• Journal of the Korean Society of Safety
• /
• v.20 no.4 s.72
• /
• pp.114-121
• /
• 2005

Railroad shunting movements connecting and disconnecting train sets are very susceptible to human errors since they depend on human decision-making and action procedure that are variable to situation to situation. Nevertheless, in the investigation of railroad accidents, all the accident causes related with human factors have merely been categorized as 'careless treatment' of the workers without any systematic approach of behavioral sciences or the analysis of human errors. In this research, therefore, 137 accident cases occurred during railroad shunting movements and 435 accident cases occurred during driving were analyzed with a special interest of human errors. According to results, the traditional accident investigation scheme used for last several decades did not seem to be appropriate for catching up true accident causes with respect to human errors. In addition, both signal men and locomotive drivers made many mistakes in judgement/action stage while the former mainly commit judgement tasks where as the latter mainly commit cognition tasks. Ant those tasks such as 'confirmation of signal and route', 'location check-up of connected train sets', and 'route identification for a shift of track' ranked highly for accident susceptibility.

• Proceedings of the KSR Conference
• /
• 2008.11b
• /
• pp.1916-1932
• /
• 2008

The railway accident is characterized by widespread and great(immerse) damage to our society when an accident occurred even if the probability of a railway accident is low. It is no doubt that the role of a rolling stock driver who are active in the first line of the rail safety driving is the most important to prevent an unexpected railway accident. Before new railway safety law being taken into effect, each railway operating organization had the responsibility of training a rolling stock driver by qualification criteria (the requirements or standards of competency), training period, probation period decided on their own standards. To train drivers driving same locomotive model, each railway operating organization had a variety of standards to draft and train trainees. Consequently, It was considered inefficient way to train and manage rolling stock drivers. However, after implementing the new railway safety law, a new system including training and evaluation process to issue driver's license on the condition that entire requirements are satisfied has been established. To prove the competency of the rolling stock driver objectively, training center and third-party evaluation center are divided individually and assigned by government office. Training center should design the training program to allot theatrical education time for different trainee group and to secure maximum practical training time utilizing simulator system to enhance the rolling stock driver's competency. This study analyzes certain conditions necessary to maximize the effectiveness and practical use of training for aquisition of a driver's license and the problems of the existing rail safety law and the ways to improve them.

• Proceedings of the KSR Conference
• /
• 2011.10a
• /
• pp.3180-3186
• /
• 2011

A segmented scissors type switch has been developed for the urban transit maglev demonstration line to be commercialized near Incheon International Airport in 2013. Based on the design of the previous segmented 3-way switch, the scissors switch is composed of four segmented 2-way switches up/down and left/right and a turn table in the mid point. The main function of the scissors switch is to change the running direction of the train at end terminals. The developed scissors switch is planned to be installed in front of the 102 station, which has a side platform, of the demonstration line. The total length of the switch is 65m and the distance between the up and down track centerlines is 6m. The 2-way switches and turn table are made of steel box type beams, and have their own driving unit, locking unit, control unit, levitation and propulsion rails, and so on. Installed in the factory, a 100,000-cycle continuous operation test was carried out after manual and automatic test operations. The applicapability of the developed switch was verified through the measurements of the linearity of the track after repetitive operations, the mechanical operation noise, the load of the main driving motor, the safety of the control panel, the natural frequency of the girder, the deformation of the girder, and so on.

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.23 no.1
• /
• pp.7-14
• /
• 2014

The core of the automotive industry's strategy to handle the climate change can be explained as the development and distribution of the vehicles with high fuel efficiencies and low emission. Clean Diesel, hydrogen fuel cell, electric, and especially hybrid power-train vehicles have been actively studied. This paper dynamically analyzes the performance of a hybrid system's five driving modes. The research subject consists of one engine, two electric motors, two simple planetary gears, and one compound planetary gears with five clutches. To define the steady state equation of the system, interaction formulas of five driving modes are introduced with motion variables and torque variables. These formulas are then used to analyze the speeds, torques, and power flows of each mode.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.16 no.7 s.112
• /
• pp.697-703
• /
• 2006

This article proposes a new technique for the reduction of vibration and noise in the geared system by adjusting the distance between gear shafts. The vibration and noise may be produced by the abnormal force applied to the tooth face. And the force may be the cause of ununiform velocity in the driven shaft. If the velocity is obtained to be uniform by adjusting the distance between shafts. the vibration and noise may be reduced to some extent. In order to review, a dynamic analysis model for the gear train used in a mill turret and a test rig are developed. The velocities in the driven shaft are calculated by dynamic simulations for the model and noises in the test rig are measured with varying of the distance between shafts. The comparison of simulation and test data shows that the distance between shafts at the most uniform velocity has the lowest level of noise.

• Proceedings of the KSR Conference
• /
• 2007.05a
• /
• pp.755-760
• /
• 2007

The research will be essential that the method for strengthening safety are safe installations in a hardware portion, a safe running ceremony of the vehicle employee (an engine driver) and the estimate of the situation which is accurate. In railway car the window wiper which is auxiliary equipment for safe operation is changed the existing manual control method with automatic correspondence precipitation used raindrops censor which can assure the vision of driving. It can be easy to safe driving as might have expected the object. We carried out studies technical review investigation which can accomplish the best suited wiping speed in maintaining an introduction existing system in the manual control system applied the existing train and corresponding to precipitation. The observation fact in the studies can be possible to ensure safety operation. That is, Manual operation method (of the driver) to add the automatic speed adjust function can operate improved window wiper drive system when it rains and, as was expected, it is able to ensure the range of a good railroad driver's vision and to concentrate in working at rainfall.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.21 no.10
• /
• pp.951-959
• /
• 2011

In this study, an OTM(on-the-move) antenna which is mounted on ground vehicles and is used for mobile communication between vehicle and satellite while moving was addressed. Since LOS(line-of-sight) of antenna should direct satellite consistently while vehicle moving to guarantee high satellite communication quality, active antenna LOS stabilization is a core technology for OTM antenna. Stabilization of a satellite tracking antenna which consists of 2-DOF gimbals, an elevation gimbal over an azimuth gimbal, was considered in this study. In consideration of driving mechanism which consists of gear train and flexible driving shafts, a two-mass-system dynamic model coupled with vehicle motion was presented. An internal PI-control loop + outer PI-control loop structure has been suggested in order to damp the torsional vibration and stabilize control system. The classical pole-placement method was applied to design control gains. In addition, a vehicle motion compensation control beside of the feedback control loop has been suggested to improve LOS stabilization performances. The feasibility of the proposed control design was verified along with some experimental results.

• Proceedings of the KSR Conference
• /
• 2004.10a
• /
• pp.455-460
• /
• 2004

The Fuel cell Rubber tired Train (FRT), which is now getting the attention as the next generation vehicle with its environment-friendliness, is the transportation for smooth connections of city traffic. It is the revival of the surface-car system with its revaluation of the function and technological development. Accordingly, fixed time operation and high speed driving became possible. FRT is operated together with other vehicles on the regular drive way. While this vehicle can solve the problem of traffic congestion in the urban area, it also can be cost-effective when it is compared to the cost of subway construction. It is also designed to minimize the underground or elevated traffic lane, to introduce the new construction technology, to reduce a term of works, and to cut down the operation cost by unmanned automatic driving system. Furthermore, it is considered as the alternative measure of other transportation due to its potential for the ecological way of speed improvement and the accessability to the disabled, elderly and children by developing the vehicle with folding steps or by building the high boarding platforms. In this research, I concentrated on the user oriented interior design of the FRT to make it passenger-friendly and safe in order to maximize the utilization of the vehicle so that all users including wheelchaired, user with baby carriage, elderly and children can conveniently use this vehicle.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.21 no.2
• /
• pp.136-145
• /
• 2013

It is very important to determine specifications of components included in the drive-train of vehicles at the initial design stage. In this study, component sizing process and performance analysis for Extended-Range Electric Vehicles (E-REV) are discussed based on the foundation of determined system configuration and performance target. This process shows sizing results of an electric driving motor, a final drive gear ratio and a battery capacity for target performance including All Electric Range (AER) limit. For E-REV driving mode, the constant output power of a Gen-set (Engine+Generator) is analyzed in order to sustain State of Charge (SOC) of the battery system.

• The Journal of Korea Robotics Society
• /
• v.18 no.3
• /
• pp.346-351
• /
• 2023

With the release of numerous open driving datasets, the demand for domain adaptation in perception tasks has increased, particularly when transferring knowledge from rich datasets to novel domains. However, it is difficult to solve the change 1) in the sensor domain caused by heterogeneous LiDAR sensors and 2) in the environmental domain caused by different environmental factors. We overcome domain differences in the semi-supervised setting with 3-stage model parameter training. First, we pre-train the model with the source dataset with object scaling based on statistics of the object size. Then we fine-tine the partially frozen model weights with copy-and-paste augmentation. The 3D points in the box labels are copied from one scene and pasted to the other scenes. Finally, we use the knowledge distillation method to update the student network with a moving average from the teacher network along with a self-training method with pseudo labels. Test-Time Augmentation with varying z values is employed to predict the final results. Our method achieved 3rd place in ECCV 2022 workshop on the 3D Perception for Autonomous Driving challenge.