• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.51 no.6
• /
• pp.258-266
• /
• 2002

This paper proposes the methodology of the image processing algorithm that estimates the pose of the inner-pipe crawling robot. The inner-pipe crawling robot is usually equipped with a lighting device and a camera on its head for monitoring and inspection purpose of defects on the pipe wall and/or the maintenance operation. The proposed methodology is using these devices without introducing the extra sensors and is based on the fact that the position and the intensity of the reflected light from the inner wall of the pipe vary with the robot posture and the camera. The proposed algorithm is divided into two parts, estimating the translation and rotation angle of the camera, followed by the actual pose estimation of the robot . Based on the fact that the vanishing point of the reflected light moves into the opposite direction from the camera rotation, the camera rotation angle can be estimated. And, based on the fact that the most bright parts of the reflected light moves into the same direction with the camera translation, the camera position most bright parts of the reflected light moves into the same direction with the camera translation, the camera position can be obtained. To investigate the performance of the algorithm, the algorithm is applied to a sewage maintenance robot.

• Proceedings of the KSME Conference
• /
• 2000.04a
• /
• pp.523-528
• /
• 2000

In this paper, the new type biped walking robot which is composed of the minimum number or links just for walking and its appropriate gaits are proposed. The proposed new gaits for this robot are four-crossing, crawling, standing and turning gait. In designing the biped robot we propose the Performance Index which means the needed torque per a moving distance and generate foot trajectories by $3^{rd}$ order spline Interpolation. Among those, numerically we find the optimal conditions which minimize the Performance Index. Dynamically stable walking of the biped robot is realized by satisfying the stability condition of ZMP(zero moment point), which is related to maintaining the ZMP within the region of the supporting foot during the s1n91e leg support phase. We determine the region of mass center from the stability condition of ZMP and plan references which track the mass conte. trajectory of constant velocity. Finally we implement the gaits statically tracking the planned trajectories using PD control method.

• Journal of the Korean Society of Mechanical Technology
• /
• v.20 no.6
• /
• pp.912-916
• /
• 2018

Crawling robots are advantageous in overcoming obstacles. These robots have characteristics such as light weight and outstanding mobility. In case of large robots, they have difficulties passing narrow gaps or entering the cave. In this paper, we propose a milli-scale hexapedal robot using 4-bar linkages. Two conditions are necessary to enable efficient walking. In short, the trajectory of the foot must be elliptical, and the lowest point of the foot should be the same. These conditions are satisfied with a novel leg design. The robot has a pair of three legs and the legs are coupled to operate simultaneously. Each set of the legs are installed to robot's both sides and the legs satisfy the equal lowest foot point and elliptical trajectory. As a result, this hexapedal robot can crawl with 0.56m/s speed.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.4
• /
• pp.519-524
• /
• 2009

Recently, reinforcement learning methods have drawn much interests in the area of machine learning. Dominant approaches in researches for the reinforcement learning include the value-function approach, the policy search approach, and the actor-critic approach, among which pertinent to this paper are algorithms studied for problems with continuous states and continuous actions along the line of the actor-critic strategy. In particular, this paper focuses on presenting a method combining the so-called ACFRL(actor-critic fuzzy reinforcement learning), which is an actor-critic type reinforcement learning based on fuzzy theory, together with the RLS-NAC which is based on the RLS filters and natural actor-critic methods. The presented method is applied to a control problem for crawling robots, and some results are reported from comparison of learning performance.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.645-650
• /
• 2003

Legged robots can avoid an obstacle by crawling-over or striding, according to the obstacle’s nature and the current state of the robot. Thus, it can be observed that the mobility efficiency to reach a destination is improved by such action. Moreover, if robots have many legs like 4-legged or 6-legged types, then the robot movement range is affected by the order of swing leg. In this paper, the avoidance action of a quadruped robot is generated by a neural network (NN) whose inputs are information on the position of the destination, the obstacle configuration and the robot's self-state. To realize a free gait in static walking, the order of swing leg is determined using an another NN whose inputs are the amount of movements and the robot’s self-state. The design parameter of the latter NN is adjusted by using genetic algorithm (GA).

• The KIPS Transactions:PartD
• /
• v.11D no.3
• /
• pp.563-570
• /
• 2004

As the use of the Internet becomes more popular, a huge amount of information is published on the Web, and users can access the information effectively with Web search services. Since Web search services retrieve relevant documents from those collected by Web robots we need to improve the crawling quality of Web robots. In this paper, we suggest evaluation criteria for Web robots such as efficiency, continuity, freshness, coverage, silence, uniqueness and safety, and present various functions to improve the performance of Web robots. We also investigate the functions implemented in the conventional Web robots of NAVER, Google, AltaVista etc. It is expected that this study could contribute the development of more effective Web robots.