• Proceedings of the Korea Information Processing Society Conference
• /
• 2005.05a
• /
• pp.1619-1622
• /
• 2005

본 논문에서는 로봇이 작업을 계획하기 위해 필요한 3차원 작업 공간을 세 가지의 컨텍스트(context)들을 이해함으로써 빠르게 모델링하는 새로운 기법을 소개 하고 있다. 로봇이 사람과 비슷한 속도와 정확도로 작업 공간을 이해하고 모델링하는 것에 초점을 두고 있으며 이를 위해 작업 공간상의 특징적인 세 가지의 컨텍스트(작업공간의 간략화를 위한 전체 공간상의 평면특징, 데이터베이스에 미리 정의된 물체 그리고 로봇의 주어진 작업에 따라 다양한 상세함을 갖는 그 외의 장애물)를 정의하였고, 그것들을 빠르게 이해함으로써 어떻게 3차원 작업 공간을 형성하는지 설명하고 있다. 본 논문에서 3 차원 정보를 갖는 scale invariant feature transformation(SIFT)를 stereo-sis SIFT 로 간주했으며 이를 이용하여 위에서 언급한 컨텍스트들을 이해하였고 다양한 카메라의 위치로부터 얻어지는 여러 개의 장면들을 정합하였다. 또한, 실험을 통해 제안한 방법의 타당성도 검증하였다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• v.9 no.1
• /
• pp.295-299
• /
• 2005

In this paper, path planning and tracking problems are mentioned to guarantee efficient and safe navigation of autonomous mobile robots. We focus on the path planning and also deal with the path tracking and obstacle avoidance. We improved the conventional distance transform (DT) algorithm for the path planning. Using the improved DT algorithm, we obtain paths with shorter distances compared to the conventional DT algorithm. In the stage of the path tracking, we employ the fuzzy logic controller to conduct the path tracking behavior and obstacle avoidance behavior. Through computer simulation studies, we show the effectiveness of the proposed navigational algorithm for autonomous mobile robots.

• The KIPS Transactions:PartB
• /
• v.16B no.6
• /
• pp.489-496
• /
• 2009

Hierarchical Task Network(HTN) planning, a typical planning method for effectively taking advantage of domain-specific control knowledge, has been widely used in complex real applications for a long time. However, it still lacks theoretical formalization and standardization, and so there are some differences among existing HTN planners in terms of principle and performance. In this paper, we present an effective way to translate a HTN planning domain specification into the corresponding standard PDDL specification. Its main advantage is to allow even many domain-independent classical planners to utilize domain-specific control knowledge contained in the HTN specifications. In this paper, we try our translation-based approach to three different domains such as Blocks World, Office Delivery, Hanoi Tower, and then conduct some experiments with a forward-chaining heuristic state-space planner, FF, to analyze the efficiency of our approach.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.15 no.12
• /
• pp.6986-6992
• /
• 2014

The generalized Voronoi graph (GVG) is a topological map of a constrained environment. This is defined in terms of workspace distance measurements using only sensor-provided information, with a robot having a maximum distance from obstacles, and is the optimum for exploration and obstacle avoidance. This is the safest path for the robot, and is very significant when studying the GVG edges of highly articulated robots. In previous work, the point-GVG edge and Rod-GVG were built with point robot and rod robot using sensor-based control. An attempt was made to use a higher degree of freedom robot to build GVG edges. This paper presents GVG-based a new local roadmap for the two-link robot in the constrained two-dimensional environment. This new local roadmap is called the two-identical-link generalized Voronoi graph (L2-GVG). This is used to explore an unknown planar workspace and build a local roadmap in an unknown configuration space $R^2{\times}T^2$ for a planar two-identical-link robot. The two-identical-link GVG also can be constructed using only sensor-provided information. These results show the more complex properties of two-link-GVG, which are very different from point-GVG and rod-GVG. Furthermore, this approach draws on the experience of other highly articulated robots.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.12 no.6
• /
• pp.689-696
• /
• 2009

In this paper, a new user interface for unmanned combat vehicle(UCV) is developed based on the mission planning and global path planning. In order to complete a tactical mission given to an UCV, it is essential to design an effective interface scheme between human and UCV considering changing combat environment and characteristics of the mission. The user interface is mainly composed of two parts, mission planning and global path planning, since they are important factors to accomplish combat missions. First of all, mission types of UCV are identified. Based on mission types, the concept of mission planning for UCVs is presented. Then a new method for global path planning is devised. It is capable of dealing with multiple grid maps to consider various combat factors so that paths suitable for the mission be generated. By combining these two, a user interface method is suggested. It is partially implemented in the Dog-horse Robot of ADD and its effectiveness is verified.

• Journal of Navigation and Port Research
• /
• v.42 no.3
• /
• pp.237-244
• /
• 2018

When underwater rubble leveling work is carried out by a robot, real-time information on the topography around the robot is required for remote control. If the topographical information with respect to the current position of the robot is displayed as a 3D graphic image, it allows the operator to plan the working schedules and to avoid accidents like rollovers. Up until now, the topographical recognition was conducted by multi-beam sonars, which were only used to assess the quality before and after the work and could not be used to provide real-time information for remote control. This research measures the force delivered to the bucket which presses the mound to determine whether contact is made or not, and the contact position is calculated by reading the cylinder length. A variable bang-bang control algorithm is applied to control the heavy robot arms for the positioning of the bucket. The proposed method allows operators to easily recognize the terrain and intuitively plan the working schedules by showing relatively 3-D gratifications with respect to the robot body. In addition, the operating patterns of a skilled operator are programmed for raking, pushing, moving, and measuring so that they are automatically applied to the underwater rubble leveling work of the robot.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.41 no.7
• /
• pp.655-663
• /
• 2017

With the development of ICT technology, the indoor utilization of robots is increasing. Research on transportation, cleaning, guidance robots, etc., that can be used now or increase the scope of future use will be advanced. To facilitate the use of mobile robots in indoor spaces, the problem of self-location recognition is an important research area to be addressed. If an unexpected collision occurs during the motion of a mobile robot, the position of the mobile robot deviates from the initially planned navigation path. In this case, the mobile robot needs a robust controller that enables the mobile robot to accurately navigate toward the goal. This research tries to address the issues related to self-location of the mobile robot. A robust position recognition system was implemented; the system estimates the position of the mobile robot using a combination of encoder information of the mobile robot and the absolute space coordinate transformation information obtained from external video sources such as a large number of CCTVs installed in the room. Furthermore, vector field histogram method of the pass traveling algorithm of the mobile robot system was applied, and the results of the research were confirmed after conducting experiments.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.1
• /
• pp.36-41
• /
• 2011

Configuration space(C-space) including configuration obstacle(C-obstacle) is one of the most important concepts in mobile robot path planning. Using C-space and C-obstacles, the robot with different shapes and moving mechanisms can be considered as a point in the C-space. And, as a result, the collision-free path for the robot can be easily achieved. To make C-space including C-obstacle, many researchers used circular approximation method for the efficient caluculation time. This method can help us to save our time by approximating the shape of a robot as the minimum sized circle which can cover all the area of robot. But, by using the circle larger than the robot, more space are considered as the part of robot and, as a result, some obstacles which are very near each other may be considered as a combined one obstacle. To solve this problem, multi-level configuration space is proposed by this paper. This multi-level method also use the circular approximation method as the initial step. But, after finding the initial path, it will check how many obstacles are combined. And then, for each combined obstacle, more accurate C-space generation will be continued. To check the efficiency of the proposed algorithm, time for c-space generation are compared with the well-known accurate C-space generation method using various types of robot shape.

• Proceedings of the Korean Society for Agricultural Machinery Conference
• /
• 2017.04a
• /
• pp.149-149
• /
• 2017

본 논문은 생산라인 자동화를 위한 로봇으로 제조업 시장 확대에 가장 큰 걸림돌이 되는 가격 경쟁력 및 인력난 해소를 위해 설계하였으며 다양한 소재에 대응하기 위해 그리퍼를 교체하여 적용이 되도록 하였다. 자동화를 위한 로봇은 소재의 내외경 가공 및 검사까지 모든 공정이 일괄적으로 이루어 져야하며 LCD 모니터에 생산수량 및 불량률 등의 정보를 실시간으로 나타내어 효율적인 생산계획을 수립할 수 있도록 하였다. 생산라인 자동화를 위해 로봇의 설계는 Auto CAD를 이용하였다. 부품의 가공은 CNC에 적용하기 위해 자동공급장치를 설계하였다. 가공이 완료된 후 측정한 값을 LCD모니터를 통하여 작업자가 알아볼 수 있게 나타냈다. 외경 1은 40.405, 외경2는 32.201, 내경 1은 23.346, 내경 2는 34.302로 나타났다. 측정결과 불량 측정을 위해 측정부의 결과 값이 나타나며 불량이 발생하면 그래프를 이용하여 어떤 부위에서 발생했는지를 알 수 있도록 하였다. 또한 결과 값은 자동으로 저장되도록 하였다. 생산라인 자동화를 위해 100EA를 측정한 결과 외경 1은 40.40438, 외경2는 32.20164, 내경 1은 23.34830 내경 2는 34.30033의 평균값을 나타냈다. 측정값의 검증은 하이트게이지로 측정한 결과 0.003 이내의 결과를 나타냈다. 따라서 본 로봇 자동화 시스템을 적용한다면 생산성 향상 및 불량률 감소가 가능하여 인력대체 및 가격경쟁력이 가능하다고 판단된다.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.42 no.5 s.305
• /
• pp.1-12
• /
• 2005

A Mobile manipulator - a serial connection of a mobile robot and a task robot - is a very useful system to achieve various tasks in dangerous environment. because it has the higher performance than a fixed base manipulator in regard to the size of it's operational workspace. Unfortunately the use of a mobile robot introduces non-holonomic constraints, and the combination of a mobile robot and a manipulator generally introduces kinematic redundancy. In this paper, first a method for estimating the position of object at the cartesian coordinate system acquired by using the geometrical relationship between the image captured by 2-DOF active camera mounted on mobile robot and real object is proposed. Second, we propose a method to determine a optimal path between current the position of mobile manipulator whose mobile robot is non-holonomic and the position of object estimated by image information through the global displacement of the system in a symbolic way, using homogenous matrices. Then, we compute the corresponding joint parameters to make the desired displacement coincide with the computed symbolic displacement and object is captured through the control of a manipulator. The effectiveness of proposed method is demonstrated by the simulation and real experiment using the mobile manipulator.