• Journal of Korea Multimedia Society
• /
• v.23 no.11
• /
• pp.1349-1360
• /
• 2020

In recent years, Convolutional Neural Networks (CNNs) have been successfully implemented in different tasks of computer vision. Since CNN models are the representatives of supervised learning algorithms, they demand large amount of data in order to train the classifiers. Thus, obtaining data with correct labels is imperative to attain the state-of-the-art performance of the CNN models. However, labelling datasets is quite tedious and expensive process, therefore real-life datasets often exhibit incorrect labels. Although the issue of poorly labelled datasets has been studied before, we have noticed that the methods are very complex and hard to reproduce. Therefore, in this research work, we propose Deep CleanNet - a considerably simple system that achieves competitive results when compared to the existing methods. We use K-means clustering algorithm for selecting data with correct labels and train the new dataset using a deep CNN model. The technique achieves competitive results in both training and validation stages. We conducted experiments using MNIST database of handwritten digits with 50% corrupted labels and achieved up to 10 and 20% increase in training and validation sets accuracy scores, respectively.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.14 no.10
• /
• pp.4665-4671
• /
• 2013

Engine mounting system is the most responsible system for NVH performance of vehicle. The vibration at idle shake, road shake, Key ON/OFF, gear shift tuned by the engine mount position and stiffness. Previously described Engine mounting system theory investigated and summarized in this paper. Decoupling of the Power train rigid mode and Reducing the angle between Torque-Roll-Axis and Elastic-roll-Axis is starting point of optimization. Multi-optimization analysis was performed because of variety simulation case and FE-model. Eventually, Find the best mount location and the stiffness has improved the performance of the vehicle NVH.

• Tribology and Lubricants
• /
• v.36 no.2
• /
• pp.96-104
• /
• 2020

The electric vehicle industry is rapidly developing because of enforced environmental regulations, and several studies have been conducted on the multispeed transmission to improve the fuel efficiency of electric vehicles. Among these studies, research on the power density improvement of electric vehicle transmission is critical. Thus, the optimal design of the gear train is necessary to enhance transmission efficiency. In this study, an optimal design methodology for the lightweight two-speed transmission of electric vehicles is proposed. Because a multispeed transmission has many operating conditions and equality and inequality constraints, a new gear design method that combines analytical and iterative methods is applied without using complex optimization algorithms. Sets of possible design variables are generated considering the operating conditions and various design variables. The modules and face width ratios of each stage gear that satisfy the corresponding operating conditions are analytically calculated. The volume of the gear train is calculated, evaluated, and arranged using these values to determine the optimal solution for minimizing the volume, and the proposed methodology is applied to the actual model to verify its effectiveness. The design of a two-speed transmission with multiple operating conditions and constraints without complicated optimization algorithms can be optimized.

• Proceedings of the KIEE Conference
• /
• 2004.07d
• /
• pp.2631-2633
• /
• 2004

This paper describes the construction of a half sphere screen driving tilting simulator that can perform six degree-of-freedom( DOF) motions simmulator to a tilting train. The mathematical equations of Tilting Train dynamics are first derived from the 6-DOF bicycle model and incorporated with the bogie. carbody, and suspension subsystems. The equations of motion are then programmed by visual C++ code. To achieve the simulator functions. a motion platform that is constructed by six electric-driven actuators is designed. and its kinetics/inverse kinetics analysis is also conducted. Driver operation signals such as carbady angle, accelerator. and tilting positions are measured to trigger the Tilting dynamics calculation and further actuate the cylinders by the motion platform control program. In addition. a digital PID controller is added to achieve the stable and accurate displacements of the motion platform. The experiments prove that the designed simulator is adequate in performing some special rail mad driving situations discussed in this paper.

• Proceedings of the KSR Conference
• /
• 2010.06a
• /
• pp.1506-1512
• /
• 2010

In this study, model tests were performed to investigate the settlement characteristics of the rock/soil mixture subgrade with the many portion of mudstone due to the cyclic train loading in high-speed railway. Specially, from the tests varying initial degree of saturation, effects of increment of the degree of saturation in the subgrade due to rainfall or elevation of ground water table under cyclic train loading on the deformation characteristics were analyzed. From the results, in the low degree of saturation, settlement converged to some value. However, in the condition of degree of saturation larger than some value, settlement rapidly increased. Therefore, it was found that it is important to maintain the degree of saturation of subgrade below the specific level to prevent the settlement of subgrade.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.24 no.3
• /
• pp.253-260
• /
• 2014

conducted. In-situ measurements were conducted to investigate the noise characteristics in high-speed trains such as KTX and KTX-Sancheon. Binaural microphones and spherical microphone were adopted to record and evaluate the spectral characteristics and noise incidence characteristics at the seats in the passenger cars. Environmental elements were also recorded during the noise measurements using a camcorder and GPS application. In addition, auditory tests were conducted to investigate annoyance of interior noise of high-speed trains. Based on the objective measure and subjective responses on interior noise, statistical analysis such as correlation and regression analyses were carried out. From the results perception models for evaluating the interior noise of high-speed train were derived.

• Journal of Advanced Marine Engineering and Technology
• /
• v.34 no.8
• /
• pp.1057-1067
• /
• 2010

Vibration and noise analysis as well as strength of gear teeth, gear profile design are considered in order to develop the multi-gear train devices of synchro system for the guns of a warship. A new approach to the critical speed calculation of practical industrial multi-mesh geared system is presented. A transfer matrix model based on Hibner's branch method is developed and the natural properties of the branched rotor system are calculated with using the ${\lambda}$-matrix formulation. A Campbell diagram, in which the excitation sources caused by the mass unbalance of the rotors and the transmitted errors of the gearing are considered, shows that, at theoperating speed, there are not the critical speed.

• Structural Engineering and Mechanics
• /
• v.75 no.2
• /
• pp.271-282
• /
• 2020

The preservation of historic masonry-arch railway bridges is of paramount importance due to their economic benefits. These bridges which belong to past centuries may nowadays be expected to carry loads higher than those for which they were designed. Such an increase in loads may be because of increase in transportation speed or in the capacity of freight-wagons. Anyway, adequate increase in their load-carrying-capacity through structural-strengthening is required. Moreover, the increasing costs of material/construction urge engineers to optimize their designs to obtain the minimum-cost one. This paper proposes a novel numerical optimization method to minimize the costs associated with strengthening of masonry-arch railway bridges. To do so, the stress/displacement responses of Sahand-Goltappeh bridge are assessed under ordinary train pass as a case study. For this aim, 3D-Finite-Element-Model is created and calibrated using experimental test results. Then, it is strengthened such that following goals are achieved simultaneously: (1) the load-carrying-capacity of the bridge is increased; (2) the structural response of the bridge is reduced to a certain limit; and, (3) the costs needed for such strengthening are minimized as far as possible. The results of the case study demonstrate the applicability/superiority of the proposed approach. Some economic measures are also recommended to further reduce the total strengthening cost.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.10 no.5
• /
• pp.742-750
• /
• 1986

The lateral vibrating motion of a railway vehicle over a certain critical speed is a well known problem in the field of train dynamics. It is known that the train equations of motion are strongly coupled and highly nonlinear with the motion and causing that it is very difficult to solve the equations simultaneously. In this paper, a 8 degree of feedom model of a railway vehicle was suggested to solve the rail vehicle lateral motion. In stead of solving the nonlinear equation simultaneously, statistical linearization technique was adopted to solve those equations. The analysis results from the statistical linearization method were directly compared with those from direct nonlinear equations and found that the linearization technique can be very effective and economical for railroad vehicle analysis. By the way, it was found that the analysis results can analytically explain the intermittent hunting phenomena which has been frequently observed in experiments.

• The Proceeding of the Korean Institute of Electromagnetic Engineering and Science
• /
• v.6 no.7
• /
• pp.37-47
• /
• 1995

In this paper, we have measured eletromagnetic noises from eletcric trains and the measurment data are treated statisticlly for monig. In order to measure the noise of electric tran, we set up an automatic measuring system andd measured the magnetic field over 9 kHz ~ 30 MHz range and electric field over 30 MHz - 1, 000 MHz range. The computer controlled measurint system yields efficiently experimental APD (Amplitude Probability distribution) data each of national rail road and subway train. The measured APD curves are analysed in terms of sensitivity study of Middleton's model through 6-parameter variation. Optimal parameters are obtained from measured data using Composite Approximation Algorithm.