• 한국항공우주학회지
• /
• 제38권8호
• /
• pp.790-797
• /
• 2010

마하수 1.0~1.2 범위의 초음속 제트에서 발생되는 축대칭 스크리치 톤 소음의 모드 전이 현상이 수치적으로 분석되었다. 이를 위해 k-e 난류모델을 가진 축대칭 Navier-Stokes 방정식이 사용되었으며 수치기법으로서 공간에 대해 소산관계보존기법과 시간에 대해 최적화된 4단계 시간 적분법이 사용되었다. 특히 낮은 마하수에서 발생하는 축대칭 A1 모드의 경우, 비선형성에 기인한 부가적인 음향파 발생을 확인하였으며, 강한 와류의 발생으로 인해 스크리치 소음 주파수는 와류 통과 주파수와 일치함을 알게 되었다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권4호
• /
• pp.1768-1789
• /
• 2016

In contemporary video streaming systems based on HTTP protocol, video players at the client side are responsible for adjusting video quality to network conditions and user expectations. However, when multiple video clips are streamed simultaneously, an intricate application logic implemented in the video players overlays the TCP mechanism which is responsible for a balanced access to a shared network link. As a result, some video players may not obtain a fair share of network throughput and may be vulnerable to an unstable video bit-rate. Therefore, we propose to simplify the algorithms implemented in the video players, which are responsible for the adjustment of video quality and constrain their functionality only to sending feedback to a server about a state of the player buffer. The main logic of the system is shifted to the server, which is now responsible for bit-rate selection and prioritisation of the video streams transmitted to multiple clients. To verify our proposition, we performed several experiments in a laboratory environment which show that when the server cooperates with the clients, the video players experience fewer quality switches and the system achieves better fairness when allocating network throughput among the video players. However, this comes at the cost of worse utilisation of network bandwidth.

• 한국시스템다이내믹스연구
• /
• 제7권1호
• /
• pp.5-26
• /
• 2006

Ever increasing dependence of organization on information technology stimulates interactions between individuals and groups in the process of knowledge creation, which overall impies that a reciprocal mechanism lies within the structure of the growth of group intelligence. Individual's intelligence may affect the group intelligence, and vise versa. However, the level of group intelligence is not necessarily determined by the sum of individual's intelligence but the quality of the interactions among the individuals. This study thus aims to conceptually identify the dynamic structure of interactions among the factors influencing the group intelligence level, which is believed to be used as a tool to measure the difference of intelligence between groups. To achieve this goal several attempts were made. First, determinants of intelligence at indiviual level and group level and similarities and differences between individual's and group intelligence were identified from the previous research. Second, causal loop diagrams were developed, which show how individual's intelligence influences group intelligence and vise versa. Third, it was attempted to identify and interpret which feedback loops are most influential in either improving or hapering group intelligence as a whole. Since this study remains only at exploratory level, a more detailed and workable model for field applications has to be developed in the future.

• 제어로봇시스템학회논문지
• /
• 제20권10호
• /
• pp.1008-1013
• /
• 2014

This paper introduces an adaptive anti-sway tracking control algorithm for container cranes with unknown payloads and friction between the trolley and the rail. If the friction effects in the system can be modeled, there is an improved potential to design controllers that can cancel these effects. The proposed control improves the sway suppressing and the positioning capabilities of the trolley and hoisting against uncertain payload and friction. The variable structure controls are first designed based on a class of feedback linearization methods for the stabilization of the under-actuated sway dynamics. The adaptation mechanism are then designed with parameter estimation of unknown payload and friction compensation for the trolley and hoisting, based on Lyapunov stability methods for the accurate positioning and fast attenuation of trolley oscillation due to frictions in the vicinity of the target position. The asymptotic stability of the overall closed-loop system is assured irrespective of variations of rope length. Simulations are shown under various frictions and external winds in the case of no priori information of payload mass.

• 전자공학회논문지S
• /
• 제35S권4호
• /
• pp.67-75
• /
• 1998

In general, Evoluationary Algorithm(EAs) are refered to as methods of population-based optimization. And EAs are considered as very efficient methods of optimal sytem design because they can provice much opportunity for obtaining the global optimal solution. This paper presents a co-evolution scheme of artifical neural networks, which has two different, still cooperatively working, populations, called as a host popuation and a parasite population, respectively. Using the conventional generatic algorithm the host population is evolved in the given environment, and the parastie population composed of schemata is evolved to find useful schema for the host population. the structure of artificial neural network is a diagonal recurrent neural netork which has self-feedback loops only in its hidden nodes. To find optimal neural networks we should take into account the structure of the neural network as well as the adaptive parameters, weight of neurons. So we use the genetic algorithm that searches the structure of the neural network by the co-evolution mechanism, and for the weights learning we adopted the evolutionary stategies. As a results of co-evolution we will find the optimal structure of the neural network in a short time with a small population. The validity and effectiveness of the proposed method are inspected by applying it to the stabilization and position control of the invered-pendulum system. And we will show that the result of co-evolution is better than that of the conventioal genetic algorithm.

• 한국통신학회논문지
• /
• 제22권9호
• /
• pp.2027-2034
• /
• 1997

프레임 릴레이는 망에서의 오류제어와 흐름제어 기능을 제거하여, TI 또는 EI의 속도로 원거리에서 다양한 길이를 갖는 프레임을 전송하고 다중화하는 고속 패킷 교환 기술이다. 그리고 오류제어와 흐름제어의 생략의 결과로 발생하는 망에서의 혼잡을 피하고 복구하기 위해서는 일반적으로 몇가지의 혼잡제어 방식이 쓰이는데, 본 연구에서는 프레임 릴레이망에서 DE비트를 사용하여 전송률을 강제적으로 조정하는 혼잡제어 방식에 대해서 수치적으로 성능해석을 수행하였다. 수치적인 해석을 위하여 각각의 프레임이 Poisson 분포를 따르며 지수분포 길이로 도착한다고 가정하고 시스템의 크기가 K인 경우, 프레임 릴레이 교환기를 하나의 서버와 K-1(1

• Journal of Communications and Networks
• /
• 제13권1호
• /
• pp.17-25
• /
• 2011

A convergecast is a popular routing scheme in wireless sensor networks (WSNs) in which every sensor node periodically forwards measured data along configured routing paths to a base station (BS). Prolonging lifetimes in energy-limited WSNs is an important issue because the lifetime of a WSN influences on its quality and price. Low-energy adaptive clustering hierarchy (LEACH) was the first attempt at solving this lifetime problem in convergecast WSNs, and it was followed by other solutions including power efficient gathering in sensor information systems (PEGASIS) and power efficient data gathering and aggregation protocol (PEDAP). Our solution-chain routing with even energy consumption (CREEC)-solves this problem by achieving longer average lifetimes using two strategies: i) Maximizing the fairness of energy distribution at every sensor node and ii) running a feedback mechanism that utilizes a preliminary simulation of energy consumption to save energy for depleted Sensor nodes. Simulation results confirm that CREEC outperforms all previous solutions such as LEACH, PEGASIS, PEDAP, and PEDAP-power aware (PA) with respect to the first node death and the average lifetime. CREEC performs very well at all WSN sizes, BS distances and battery capacities with an increased convergecast delay.

• 한국생산제조학회지
• /
• 제21권1호
• /
• pp.161-166
• /
• 2012

Intravenous (IV) injection is widely used to supply Ringer solution directly into a vein in hospital. Generally, a passive injection method has been used, which causes the inconsistent flow rate of fluid and inappropriate control of injection time by a patient. It leads to an unnecessary nurse's overwork and decrement of IV injection's effect. To solve these problems, flow control infusion pumps have been developed. But because of relatively heavy weight and high price, its usage has been limited. In the present study, a new automatic IV injection system is developed. It is installed with a small pressing mechanism driven by a small electric motor to regulate the flow rate by pressing tube. Proportional integral derivative (PID) feedback control algorithm is applied to control the electric motor. The system is smaller in size and uses lower power than the existing commercial product. The newly developed system is also installed with networking capability, which enables monitoring the status of several automatic IV injection system at the same time.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2008년도 추계학술대회B
• /
• pp.3175-3180
• /
• 2008

A metal-fueled pool-type liquid metal fast reactor (LMFR) provides large margins to sodium boiling and fuel damage under accident conditions. The favorable passive safety results are obtained by both a reactivity feedback mechanism in the core and a passive decay heat removal system. Among the various reactivity feedbacks, the ones by a thermal expansion of a radial dimension of the core and by the control rod drivelines are strongly dependent on the flow conditions in the core and the hot pool, respectively. The effects of multidimensional thermal hydraulic characteristics on these reactivity feedbacks are investigated by the system-wide safety analysis code SSC-K with advanced thermal hydraulics models. Particularly a detailed three dimensional thermal hydraulics reactor core model is integrated into SSC-K for use in a whole system analysis of the passive safety aspects of LMR designs. The model provides fuel and cladding temperatures for every fuel pin in a reactor and coolant temperatures for every coolant sub-channel in the reactor.

• 산업융합연구
• /
• 제18권5호
• /
• pp.102-107
• /
• 2020

본 연구의 목적은 대학 교과과정 내에서 리빙랩을 활용하여 현장 맞춤형 교육 프로그램 개발과 평가를 수행하는 것이다. 개발 과정은 리빙랩의 작동 메커니즘에 따라 아이디어 발현, 의제 설정, 실제 활동, 피드백의 단계로 수행되었다. 개발한 유아교육 프로그램은 4가지 주제, 16가지 소주제, 53개의 활동으로 구성되었다. 프로그램에 대한 평가는 부산 소재 유아교육기관에 근무하는 교사 68명을 대상으로 만족도 조사를 실시하였다. 그 결과, 모든 문항에서 대부분 80%이상 만족스럽다는 반응을 나타냈다. 본 연구는 대학 교과에 리빙랩을 활용한 연구로 이후 대학의 다양한 전공 및 여러 교과에서 리빙랩을 적용하는데 기초자료로 의의가 있다.