. Implemented embedded system game measures gyro-sensor to recognize human movement and detects the attack success of the opponent by using touch sensor. Moreover, health of the game player is updated in the real time through the android phone-based database. In this paper, implemented embedded system-based game provides GUI screen of android phone. It is possible to select watching mode and competition mode. Also, it has low energy consumption and easy to expand because it send and receive data packet through recent Bluetooth communication.

• Journal of Korea Game Society
• /
• v.10 no.6
• /
• pp.157-168
• /
• 2010

A terrain is an essential element for constructing a virtual world in which game characters and objects make various interactions with one another. Creating a terrain requires a great deal of time and repetitive editing processes. This paper presents a 3D terrain reconstruction system to create 3D terrain in virtual space based on real terrain data. In this system, it converts the coordinate system of the height maps which are generated from a stereo camera and a laser scanner from global GPS into 3D world using the x and z axis vectors of the global GPS coordinate system. It calculates the movement vectors and the rotation matrices frame by frame. Terrain meshes are dynamically generated and rendered in the virtual areas which are represented in an undirected graph. The rendering meshes are exactly created and updated by correcting terrain data errors. In our experiments, the FPS of the system was regularly checked until the terrain was reconstructed by our system, and the visualization quality of the terrain was reviewed. As a result, our system shows that it has 3 times higher FPS than other terrain management systems with Quadtree for small area, improves 40% than others for large area. The visualization of terrain data maintains the same shape as the contour of real terrain. This system could be used for the terrain system of realtime 3D games to generate terrain on real time, and for the terrain design work of CG Movies.

• Journal of Intelligence and Information Systems
• /
• v.11 no.1
• /
• pp.153-167
• /
• 2005

In this paper, we consider the problem of exploring unknown environments with a mobile robot or an autonomous character agent. Traditionally, research efforts to address the space exploration problem havefocused on the graph-based space representations and the graph search algorithms. Recently EXPLORE, one of the most efficient search algorithms, has been discovered. It traverses at most min$min(mn, d^2+m)$ edges where d is the deficiency of a edges and n is the number of edges and n is the number of vertices. In this paper, we propose DFS-RTA* and DFS-PHA*, two real-time graph search algorithms for directing an autonomous agent to explore in an unknown space. These algorithms are all built upon the simple depth-first search (DFS) like EXPLORE. However, they adopt different real-time shortest path-finding methods for fast backtracking to the latest node, RTA* and PHA*, respectively. Through some experiments using Unreal Tournament, a 3D online game environment, and KGBot, an intelligent character agent, we analyze completeness and efficiency of two algorithms.