• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 한국소음진동공학회 2004년도 추계학술대회논문집
• /
• pp.419-422
• /
• 2004

The paper classified SI schemes for structural engineering applications based on the type of measured data. Only parametric SI algorithms with optimization processes were reviewed where optimal structural parameters are estimated by minimizing an output error between measured and computed responses. Some important issues in applying SI schemes were analyzed with the definition of an analytical model, noise and sparseness in measured data. As a sample study, the application of a nonlinear time-domain SI algorithm for a shear building was examined.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 한국소음진동공학회 2004년도 추계학술대회논문집
• /
• pp.435-440
• /
• 2004

In this paper, the tension of assembly stay cable connected with massive anchorage block was calculated through back analysis of in-situ frequencies measured from a stadium structure. Direct approach to back analysis is adopted using the univariate method among the direct search methods as an optimization technique. The univariate method can search the optimal tension without regard to the initial ones and has a rapid convergence rate. To verify the reliability of back analysis, Tension formulas proposed by Zui et al. and Shimada were used. Tensions estimated by three methods are compared with the design tension, and are in a reasonable agreement with an error of more or less than 15%. Therefore, it is shown that back analysis applied in this paper is appropriate for estimation of cable tension force.

• Proceedings of the Earthquake Engineering Society of Korea Conference
• /
• 한국지진공학회 2003년도 춘계 학술발표회논문집
• /
• pp.216-223
• /
• 2003

A time-domain system identification (SI) method is developed for seismic damage assessment on structures. SI algorithms for complete measurements with respect to degrees-of-freedom are proposed. To take account of nonlinear dynamic response, an equation error in the incremental dynamic governing equation is defined for complete measurement between measured and computed acceleration. Variations of stiffness and damping parameters during earthquake vibration are chased by utilizing a constrained nonlinear optimization tool available in MATLAB. A simulation study has been carried out to identify damage event and to assess damage severity by using measured acceleration time history. Mass properties are assumed as known a priori. The effects of measurement noise on the identification are also investigated.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2001년도 ICCAS
• /
• pp.41.4-41
• /
• 2001

In this paper, we propose a LQR(Linear Quadratic Regulator) controller design for the active suspension using two-degree-of-freedom quarter-car model. We can improve the inherent suspension problem, the tradeoff between ride quality and suspension travel by selecting appropriate weights in the LQR-objective function. Because any definite rules for selecting weights do not exist, we replace the designer´s trial and error with the optimization-algorithm, ES(Evolution Strategy). Using the ES, we can find the proper control gains for selected frequencies, which have major effects on the vibrations of the vehicle´s state variables.

• Korean Journal of Chemical Engineering
• /
• 제36권3호
• /
• pp.356-367
• /
• 2019

This work presents an advanced and systematic approach to analytically design the optimal parameters of a two parameter second-order internal model control (IMC) filter that satisfies operational constraints on the output process, the manipulated variable as well as rate of change of the manipulated variable, for a first-order plus dead time (FOPDT) process. The IMC parameters are designed to minimize a control objective function composed of the weighted sum of the error between the process variable and the set point, and the rate of change of the manipulated variable, and to satisfy the desired constraints. The feasible region of the constrained IMC control parameters was graphically analyzed, as the process parameters and the constraints varied. The resulting constrained IMC control parameters were also used to find the corresponding industrial proportional-integral controller parameters of a Smith predictor structure.

• Journal of Integrative Natural Science
• /
• 제11권4호
• /
• pp.204-208
• /
• 2018

Comparison of different optimizer performance in photovoltaic power modeling using artificial neural deep learning techniques is described in this paper. Six different deep learning optimizers are tested for Long-Short-Term Memory networks in this study. The optimizers are namely Adam, Stochastic Gradient Descent, Root Mean Square Propagation, Adaptive Gradient, and some variants such as Adamax and Nadam. For comparing the optimization techniques, high and low fluctuated photovoltaic power output are examined and the power output is real data obtained from the site at Mokpo university. Using Python Keras version, we have developed the prediction program for the performance evaluation of the optimizations. The prediction error results of each optimizer in both high and low power cases shows that the Adam has better performance compared to the other optimizers.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 한국정보통신학회 2016년도 추계학술대회
• /
• pp.161-163
• /
• 2016

Organic semiconductors, in contrast with inorganic semiconductors, have DOS of the Gaussian distribution function which leads to difficulties to obtain the electron density with a closed equation. For this reason, the prediction of the device operation is extremely difficult in the development of OLED and the repetitive trial-and-error is required to find the appropriate electrical and optical properties of the devices. In this paper, the standard to analyze organic semiconductor is proposed by the optimization of the time and the accuracy which is necessary for the numerical calculation of the electron density in organic semiconductors.

• Corrosion Science and Technology
• /
• 제21권2호
• /
• pp.121-129
• /
• 2022

A study on water-based epoxy coated on mild steel using the electroplating method was conducted to optimize the process parameters for dry film thickness to achieve superior paint quality at optimal cost in an automotive plant. The regression model was used to adjust various parameters such as electrode voltage, bath temperature, processing time, non-volatile matter, and surface area to optimize the dry film thickness. The average dry film thickness computed using the model was in the range of 15 - 35 ㎛. The error in the computed dry film thickness with reference to the experimentally measured dry film thickness value was - 0.5809%, which was well within the acceptable limits of all paint shop standards. Our study showed that the dry film thickness on mild steel was more sensitive to electrode voltage and bath temperature than processing time. Further, the presence of non-volatile matter was found to have the maximum impact on dry film thickness.

• Advances in Energy Research
• /
• 제8권3호
• /
• pp.145-153
• /
• 2022

Recent concerns about rising fuel prices and greenhouse gas emissions have focused attention on alternative energy sources, particularly in the transport sector. Transportation consumes 40% of overall fuel usage. As a result, a growing majority of researches on Electric Vehicles (EVs) and their Energy Management Systems (EMS) have been done. In order to enhance the performance and to meet the needs of drivers, more information regarding the EMS is needed. A new Energy Management System is proposed using a FOPID controller. To put the concept into practice, state equations are utilised. The fifth-order state-space model under study is a linked model with several inputs and outputs and the transfer matrices are calculated for decoupling the system. Utilizing these transfer matrices to decouple the system and FOPID controller is used to tune the system. The tuned parameters are minimized using a Particle Swarm Optimization (PSO) approach with Integral Time Absolute Error (ITAE) as the goal. When the suggested FOPID system's results are compared to those of PID-controlled systems, a sizable improvement is observed, which is explained by the results.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 한국정보통신학회 2022년도 추계학술대회
• /
• pp.93-96
• /
• 2022

DTW can eliminate gaps between sequences of different lengths and find out the similarity of patterns, but due to the time and space complexity, it requires a high computational cost on large datasets. In this paper, we propose a DDTW algorithm that not only reduces computational costs but also has no error in the results. In addition, the algorithm complexity of DTW and DDTW is compared by measuring the computational time according to the length of the sequence. Simulation results show a noticeable reduction in computational time in DDTW compared to DTW.