• Journal of KIISE:Software and Applications
• /
• v.30 no.11
• /
• pp.1054-1061
• /
• 2003

Q-Learning, a representative algorithm of reinforcement learning, experiences repeatedly until estimation values about all state-action pairs of state space converge and achieve optimal policies. When the state space is high dimensional or continuous, complex reinforcement learning tasks involve very large state space and suffer from storing all individual state values in a single table. We introduce Q-Map that is new function approximation method to get classified policies. As an agent learns on-line, Q-Map groups states of similar situations and adapts to new experiences repeatedly. State-action pairs necessary for fine control are treated in the form of rule. As a result of experiment in maze environment and mountain car problem, we can achieve classified knowledge and extract easily rules from Q-Map

• Korean Journal of Ecology and Environment
• /
• v.40 no.3
• /
• pp.370-378
• /
• 2007

The objective of this research was to diagnose integrative ecological health in Bansuk Stream, one of the tributaries of Gap Stream, using the fish assemblage during July 2006${\sim}$April 2006. For this research, we selected six sampling sites and applied some approaches such as the Index of Biological Integrity (IBI), Qualitative Habitat Evaluation Index (QHEI), and necropsy-based Health Assessment Index (HAI). The stream health condition, based on the IBI values, averaged 24 (n= 18, range: $10{\sim}46$), indicating "poor${\sim}$fair" condition according to the criteria of US EPA (1993). Physical habitat condition, based on the QHEI, averaged 116 (n=6, range: $77{\sim}139$), indicating "fair${\sim}$good" condition. Values of IBI were more correlated with 3 metrics of instream cover ($M_1$, r=0.553, p=0.017, n=18), flow/velocity ($M_3$, r=0.627, p=0.005, n=18), and riffes/bends ($M_7$, r=0.631, p=0.005, n=18) than other metrics. Value of HAI in the control was zero (i.e., excellent condition), while the values in the T1 and T2 treatments were 5 (range: 0${\sim}$30) and 50 (range: 40${\sim}$80), respectively. The maximum values of IBI (46) were coincided with zero of HAI. Thus, these approaches seem to be a good tool for a diagnosis and evaluations of stream ecosystem health.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.26 no.1
• /
• pp.104-112
• /
• 1989

A design automation method for sequential circuits implementation by mans of PLA is discussed, and an optimal state-code assignment algorithm to minimize the PLA area is proposed. In order to design sequential circuit automatically, DASL (Design Automation Support Language) [8] which is easy to describe and powerful to synthesize, is proposed and used to describe sequential circuit, An optimal statecode assignment algorithm which considers next states and outputs simultaneously is proposed, and by adopting this algorithm to various examples, the area of PLA is reduced by 10% comparing privious methods. This system is constructed to design microinstruction, FSM, VLSI control part synthesis.

• Proceedings of the Korean Institute Of Construction Engineering and Management
• /
• 2008.11a
• /
• pp.149-154
• /
• 2008

Service life of bridges should be evaluated by physical life considering damage/deterioration. But it is difficult to identify optimal maintenance scenario due to insufficient research related to that. To identify optimal maintenance scenario, it is needed to develope life cycle profile model of condition state variation by deterioration factor. The LCP model has been developed in consideration of regression analysis and survey in this study. It is expected that the LCP model could help to achieve HBMS system improvement.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.6 no.3
• /
• pp.51-62
• /
• 1996

Electro sensitive traffic system can not consider passenger car unit, so it causes start up delay time and passenger waiting time. In this paper, it antecedently creates passenger car unit at the bottom intersection using neural network. But, sometimes it can make mistakes due to changes in car weight, car speed, and passing area. Therefore, it consequently reduces the car waiting time and start-up delay time using fuzzy control of feed-back data. Moreover, to prevent spillback, it can adapt control even though upper traffic intersection has a different saturation rate, road length, road slope and road width.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.1 no.3
• /
• pp.131-140
• /
• 1977

전기유압식 액튜에이터의 시간영역내에서의 성능을 향상시키기 위해서 최적제어이론을 적용하였다. 계통의 정량적인 성능을 이차성능지수로 표시햐였다. 유압 액튜에이터와 밸브의 선형적인 전달특성을 이용해서 계통의 상태방정식을 세우고 리카티 방정식의 해를 컴퓨터로 구해서 계통의 최적입력을 결정하였다. 이상과 같이 만들어진 최적제어계통의 변위, 속도, 가속도의 과도응답을 구하기 위해서 아날로그 컴퓨터를 사용하고 그 응답과 P.W.M. 계통의 응답을 비교한 결과, 최적제어계통이 더욱 빠르고 안정된 응답을 나타냄을 알았다. 그 비교를 구체적이고 정량적으로 행하기 위해서 성능지수곡선을 구해서 비교한 결과 그 성능지수의 척도로 볼때 최적제어계통이 P.W.M. 계통보다 약 35% 까지 우수하다는 결론을 얻었다.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.15 no.1
• /
• pp.155-164
• /
• 2002

The journal bearings use in machine parts which move relative to each other and those reduce friction and wear of journals. The journal bearings are designed to operate in the hydyodynamic lubrication regime, but elastohydrodynamic lubrication nay occur if the pressures are too high or the running speeds are too low at machine elements. It is the phenomenon that the lubricant film is broken and some parts of surfaces are in rolling contact, so that wear will increase in mixed lubrication regime. The purpose of this study is to minimize the wear rate of journal bearings for extending machine life. The wear mate in mixed lubricated regime is selected as objective function because most of wear of the journal bearings develops in elastohydrodynamic lubrication. The journal bearings we represented by a bearing radius, shaft radius, and bearing width, but the bearing radius only is selected as design variables due to a bearing radius has an influence on friction loss, stability limit velocity, and film parameter, which are used as constraints. For numerical calculation, PLBA, that is a class of the RQP algorithm, is used.

• Proceedings of the Korean Institute of Communication Sciences Conference
• /
• 1990.06a
• /
• pp.563-566
• /
• 1990

• Journal of the Earthquake Engineering Society of Korea
• /
• v.5 no.6
• /
• pp.37-46
• /
• 2001

In this research, a controller design method based on optimization is proposed that can satisfy constraints on maximum responses of building structures subject to around excitation modeled by partially stochastic process. The class of controllers to be optimized is restricted to LQR. Weighting matrix on controlled outputs is used as design variable. Objective function, constraint functions and their gradients are computed by the parameterization of control gain with Riccati matrix. Full state feedback controllers designed by proposed optimization method satisfy various design objectives and their necessary maximum control forces are computed for the production of actuator. LQG controllers composed of Kalman filter and LQR designed by proposed method perform well with little deterioration. So it is possible to design output feedback controllers satisfying constraints on various maximum responses of structures.

• Journal of the KSME
• /
• v.32 no.11
• /
• pp.917-926
• /
• 1992

트라이볼로지의 핵심분야인 마찰, 마멸에 대해 간단히 살펴보았다. 과학의 급속한 발전에도 불 구하고 아직도 이들에 대한 완전한 지식은 아직 요원한 상태이다. 그러나 현재까지 알려진 지 식으로도 기계시스템의 최적설계에 크게 도움이 된다는 사실에 대해서는 의심의 여지가 없다. 앞에서 함께 살펴본 마멸 입자에 대한 소개는 마찰과 마멸의 복합적인 산물로서 이의 관찰을 통해 분석하고자 하는 기계시스템의 상태를 효과적으로 진단하여 트라이볼로지적 거동을 이해 하는 데에 중대한 정보가 된다. 따라서 마찰과 마멸에 대한 궁극적인 연구와 함께 그 산물인 마멸입자에 대한 보다 적극적인 이해와 지식의 축적은 요소(tribo-elements)의 고성능, 고정밀 화에 크게 이바지할 것이다.