• 한국공작기계학회:학술대회논문집
• /
• 한국공작기계학회 2000년도 추계학술대회논문집 - 한국공작기계학회
• /
• pp.466-470
• /
• 2000

Externally pressurized air journal bearing has been widely used in high-speed spindle system and precision machinery because of its characteristics such as substantially low frictional loss, low heat generation and averaging effect leading better running accuracy. But air journal bearing have a disadvantage of the low load capacity due to the low viscosity. In this paper, The air journal bearing design to overcome the defects of air bearing such as low stiffness and dimping coefficients was investigated theoretically.

• 한국항해학회지
• /
• 제6권2호
• /
• pp.1-12
• /
• 1982

The tedious work, connected to the traditional computation of altitudes and azimuths and the plotting of the position lines, has been a severe objection to celestial fixes. But recently computers have become to be used generally for computations of altitudes and aximuths and the computing objection seems to be practically overruled. Now it seems appropriate to concentrate on other problems which are the procedure of improving accuracy of ship's position and the design of a general computing procedure to determine the coordinates of the optimally estimated ship's position. In this paper, such procedures as an application of Kalman filter and the results of the Digital simulation conducted under various noise conditiions are presented. The positions estimated by Kalman filter are compared with the running fixes and the most probable positions obtained from a single position line, and it is confirmed that the resutls of the proposed method is evidently accurate than others.

• 대한임베디드공학회논문지
• /
• 제16권5호
• /
• pp.179-186
• /
• 2021

For stereo matching based on deep learning, the design of network structure is crucial to the calculation of matching cost, and the time-consuming problem of convolutional neural network in image processing also needs to be solved urgently. In this paper, a method of stereo matching using sparse loss volume in parallax dimension is proposed. A sparse 3D loss volume is constructed by using a wide step length translation of the right view feature map, which reduces the video memory and computing resources required by the 3D convolution module by several times. In order to improve the accuracy of the algorithm, the nonlinear up-sampling of the matching loss in the parallax dimension is carried out by using the method of multi-category output, and the training model is combined with two kinds of loss functions. Compared with the benchmark algorithm, the proposed algorithm not only improves the accuracy but also shortens the running time by about 30%.

• 천문학회보
• /
• 제44권1호
• /
• pp.46.1-46.1
• /
• 2019

We present an accurate and efficient method to calculate the gravitational potential of an isolated system in three-dimensional Cartesian and cylindrical coordinates subject to vacuum (open) boundary conditions. Our method consists of two parts: an interior solver and a boundary solver. The interior solver adopts an eigenfunction expansion method together with a tridiagonal matrix solver to solve the Poisson equation subject to the zero boundary condition. The boundary solver employs James's method to calculate the boundary potential due to the screening charges required to keep the zero boundary condition for the interior solver. A full computation of gravitational potential requires running the interior solver twice and the boundary solver once. We develop a method to compute the discrete Green's function in cylindrical coordinates, which is an integral part of the James algorithm to maintain second-order accuracy. We implement our method in the {\tt Athena++} magnetohydrodynamics code, and perform various tests to check that our solver is second-order accurate and exhibits good parallel performance.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1999년도 제14차 학술회의논문집
• /
• pp.244-247
• /
• 1999

This paper presents a methodology of the technique for controlling DC motor drive by implementation of 2-quadrant operating mode which can ensure the torque controlling and speed with response time less than 2 seconds at all loading conditions. By implementation of BRM technique, energy is fed with definite values of BRM 256 bits, with different patterns of high accuracy, and fixing scan time at 0.667 ms, the ripple is less than 1%, thus high efficiency can be achieved, from the consequence of the accuracy of energy feeding at low current. The stability of the whole system can be determined from circle criterion by root locus method . The instant reverse direction of rotation can be done by decreasing the energy to the lowest level while motor is running with no load and variable load at the speed about 100-120 rpm and 50-60 rpm.

• International Journal of Precision Engineering and Manufacturing
• /
• 제7권3호
• /
• pp.3-7
• /
• 2006

For the purpose to improve and to automate designing of high-speed spindle units (SU's), we have developed the mathematical models and software to estimate SU performance characteristics, including the run-out of spindles running on ball bearings. In order to understand better the mechanics of high-speed SUs, the dynamic interaction of ball bearings and SU, and the influence of the bearing imperfections and SU's operational conditions on the run-out, we have carried out computer simulation and experimental studies. Through the study of SU's, we have found out that run-out of SU can vary drastically with variation of rpm. The influences of rotation speed and of accuracy parameters of bearings on the SU accuracy have the greatest importance. The influence of bearing preload has a secondary importance. Comparison of the results of these studies has demonstrated adequacy of the models and software developed to the real SU's.

• Journal of Electrical Engineering and Technology
• /
• 제11권3호
• /
• pp.769-774
• /
• 2016

In the design of sensorless control system for induction motor, high-precision speed estimation is one of the most difficult problems. To solve this problem, the common method is model reference adaptive method (MRAS). MRAS requires accurate motor parameters to estimate rotor speed precisely. However, when motor is running, the variety of temperature and magnetic saturation will lead to the change of motor parameters such as stator resistance and rotor resistance, which will lower the accuracy of the speed estimation. To improve the accuracy and rapidity of speed estimation, this paper analyses the mutual MRAS speed identification based on rotor flux linkage, and proposes an improved mutual MRAS speed identification based on back-EMF. The improved method is verified by Simulink simulation and motor experimental platform based on DSP2812. The results of simulation and experiment indicate that the method proposed by this paper can significantly improve the accuracy of speed identification, and speed up the response of identification.

• 대한임베디드공학회논문지
• /
• 제16권1호
• /
• pp.9-15
• /
• 2021

Recently, it is increasingly necessary to run high-end neural network applications with huge computation overhead on top of resource-constrained embedded systems, such as wearable devices. While the huge computational overhead can be alleviated by distributed neural networks running on multiple separate devices, existing distributed neural network techniques suffer from a large traffic between the devices; thus are very vulnerable to communication failures. These drawbacks make the distributed neural network techniques inapplicable to wearable devices, which are connected with each other through unstable and low data rate communication medium like human body communication. Therefore, in this paper, we propose a distributed neural network partitioning technique that is resilient to communication failures. Furthermore, we show that the proposed technique also improves the inference accuracy even in case of no communication failure, thanks to the improved network partitioning. We verify through comparative experiments with a real-life neural network application that the proposed technique outperforms the existing state-of-the-art distributed neural network technique in terms of accuracy and resiliency to communication failures.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2023년도 학술발표회
• /
• pp.175-175
• /
• 2023

Detection and monitoring of wildfires are essential for limiting their harmful effects on ecosystems, human lives, and property. In this research, we propose a novel method running in the Google Earth Engine platform for identifying and characterizing burnt regions using a hybrid of Sentinel-1 (C-band synthetic aperture radar) and Sentinel-2 (multispectral photography) images. The 2022 Uljin wildfire, the severest event in South Korean history, is the primary area of our investigation. Given its documented success in remote sensing and land cover categorization applications, we select the Random Forest (RF) method as our primary classifier. Next, we evaluate the performance of our model using multiple accuracy measures, including overall accuracy (OA), Kappa coefficient, and area under the curve (AUC). The proposed method shows the accuracy and resilience of wildfire identification compared to traditional methods that depend on survey data. These results have significant implications for the development of efficient and dependable wildfire monitoring systems and add to our knowledge of how machine learning and remote sensing-based approaches may be combined to improve environmental monitoring and management applications.

• 대한의용생체공학회:의공학회지
• /
• 제27권6호
• /
• pp.301-309
• /
• 2006

The main purpose of this study is to derive a regression equation that predicts the individual differences in activity energy expenditure (AEE) using accelerometer during different types of activity. Two subject groups were recruited separately in time: One is a homogeneous group of 94 healthy young adults with age ranged from $20\sim35$ yrs. The other subject group has a broad spectrum of physical characteristics in terms of age and fat ratio. 226 adolescents and adults of age ranged from $12\sim57$ yrs and fat ratio from $4.1\sim39.7%$ were in the second group. The wireless 3-axis accelerometers were developed and carefully fixed at the waist belt level. Simultaneously the total calorie expenditure was measured by gas analyzer. Each subject performed walking and running at speeds of 1.5, 3.0, 4.5, 6.0, 6.5, 7.5, and 8.5 km/hr. A generalized sensor-independent regression equation for AEE was derived. The regression equation was developed fur walking and running. The regression coefficients were predicted as functions of physical factors-age, gender, height, and weight with multivariable regression analysis. The generalized calorie estimation equation predicts AEE with correlation coefficient of 0.96 and the average accuracy of the accumulated calorie was $89.6{\pm}7.9%$.