• 대한인간공학회지
• /
• 제11권2호
• /
• pp.15-21
• /
• 1992

This paper presents the ROBOT MODAPTS(Modular Arrangement of Predetermi- ned Time Standards) technique which can be applied to the robot motion analysis. Robot motions are easily divided into several movement activities and terminal activi- ties by means of the ROBOT MODAPTS technique Each link of robot arm is numbered such that finger is 1, hand is 2, elbow is 3, and so on. The robot motion time of each link is counted by multipling its given number to some time values observed at the finger movement. We can easily estimate robot teaching task time with this technique. This technique can be applied efficiently to establishing an adaptable robot motion ergonomically.

• 한국지능시스템학회논문지
• /
• 제19권5호
• /
• pp.661-666
• /
• 2009

뱀형 모듈라 로봇은 모듈을 일련의 형태로 연결하여 구성한 것으로, 다양한 환경에 대해서 강인성을 가지고 있고, 모듈 일부의 고장에도 이동할 수 있는 장점을 가진다. 그러나 이동 제어 방법이 어렵고, 아직까지 효율적이고 다양한 이동법의 개발이 미비한 편이다. 본 연구에서는 뱀형 로봇의 이동제어를 위하여 GA(Genetic Algorithm)기반의 위상생성 방식과 임의의 궤적 생성방식을 비교하고, 이를 확장하여 일부 모듈만의 선택에 따른 영향을 분석하기 위해서 GA를 통한 모듈 선택 실험을 수행하였다. KMC사의 뱀형 로봇을 대상으로 먼저 webots 시뮬레이터 상에서 모델링 및 시뮬레이션 환경을 구축하고, 위의 GA 기반 이동 생성 실험들을 수행하였다.

• Advances in robotics research
• /
• 제1권1호
• /
• pp.101-126
• /
• 2014

In this article we present modular neural control for a leg-wheel hybrid robot consisting of three legs with omnidirectional wheels. This neural control has four main modules having their functional origin in biological neural systems. A minimal recurrent control (MRC) module is for sensory signal processing and state memorization. Its outputs drive two front wheels while the rear wheel is controlled through a velocity regulating network (VRN) module. In parallel, a neural oscillator network module serves as a central pattern generator (CPG) controls leg movements for sidestepping. Stepping directions are achieved by a phase switching network (PSN) module. The combination of these modules generates various locomotion patterns and a reactive obstacle avoidance behavior. The behavior is driven by sensor inputs, to which additional neural preprocessing networks are applied. The complete neural circuitry is developed and tested using a physics simulation environment. This study verifies that the neural modules can serve a general purpose regardless of the robot's specific embodiment. We also believe that our neural modules can be important components for locomotion generation in other complex robotic systems or they can serve as useful modules for other module-based neural control applications.

• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 2001년도 The Pacific Aisan Confrence On Intelligent Systems 2001
• /
• pp.59-64
• /
• 2001

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 chose 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 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 it 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 effectively, AMMQL can show higher adaptability to environmental changes than pure MQL. This paper introduces the concept of AMMQL and presents details of its application into dynamic positioning of robot soccer agents.

• 한국시뮬레이션학회:학술대회논문집
• /
• 한국시뮬레이션학회 2001년도 The Seoul International Simulation Conference
• /
• pp.321-324
• /
• 2001

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 Beaming 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 loaming is different from supervised teaming in the sense that there is no presentation of input-output pairs as training examples. Furthermore, model-free reinforcement loaming algorithms like Q-learning do not require defining or loaming any models of the surrounding environment. Nevertheless it 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. This paper introduces the concept of AMMQL and presents details of its application into dynamic positioning of robot soccer agents.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2005년도 ICCAS
• /
• pp.486-491
• /
• 2005

There come various types of robot with researches for mobile robot. This paper introduces the multi-joint snake robot having 16 degree of freedom and composing of eight-axis. The biological snake robot uses the forward movement friction and the proposed artificial snake robot uses the un-powered wheel instead of the body of snake. To determine the enable joint angle of each joint, the controller inputs are considered such as color and distance using PC Camera and ultra-sonic sensor module, respectively. The movement method of snake robot is sequential moving from head to tail through body. The target for movement direction is decided by a certain article be displayed in the PC Camera. In moving toward that target, if there is any obstacle then the snake robot can avoid by itself. In this paper, we show the method of snake robot for tracing the target with experiment.

• 한국해양공학회지
• /
• 제29권5호
• /
• pp.380-385
• /
• 2015

In the development of an underwater construction robot, the reliability of the operating system is the most important issue because of its huge maintenance cost, especially in a deep sea application. In this paper, we propose a new redundant architectural design for the hydraulic control system of an underwater construction robot. The proposed architecture consists of dual independent modular redundancy management systems linked with a commercial profibus network. A cold standby redundancy management system consisting of a preprocessing switch circuit is applied to the signal network, and a hot standby redundancy management system is adapted to utilize two main controllers.

• 국제학술발표논문집
• /
• The 9th International Conference on Construction Engineering and Project Management
• /
• pp.336-343
• /
• 2022

Industrial robotic arms are widely adopted in numerous industries for manufacturing automation under factory settings, which eliminates the limitations of manual labor and provides significant productivity and quality benefits. The U.S. modular construction industry, despite having similar controlled factory environments, still heavily relies on manual labor. Thus, this study investigates the U.S., Canada, and Europe-based leading modular construction companies and research labs implementing industrial robotic arms for manufacturing automation. The investigation mainly considered the current research scope, industry state, and constraints, as well as identifying the types and specifications of the robotic arms in use. First, the study investigated well-recognized modular building associations, the Modular Building Institute (MBI), and renowned architecture design magazine, Dezeen to gather industry updates. The authors discovered one university lab and a few companies that adopted Switzerland-based robotic arms, ABB. Researching ABB robotics led to the discovery of ABB's competitor, Germany-based KUKA robotic arms. Consequently, research extended to the companies and labs adopting KUKA models. In total, this study has identified seven modular companies and four research labs. All companies employed robotic arms and gantry robot combinations in a production-line-like system for partial automation, and some adopted design standardization for optimization. The common goal among the labs was to achieve greater flexibility and full automation with robotic arms. This study will help companies better implement robotic arm automation by providing recommendations from investigating its current industry status.

• 한국지능시스템학회:학술대회논문집
• /
• 한국지능시스템학회 2008년도 춘계학술대회 학술발표회 논문집
• /
• pp.347-350
• /
• 2008

모듈라 뱀형 로봇은 고장에 대한 강인성과 환경에 유연한 이동 특성을 가지고 있으나, 제어가 어렵다는 단점이 있다. 진화연산을 로봇에 이용한 많은 연구가 진행되어 왔지만, 어떤 기법의 진화연산이 문제에 더 적합하고, 높은 성능을 얻을 수 있는지에 대한 비교는 거의 이루어지지 않고 있다. 본 논문은 두 가지 대표적인 진화기법인 GA와 GP를 이용하여 모듈라 뱀형 로봇의 이동 제어를 수행하였다. 대상 로봇은 H/W로 구현이 가능한 실제 모듈로 구성되었고, Webots을 사용하여 시뮬레이션 실험을 수행하였으며, GA와 GP 기법에 의한 결과를 비교 분석하였다.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2002년도 추계학술대회 및 정기총회
• /
• pp.415-418
• /
• 2002

MR의 자율주행 기능에는 추종, 접근, 충돌회피, 경고 등의 여러 기능이 있다. 이 기능들을하나의 Neural Network로 학습시키는 것은 어려운 일이다. 이것을 보안하고자 기능들을 각각의 Module로 구성하여 상황에 맞게 학습된 Module의 출력 값으로 MR을 제어하였다 로봇은 인간의 감각을 대신할 수 있는 다중 초음파 센서와 PC 카메라를 장착하고 있으며, 이곳에서 측정된 환경정보 데이터들은 Modular Neural Network을 통해 학습이 이루어진다 MNN에서의 출력값은 Gating Network(GN)에서 로봇의 진행 방향과 속도를 스위칭 출력함으로서 MR을 제어하는데 사용된다. MNN 내 EN의 활성화 함수 최적결합을 통해 효과적인 MNN을 구성하였다. 본 논문에서는 Modular Neural Network의 Expert Network(EN)을 최적설계 하였고, 제안한 MNN의 검증을 위해 실시간으로 MR에 구현하였다.