• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.78.5-78
• /
• 2002

We propose an evolutionary reinforcement learning (RL) system with time-varying parameters that can deal with a dynamic environment. The proposed system has three characteristics: 1) It can deal easily with a dynamic environment by using time-varying parameters; 2) The division of state space is acquired evolutionarily by genetic algorithm (GA); 3) One does not have to design the rules constructing an agent in advance. So far many RL systems have been proposed. These systems adjust constant or non time-varying parameters; by those systems it is difficult to realize appropriate behavior in complex and dynamic environment. Hence, we propose the RL system whose parameters can vary temporally. T...

• Smart Media Journal
• /
• v.12 no.9
• /
• pp.124-133
• /
• 2023

Currently, Korea shares culture, economy, and society together as a member of the international community along with rapid economic growth. In particular, in the 2000s, the construction of local cultural performance halls has been promoted through efforts and methods for the development of local culture. As a result, many local governments built medium-sized or larger performance halls, and achieved both quantitative and qualitative effects by satisfying citizens and attracting visitors. However, this study aims to analyze the safety of the space used by many visitors at the same time and whether an effective evacuation plan is applied to frequent disasters and disaster situations in recent years.

• Structural Engineering and Mechanics
• /
• v.75 no.1
• /
• pp.33-47
• /
• 2020

In this paper the dynamic behavior of an isolated building subjected to idealized near-fault pulses is investigated. The building is represented with a simple 2-DOF model. Both linear and non-linear behavior of the isolation system is considered. Using dimensional analysis, in conjunction with closed form mathematical idealized pulses, appropriate dimensionless parameters are defined and self-similar curves are plotted on dimensionless graphs, based on which various conclusions are reached. In the linear case, the role of viscous damping is examined in detail and the existence of an optimum value of damping along with its significant variation with the number of half-cycles is shown. In the nonlinear case, where the behavior of the building depends on the amplitude of the excitation, the benefits of dimensional analysis are evident since the influence of the dimensionless 𝚷-terms is easily examined. Special consideration is given to the normalized strength of the non-linear isolation system that appears to play a complex role which greatly affects the response of the 2-DOF. In the last part of the paper, a comparison of the responses to idealized pulses between a linear fixed-base SDOF and the respective isolated 2-DOF with both linear and non-linear damping is conducted and it is shown that, under certain values of the superstructure and isolation system characteristics, the use of an isolation system can amplify both the normalized acceleration and displacement of the superstructure.

• Proceedings of the KIEE Conference
• /
• 2000.07d
• /
• pp.2804-2807
• /
• 2000

This paper suggests a method of the forward dynamic analysis for the computer simulation on the analysis of the dynamic behavior for biped walking robot. The equations f motion of the system or the simulation are constructed by using the Method of the multibody dynamics which is powerful method for modeling of the complex biped system. For the simplicity of simulation, we consider that the sole of the contacting foot is affected by the reaction forces for tree structure system topology instead of the addition or deletion of the kinematic constraints. The ground reaction forces can be modeled using the simple spring and damper model at the three contacting points on the sole of the foot. For minimizing the errors of numerical integration, the number of equations of motion is minimized by adding the driving constraints or a controller instead of the direct driving torques.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.28 no.11
• /
• pp.1656-1663
• /
• 2004

As the structures become larger, higher and more complicated, the demand for safety level has increased. In recent years, TLD(Tuned Liquid Damper) proved to be a successful control tool for reducing structural vibrations. For this reason, the influence of some key parameters of the U-type TLD on the dynamic response is studied. And simple and effectively developed learning control logic is used to control vibration of U type Tuned Liquid Damper system. The purpose of this paper is design optimal control system to deal with unknown errors from non linearity and variation that cost modeling difficulty in complex structure and is followed with the desired behavior. Finally this hybrid control method applied to U type Tuned Liquid Damper structure gives the benefit from better performance of precision and stability of the structure by reducing vibration effect. This research leads to safety design in various structure to robust unspecified foreign disturbances such as windy-load and earthquake.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2004.10a
• /
• pp.1113-1118
• /
• 2004

This paper presents the dynamic analysis method for an electromechanical system. The engineer has at his disposal a variety of software simulation tools. However, difficulties arise when the study of the behavior of complex electromechanical systems in combination with coupling element is required. Typical examples of such systems are machines for factory automation, home automation, and office automation. Dynamic systems analysis packages or electronic systems analysis packages offer the restrictive to simulate these mixed systems such electromechanical product. Electronic circuit analysis algorithm is easily incorporated into a multi-body dynamics analysis algorithm. The governing equation of electronic circuit is formulated as a differential algebraic equation form including both electrical and mechanical variables and is simultaneously solved in every time step. This analysis method clearly demonstrates the application potential for mixed electromechanical simulation.

• ETRI Journal
• /
• v.31 no.1
• /
• pp.92-94
• /
• 2009

We propose a system for mobile-phone fraud detection based on a bidirectional artificial neural network (bi-ANN). The key advantage of such a system is the ability to detect fraud not only by offline processing of call detail records (CDR), but also in real time. The core of the system is a bi-ANN that predicts the behavior of individual mobile-phone users. We determined that the bi-ANN is capable of predicting complex time series (Call_Duration parameter) that are stored in the CDR.

• Journal of Industrial Technology
• /
• v.17
• /
• pp.323-330
• /
• 1997

In this paper, a new approach to modeling of nonlinear systems using fuzzy theory is presented. To express the various and complex behavior of nonlinear system, we combine multiple model method with hierachical prioritized structure, and the mountain clustering technique is used in partitioning of system. TSK rule structure is adopted to form the fuzzy rules, and Back propagation algorithm is used for learning parameters in consequent parts of the rules. Also we soften the paradigm of Mamdani's inference mechanism by using Yager's S-OWA operators. Computer simulations are performed to verify the effectiveness of the proposed method.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 1996.03a
• /
• pp.81-90
• /
• 1996

In this paper, a finite element based system is presented for the prediction of the distributions of the recrystallized grain sizes in the workpiece in hot forging. The system adopts fully coupled finite element thermo-mechanical model for predicting plastic deformatin and heat transfer occuring in the workpiece, and employsexisting metallurgical models relating the recrystallization behavior with the thermo-mechanical variables such as temperatures, strain, and strain rate. The system is applied to upsetting of cylindrical preform. The predicted grain sizes are compared with the measurements . It is further applied to forging of a complex-shaped product.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2010.10a
• /
• pp.571-573
• /
• 2010

Most real world system such as world economy, management, medical and engineering applications contain a series of complex phenomena. One of common methods to understand these system is to build a model and analyze the behavior of the system. As a first step, Determining the best clusters on data. As a second step, Determining the model of the cluster. In this paper, we investigated heuristic search methods for efficient clustering.