• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.27 no.8
• /
• pp.1347-1352
• /
• 2003

In periodic cyclic structures, small property irregularity of their substructured often causes significant difference in their dynamic responses, which results in unpredicted premature failures. The small irregularity and the resulting phenomenon are called the mistuning and the vibration localization, respectively. In this paper a simple coupled multi-pendulum system is employed to investigate the effects of the stiffness mistuning and the phase difference in excitation force on the vibration localization of periodic cyclic structures.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.12 no.12
• /
• pp.994-1000
• /
• 2002

In periodically repeated cyclic structures, small property irregularity of their substructures often causes significant difference in their dynamic responses. which results in unpredicted premature failures. The small irregularity and the resulting phenomenon are called the mistuning and the vibration localization. respectively. In this paper, the vibration localization phenomena due to mistuned coupling effects are investigated. To effectively achieve the objective, a simple coupled multi-pendulum system Is employed. The results show that if there exists some coupling stiffness irregularity, vibration localization may occur and becomes more predominant as the number of substructures increases.

• Journal of Advanced Marine Engineering and Technology
• /
• v.20 no.4
• /
• pp.11-19
• /
• 1996

Recently, fuzzy and neural network techniques have been successfully applied to control of complex and ill-defined system in a wide variety of areas, such as robot, water purification, automatic train operation system and automatic container crane operation system, etc. In this paper, we present a neuro-fuzzy controller which unifies both fuzzy logic and multi-layered feedforward neural networks. Fuzzy logic provides a means for converting linguistic control knowledge into control actions. On the other hand, feedforward neural networks provide salient features, such as learning and parallelism. In the proposed neuro-fuzzy controller, the parameters of membership functions in the antecedent part of fuzzy inference rules are identified by using the error backpropagation algorithm as a learning rule, while the coefficients of the linear combination of input variables in the consequent part are determined by using the least square estimation method. Finally, the effectiveness of the proposed controller is verified through computer simulation of an inverted pendulum system.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.40 no.9
• /
• pp.807-813
• /
• 2016

Recently, a spherical pendulum attached to an end-effector of a robot manipulator has been frequently used for a test bed of residual vibration suppression control in a multi-dimensional motion. However, there was no automatic tracking system to detect the current bob position on-line, and there was inconvenience to not be able to store the bob position in real time and plot the trajectory. In this study, we developed a two-dimensional, real-time bob-detecting system using a digital USB camera, of which the key is hardware component design and software C programming for fast image processing and interfacing. The developed system was applied to residual vibration suppression control of a two-dimensional spherical pendulum that is attached at the end-effector of a two degree-of-freedom SCARA robot, and the effectiveness of the developed system has been demonstrated.

• Journal of Industrial Technology
• /
• v.24 no.A
• /
• pp.99-107
• /
• 2004

This paper considers reinforcement learning control with the self-organizing map. Reinforcement learning uses the observable states of objective system and signals from interaction of the system and the environments as input data. For fast learning in neural network training, it is necessary to reduce learning data. In this paper, we use the self-organizing map to parition the observable states. Partitioning states reduces the number of learning data which is used for training neural networks. And neural dynamic programming design method is used for the controller. For evaluating the designed reinforcement learning controller, an inverted pendulum of the cart system is simulated. The designed controller is composed of serial connection of self-organizing map and two Multi-layer Feed-Forward Neural Networks.

• Earthquakes and Structures
• /
• v.13 no.1
• /
• pp.51-58
• /
• 2017

Passive dynamic vibration absorbers (DVAs) are often used to suppress the excessive vibration of a large structure due to their simple construction and low maintenance cost compared to other vibration control techniques. A new type of passive DVA consists of two pendulums connected with spring and dashpot element is investigated. This research evaluated the performance of the DVA in reducing the vibration response of a two degree of freedom shear structure. A model for the two DOF vibration system with the absorber is developed. The nominal absorber parameters are calculated using a Genetic Algorithm(GA) procedure. A parametric study is performed to evaluate the effect of each absorber parameter on performance. The simulation results show that the optimum condition for the absorber frequencies and damping ratios is mainly affected by pendulum length, mass, and the damping coefficient of the pendulum's hinge joint. An experimental model validates the theoretical results. The simulation and experimental results show that the proposed technique is able be used as an effective alternative solution for reducing the vibration response of a multi degree of freedom vibration system.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.6 no.1
• /
• pp.6-13
• /
• 2002

This paper deals with the linear matrix inequality (LMI) design procedures for multi-objective Η$_2$/Η$_{\infty}$ controllers with pole-placement constraints for an inverted pendulum system modeled as convex polytopes to ensure the stabilizing regulator and tracking performances. Polytopic models with multiple linear time-invariant models linearized at some operating points are derived to design controllers overcoming the conservativeness such as a controller may have when it is designed for a model linearized at a single operating point. Multi-objective controllers are designed for polytopic models by the LMT design technique with convex algorithms. It is observed that the inverted pendulum controlled by any controller designed for each polytopic model is stabilizingly restored to the vertical angle position for initial values of larger tilt anlges.

• Earthquakes and Structures
• /
• v.4 no.3
• /
• pp.285-297
• /
• 2013

The effect of soil structure interaction (SSI) on seismic response of a multi-degree-of-freedom structure isolated with a friction pendulum system (FPS) is studied. In the analysis, the soil is considered as an elastic continuum and is modeled using the finite element method. The effect of SSI on response of the structure is evaluated for twenty far-field and twenty near-fault earthquake ground motions. The effect of friction coefficient of sliding material of FPS on SSI is also studied. The results of the study show that the seismic response of the structure increases for majority of the earthquake ground motions due to SSI. The sliding displacement and base shear are underestimated if SSI effects are ignored in the seismic analysis of structures isolated with FPS.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.142-145
• /
• 2003

Many control applications using Neural Network need a priori information about the objective system. But it is impossible to get exact information about the objective system in real world. To solve this problem, several control methods were proposed. Reinforcement learning control using neural network is one of them. Basically reinforcement learning control doesn't need a priori information of objective system. This method uses reinforcement signal from interaction of objective system and environment and observable states of objective system as input data. But many methods take too much time to apply to real-world. So we focus on faster learning to apply reinforcement learning control to real-world. Two data types are used for reinforcement learning. One is reinforcement signal data. It has only two fixed scalar values that are assigned for each success and fail state. The other is observable state data. There are infinitive states in real-world system. So the number of observable state data is also infinitive. This requires too much learning time for applying to real-world. So we try to reduce the number of observable states by classification of states with Self-Organizing Map. We also use neural dynamic programming for controller design. An inverted pendulum on the cart system is simulated. Failure signal is used for reinforcement signal. The failure signal occurs when the pendulum angle or cart position deviate from the defined control range. The control objective is to maintain the balanced pole and centered cart. And four states that is, position and velocity of cart, angle and angular velocity of pole are used for state signal. Learning controller is composed of serial connection of Self-Organizing Map and two Multi-layer Feed-Forward Neural Networks.

• Smart Structures and Systems
• /
• v.22 no.4
• /
• pp.383-397
• /
• 2018

Traditional base isolation systems focus on isolating the seismic response of a structure in the horizontal direction. However, in regions where the vertical earthquake excitation is significant (such as near-fault region), a traditional base-isolated building exhibits a significant vertical vibration. To eliminate this shortcoming, a rocking-isolated system named Telescopic Column (TC) is proposed in this paper. Detailed rocking and isolation mechanism of the TC system is presented. The seismic performance of the TC is compared with the traditional elastomeric bearing (EB) and friction pendulum (FP) base-isolated systems. A 4-storey reinforced concrete moment-resisting frame (RC-MRF) is selected as the reference superstructure. The seismic response of the reference superstructure in terms of column axial forces, base shears, floor accelerations, inter-storey drift ratios (IDR) and collapse margin ratios (CMRs) are evaluated using OpenSees. The results of the nonlinear dynamic analysis subjected to multi-directional earthquake excitations show that the superstructure equipped with the newly proposed TC is more resilient and exhibits a superior response with higher margin of safety against collapse when compared with the same superstructure with the traditional base-isolation (BI) system.