http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
실험계획법을 기반으로 한 유연 힌지 구조의 강건한 반사경 마운트 최적 설계 연구
남성식(Sungsik Nam),남병욱(Byounguk Nam) 한국기계가공학회 2024 한국기계가공학회지 Vol.23 No.4
In this study, we present an optimal mount design that can maintain the surface accuracy of the reflector even under the influence of gravity, thermal, and fastening loads. The reflector is a key component system in which performance is determined by the trade-off between stiffness and wavefront error. The newly designed reflector mount incorporates flexibility in the radial direction through a flexible hinge, while ensuring rigidity to maintain surface accuracy without transmitting thermal and mechanical loads to the remaining degrees of freedom. The key design variables for the reflector mount are the thickness and length of the flexible hinge structure and robust structure to withstand assembly load. Through analysis, the system’s requirements were first confirmed for high natural frequencies in low modes. Additionally, wavefront errors capable of stable transmission of high energy under extreme temperature conditions were analyzed. The reflector mount design, satisfying high natural frequencies and wavefront errors, involved interpreting the fastening forces generated when applied to the system as enforced displacement values. And a reflector mount was produced and verified by a wavefront error experiment. Through those analysis and experiment, an optimal design for the reflector mount was derived.
소형무장헬기 회전형 포탑시스템의 안정화추적 성능 시험기법
남병욱,이호정,조시훈,조성훈,Nam, Byounguk,Lee, Hojung,Jo, Sihun,Cho, Sunghun 한국군사과학기술학회 2014 한국군사과학기술학회지 Vol.17 No.6
In this paper, we propose a novel method which can be applied to test and verify the stabilization-tracking performance of Gun/Turret Driving System(GTDS) for Light Armed Helicopter(LAH). In real system, GTDS is connected to TADS/SMC and drived to aim at the target with 20mm gun. But each equipment is separately developed during exploratory development stage, so GTDS cannot be tested under real system state. We suggest new configuration of test system for evaluating the stabilization-tacking performance, in which TADS and SMC are replaced by vision acquisition unit and processing unit, respectively.
반응표면법을 이용한 광학미러용 일체형 유연힌지 마운트 최적설계
이경호,남병욱,남성식,Kyoungho Lee,Byounguk Nam,Sungsik Nam 한국군사과학기술학회 2023 한국군사과학기술학회지 Vol.26 No.3
An optimal design of a simple beam-shaped flexure hinge mount supporting an optical mirror is presented. An optical mirror assembly is an opto-mechanically coupled system as the optical and mechanical behaviors interact. This side-supporting mount is flexible in the radial direction and rigid for the remaining degrees of freedom to support the mirror without transferring thermal load. Through thermo-elastic, optical and eigenvalue analysis, opto-mechanical performance was predicted to establish the objective functions for optimization. The key design parameters for this flexure are the thickness and length. To find the optimal values of design parameters, response surface analysis was performed using the design of experiment based on nested FCD. Optimal design candidates were derived from the response surface analysis, and the optimal design shape was confirmed through Opto-mechanical performance validation analysis.
An Intensity Controlled Review Dataset Construction for Automatic Review Generation
Nagyeong Kim,Hyejin Jo,Jueon Lee,Jaeho Choi,Byounguk Nam,Yuchul Jung 한국정보기술학회 2022 한국정보기술학회논문지 Vol.20 No.2
Most users refer to existing online reviews to see if other users were previously satisfied or not about the products (or foods) they want to buy. Meanwhile, some users do not want to write fruitful, realistic reviews because it is annoying and bothering. To minimize users’ writing costs, we are interested in implementing an automatic review generator to create a complete review with only scores and a few seed keywords by enabling intensity control tailored to the user’s needs. To this end, we propose an intensity controlled review data construction method for online review. Moreover, we employ GPT-2 and BART models popular for text generation tasks for the review generation experiments. Our automatic and manual evaluations for randomly sampled generation results prove the quality of our constructed dataset.