• 제어로봇시스템학회:학술대회논문집
• /
• 1993.10a
• /
• pp.660-665
• /
• 1993

Neural networks are used in the framework of sensorbased tracking control of robot manipulators. They learn by practice movements the relationship between PSD ( an analog Position Sensitive Detector) sensor readings for target positions and the joint commands to reach them. Using this configuration, the system can track or follow a moving or stationary object in real time. Furthermore, an efficient neural network architecture has been developed for real time learning. This network uses multiple sets of simple backpropagation networks one of which is selected according to which division (corresponding to a cluster of the self-organizing feature map) in data space the current input data belongs to. This lends itself to a very fast training and processing implementation required for real time control.

• Journal of Ocean Engineering and Technology
• /
• v.34 no.1
• /
• pp.13-18
• /
• 2020

An azimuth propeller can generate thrust in all directions by rotating its housing with an electric motor. An azimuth propeller can be operated using several methods to stop a ship. This study aims to derive an efficient method to stop a ship safely using an azimuth propeller through full-scale maneuvering trials with the research vessel "NARA" of Pukyong National University in 4.63 m/s (9 kts). Five methods with different azimuth propeller operations were tested to stop the ship. The test results confirmed that the simultaneous use of the thrust and the hydrodynamic force acting on the strut is the most effective method to stop the ship.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2001.04a
• /
• pp.239-243
• /
• 2001

Neural networks are used in the framework of sensorbased tracking control of robot manipulators. They learn by practice movements the relationship between PSD (an analog Position Sensitive Detector) sensor readings for target positions and the joint commands to reach them. Using this configuration, the system can track or follow a moving or stationary object in real time. Furthermore, an efficient neural network architecture has been developed for real time learning. This network uses multiple sets of simple backpropagation networks one of which is selected according to which division (corresponding to a cluster of the self-organizing feature map) in data space the current input data belongs to. This lends itself to a very training and processing implementation required for real time control.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2000.04a
• /
• pp.171-175
• /
• 2000

Neural networks are in the framework of sensorbased tracking control of robot manipulators. They learn by practice movements the relationship between PSD (an analog Position Sensitive Detector) sensor readings for target positions and the joint commands to reach them. Using this configuration, the system can track or follow a moving or stationary object in real time. Furthermore, an efficient neural network architecture has been developed for real time learning. This network uses multiple sets of simple backpropagation networks one of which is selected according to which division (corresponding to a cluster of the self-organizing feature map) in data space the current input data belongs to. This lends itself to a very training and processing implementation required for real time control.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2000.10a
• /
• pp.108-112
• /
• 2000

Neural networks are used in the framework of sensorbased tracking control of robot manipulators. They learn by practice movements the relationship between PSD (an analog Position Sensitive Detector) sensor readings for target positions and the joint commands to reach them. Using this configuration, the system can track or follow a moving or stationary object in real time. Furthermore, an efficient neural network architecture has been developed for real time learning. This network uses multiple sets of simple backpropagation networks one of which is selected according to which division (corresponding to a cluster of the self-organizing feature map) in data space the current input data belongs to. This lends itself to a very training and processing implementation required for real time control.

• Journal of the Korean Society of Industry Convergence
• /
• v.7 no.2
• /
• pp.167-172
• /
• 2004

Neural networks are used in the framework of sensor based tracking control of robot manipulators. They learn by practice movements the relationship between PSD(an analog Position Sensitive Detector) sensor readings for target positions and the joint commands to reach them. Using this configuration, the system can track or follow a moving or stationary object in real time. Furthermore, an efficient neural network architecture has been developed for real time learning. This network uses multiple sets of simple back propagation networks one of which is selected according to which division (Corresponding to a cluster of the self-organizing feature map) in data space the current input data belongs to. This lends itself to a very training and processing implementation required for real time control.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1997.11a
• /
• pp.259-262
• /
• 1997

This paper presents a visual servoing of RV-M2 robot manipulators to track and grasp moving object, using pruned dynamic recurrent neural networks(DRNN). The object is stationary in the robot work space and the robot is tracking and grasping the object by using CCD camera mounted on the end-effector. In order to optimize the structure of DRNN, we decide the node whether delete or add, by mutation probability, first in case of delete node, the node which have minimum sum of input weight is actually deleted, and then in case of add node, the weight is connected according to the number of case which added node can reach the other nodes. Using evolutionary programming(EP) that search the struture and weight of the DRNN, and evolution strategies(ES) which train the weight of neuron, we pruned the net structure of DRNN. We applied the DRNN to the Visual Servoing of a robot manipulators to control position and orientation of end-effector, and the validity and effectiveness of the pro osed control scheme will be verified by computer simulations.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.21 no.5
• /
• pp.552-557
• /
• 2015

A numerical simulation studies were performed to investigate a manoeuvring characteristics as function of stern hull form with the mathematical model. In order to consider the effect of the stern hull form and obtain the manoeuvring characteristics, a parameter($C_{wa}$) which is aft. water plane area coefficient is modified. Because modifying $C_{wa}$(${\pm}2%$) means that the stern hull form is modified to V-type or U-type, the numerical simulation was performed with this modified $C_{wa}$. A changing trend for the manoeuvring characteristics not only in deep water but also in shallow water such as directional stability, turning and zig-zag was investigated and presented as the results. Present study showed that the manoeuvrability in shallow water largely changed when the draught and water depth ratio(=d/H) become 0.5, and the stern hull form can affect to the manoeuvrability of a vessel navigating in restricted water depth. In addition, it showed that approaching the stern hull to U-type makes the advance and tactical diameter of turning motion large and the overshoot angle of zig-zag motions small. Otherwise, it showed approaching the stern hull form to V-type makes the advance and tactical diameter of turning motion small and the overshoot angle of zig-zag motions large in the present study.

• Physical Therapy Rehabilitation Science
• /
• v.3 no.1
• /
• pp.55-62
• /
• 2014

Objective: To study the effects of gait training with additional weight and gait training with non-additional weight on balance ability and gait ability in patients with chronic stroke through comparative analysis. Design: Randomized controlled trials. Methods: The subjects were divided randomly into two groups: additional weight group (AWG, n=12), and non-additional weight group (NAWG, n=10). Both groups received general physical therapy for 30 min in 1 session, 5 sessions per week during 6 months. The AWG practiced gait training with additional weight of 0.1 and 0.5 kg for 20 min a day, 3 days per week for 6 months and the NAWG practiced gait training with non-additional weight for 20 min a day, 3 days per week for 6 months. Patients in both groups were instructed to walk as fast as they could along a 35 m long track (straight for 20 m and curved for 15 m). Patients walked with their hemiplegic side on the inside of the track while a physical therapist followed along to instruct patients to maintain a straight posture. Balance ability was tested with the Functional Reach Test, the Timed Up and Go test, and the Berg Balance Scale, and gait ability was tested with GAITRite. The results of balance and gait ability were analyzed before and after interventions. Results: A significant increase in FRT, TUG, BBS was seen in both groups after intervention (p<0.05). A significant increase in gait ability was seen in the AWG after intervention (p<0.05). For balance and gait ability, the results from the AWG was significantly improved compared with the NAWG (p<0.05). Conclusions: Gait training with additional weight improves balance ability and gait ability in stroke patients, this gait training method is effective and suitable for stroke patients to increase the ability of functional performance.

• Journal of the Korean Society for Railway
• /
• v.15 no.6
• /
• pp.631-637
• /
• 2012

New Opening of Suseo-Pyeongteak High Speed Railway (HSR) will be a new leap in the Korean railway history. However if this section of HSR line around Seoul Metropolitan Area opens, the confluence of new HSR and existing HSR line in Pyeongteak-Osong section will cause a bottle neck problem. In other words, the opening of Suseo-Pyeongteak HSR line will make the capacity of track reach the limit and the section of railroad between Pyeongteak and Osong will be saturated. This will also make such troubles as restricting the number of train which stops at Cheonan-Asan station. In this study, based on the train assignment theory of TVM430 signal system, the methods of calculating headway and number of train are reviewed and the plan for application of optimal operation pattern during peak hour between Pyeongteak-Osong section is also suggested. To remove the bottle neck problem in this HSR section, 3 alternatives are suggested and the expected effects and problems of each alternative are also analyzed. The results show that the troubles caused by excess of track capacity can be removed without any additional cost if the minimum headway in operating system for HSR is adopted in this section. In the future, if these alternatives are considered to the long-term plan for operating train and signal systems, this will improve the efficiency of train operation, which can remove the bottle neck in the HSR line.