• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2006.05a
• /
• pp.53-54
• /
• 2006

The main purpose of tunnel ventilation system is to maintain CO pollutant and VI (visibility index) under an adequate level to provide drivers with safe driving condition. Moreover, it is necessary to minimize power consumption used to operate ventilation system. To achieve the objectives, the control algorithm used in this research is reinforcement teaming (RL) method. RL is a goal-directed teaming of a mapping from situations to actions. The goal of RL is to maximize a reward which is an evaluative feedback from the environment. Constructing the reward of the tunnel ventilation system, two objectives listed above are included. RL algorithm based on actor-critic architecture and natural gradient method is adopted to the system. Also, the recursive least-squares (RLS) is employed to the learning process to improve the efficiency of the use of data. The simulation results performed with real data collected from existing tunnel are provided in this paper. It is confirmed that with the suggested controller, the pollutant level inside the tunnel was well maintained under allowable limit and the performance of energy consumption was improved compared to conventional control scheme.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.29 no.4D
• /
• pp.523-533
• /
• 2009

The purpose of the present study is to examine the dynamic interaction characteristics between moving maglev vehicle and guideway bridge system. For this purpose, the dynamic governing equation of 2-dof maglev vehicle using optimal feedback control scheme of LQG was derived with or without consideration of the dynamic interaction between vehicle and guideway bridge system. From the parametric study, it was found that the dynamic interaction effect between bridge and vehicle was large in case of neglecting the railway roughness effect. But if the railway roughness effect was considered, it was observed two analysis results with or without consideration of the dynamic interaction did not show big difference. As a conclusion, it is required to take into account the dynamic interaction effect of bridge and maglev vehicle and the railway roughness for precise evaluation of runnability of maglev vehicle and impact factor of guideway.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.29 no.2D
• /
• pp.295-303
• /
• 2009

The purpose of present study is to examine the effect of guideway characteristics on runnability of low and medium speed maglev vehicle. Dynamic governing equation for 2-dof vehicle and optimal feedback control scheme are developed. And then the effect of vehicle speed, rail roughness, guideway deflection, continuity of spans, each span length on dynamic response of the UTM-01 maglev vehicle are investigated. From the numerical simulation, it is found that the gap between bogie and guideway does not increase greatly within design velocity of the vehicle. The response of vehicle are mostly affected by the guideway deflection rather than rail roughness. As a result of the present study, the runnability of maglev vehicle can be improved by reducing the maximum deflection of guideway and adopting the continuous girder systems.

• Nuclear Engineering and Technology
• /
• v.25 no.1
• /
• pp.125-135
• /
• 1993

The safe and efficient operation of nuclear plants is recognized to be accomplished through the application of plant automation using digital technology, which is one of main targets of the next generation nuclear plants. For plant level automation, it is first required that each major subsystem be digitalized, and the steam generator water level control system is discussed in this study. The transfer functions between inputs and the level are derived by employing the thermal hydraulic model of the steam generator and are applied to the analysis of the current three-element control system. Since the control scheme in this study includes the steam generator itself as a process plant, the system order is high and the numerical instability arises in digitalizing. Together with this, the unreliability of the feedwater feedback signal at low power level leads to the proposal of a two-element control system with a proper digital controller. The digital PI controller developed for this system has the initial power adaptive gain and integration time constant. And it makes the overall system response satisfy the stability and other necessary control specifications simultaneously. Since the two-element control system using this controller depends on the initial power only, it is simple to define and it shows a similar level response behavior to that of its corresponding analog system.

• Journal of KSNVE
• /
• v.10 no.6
• /
• pp.1059-1066
• /
• 2000

In this paper. a new indirect feedback active noise control (ANC) scheme barred on the fundamental frequency estimation is proposed for systems with a harmonic noise. When reference signals necessary for feedforward ANC configuration are difficult to obtain, the conventional ANC algorithms for multi-tonal noise do not measure the reference signals but generate them with the estimated frequencies.$^{(4)}$ However, the beating phenomena, in which certain frequency components of the noise vanish intermittently, may make the adaptive frequency estimation difficult. The confusion in the estimated frequencies due to the beating phenomena makes the generated reference signals worthless. The proposed algorithm consists of two parts. The first part is a reference generator using the fundamental frequency estimation and the second one is the conventional feedforward control. We propose the fundamental frequency estimation algorithm using decision rules. which is insensitive to the beating phenomena. In addition, the proposed fundamental frequency estimation algorithm has good tracking capability and lower variance of frequency estimation error than that of the conventional cascade ANF method.$^{(4)}$ We are also able to control all interested modes of the noise, even which cannot be estimated by the conventional frequency estimation method because of the poor S/N ratio. We verify the performance of the proposed ANC method through simulations for the measured cabin noise of a passenger ship and the measured time-varying engine booming noise of a passenger vehicle.

• The korean journal of orthodontics
• /
• v.41 no.1
• /
• pp.6-15
• /
• 2011

Objective: The aim of this study was to evaluate the biomechanical aspects of peri-implant bone upon root contact of orthodontic microimplant. Methods: Axisymmetric finite element modeling scheme was used to analyze the compressive strength of the orthodontic microimplant (Absoanchor SH1312-7, Dentos Inc., Daegu, Korea) placed into inter-radicular bone covered by 1 mm thick cortical bone, with its apical tip contacting adjacent root surface. A stepwise analysis technique was adopted to simulate the response of peri-implant bone. Areas of the bone that were subject to higher stresses than the maximum compressive strength (in case of cancellous bone) or threshold stress of 54.8MPa, which was assumed to impair the physiological remodeling of cortical bone, were removed from the FE mesh in a stepwise manner. For comparison, a control model was analyzed which simulated normal orthodontic force of 5 N at the head of the microimplant. Results: Stresses in cancellous bone were high enough to cause mechanical failure across its entire thickness. Stresses in cortical bone were more likely to cause resorptive bone remodeling than mechanical failure. The overloaded zone, initially located at the lower part of cortical plate, proliferated upward in a positive feedback mode, unaffected by stress redistribution, until the whole thickness was engaged. Conclusions: Stresses induced around a microimplant by root contact may lead to a irreversible loss of microimplant stability.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.10 no.2
• /
• pp.152-160
• /
• 2000

This paper addresses the analysis and design of fuzzy-model-based controller for nonlinear systems using extended PDC and optimal pole-placement schemes. In the design procedure, we represent the nonlinear system using a Takagi-Sugeno fkzy model and formulate the controller rules by using the extended parallel distributed compensator (EPDC) and construct an overall fuzzy logic controller by blending all local state feedback controllers with an optimal pole-placement scheme. Unlike the commonly used parallel distributed compensation technique, by blending a newly extended parallel distributed compensator and the optimal poleplacement schemes, we can design not only a local stable k z y controller but also an overall stable fuzzy controller to perform the tacking control objective. Furthermore, a stability analysis is carried out not only for the fuzzy model but also for a real nonlinear system. Finally. the effectiveness and feasibility of the proposed fizzy model-based controller design method has been shown through a simulation example.

• Journal of Korea Multimedia Society
• /
• v.13 no.2
• /
• pp.279-286
• /
• 2010

Unbalanced traffic load offered to the nodes making up a path in the network guaranteeing quality-of-service has been known as a main cause deteriorating the capability of the path in admitting traffic flows. Several path load balancing methods have been developed to resolve this problem which used a feedback control scheme that adjusts the delay budget of a flow allocated to each node according to the conditions of available resource in the path. Because of no consideration about the probabilistic characteristics of the service quality, it is impossible for them to prevent in advance a bottleneck on the path which leads to a native restriction in the improvement of the capability being deteriorated. This paper proposes a preventive intra-path load balancing method applicable to the RSVP system which is based on the probabilistic characteristics of the quality-of-service of the offered load. The results of the simulation of the proposed method on a simple evaluation network showed that it provides the gain of 4~22% compared to the legacy one in terms of the number of admitted flows.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.8
• /
• pp.1675-1685
• /
• 2012

DVC (Distributed Video Coding) technique plays an essential role in providing low-complexity video encoder. But, in order to achieve the better rate-distortion performances, most DVC systems need feedback channel for parity bit control. This causes the DVC-based system to have high decoding latency and becomes as one of the most critical problems to overcome for a real implementation. In order to overcome this problem and to accelerate the commercialization of the DVC applications, this paper analyzes an effect of LDPCA frame size for adaptive LDPCA frame-based parity bit request estimations. First, this paper presents the LDPCA segmentation method in pixel-domain and explains the temporal-based bit request estimation method and the spatial-based bit request estimation method using the statistical characteristics between adjacent LDPCA frames. Through computer simulations, it is shown that the better performance and fast decoding is observed specially when the LDPCA frame size is 3168 in QCIF resolution.

• Journal of the Korea Society for Simulation
• /
• v.22 no.1
• /
• pp.53-61
• /
• 2013

The purpose of this paper is to examine how the humans learn and perceive the weight of objects corresponding to visual information in virtual environment. We conducted two kinds of load-on-tasks with two virtual objects that have same weight but different visual cues; have same visual cues but changed weight by trails. We found that the subject could not generate appropriate force for the smaller and changed weight objects in the beginning of the trials. the discrepancy between the expected weight and actual force consequences sue to visually invoked size and previous experience made subjects perceive the small object were heavier. one the other hand, after the tasks were repeated, the subject responded the weights were the same or very similar when the mismatch between the expected weight and the actual weight became vanished. this means that the sensorimotor feedback influences the anticipatory control scheme and weight perception aggressively in virtual environment.