• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.110-115
• /
• 2004

The hybrid fault diagnosis method based on a combination of the signed digraph (SDG) and the partial least-squares (PLS) has the advantage of improving the diagnosis resolution, accuracy and reliability, compared to those of previous qualitative methods, and of enhancing the ability to diagnose multiple fault. In this study, the method is applied for the multiple fault diagnosis of the Tennessee Eastman challenge process, which is a realistic industrial process for evaluating process contol and monitoring methods. The process is decomposed using the local qualitative relationships of each measured variable. Dynamic PLS (DPLS) model is built to estimate each measured variable, which is then compared with the estimated value in order to diagnose the fault. Through case studies of 15 single faults and 44 double faults, the proposed method demonstrated a good diagnosis capability compared with previous statistical methods.

• Proceedings of the KSME Conference
• /
• 2004.04a
• /
• pp.1352-1357
• /
• 2004

Thermal error compensation has been developed for CNC (Computer Numerical Control) machining center with moving heat sources. The thermal error in CNC machining center has an effect on machining accuracy more than the geometric error does. Thus, temperature distributions of a spindle unit have been analyzed numerically by a Finite Differential Method and experimentally by an infrared (IR) camera in this study. A multiple variable method has been derived to estimate the thermal deformation of the machine origin stably and effectively after measuring deformation and temperature data. The experimental results for a vertical machining center have shown that the thermal errors of the machine origins were reduced more than 30% by the developed method.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.63 no.9
• /
• pp.1186-1191
• /
• 2014

Weekly and monthly electric load forecasting are essential for the generator maintenance plan and the systematic operation of the electric power reserve. This paper proposes the weekly maximum electric load forecasting model for 104 weeks with the multiple regression model. Input variables of the multiple regression model are temperatures and GDP that are highly correlated with electric loads. The weekly variable is added as input variable to improve the accuracy of electric load forecasting. Test results show that the proposed algorithm improves the accuracy of electric load forecasting over the seasonal autoregressive integrated moving average model. We expect that the proposed algorithm can contribute to the systematic operation of the power system by improving the accuracy of the electric load forecasting.

• Korean Business Review
• /
• v.4
• /
• pp.281-295
• /
• 1991

Nothing is more incorrect than forecasting. Nevertheless, forecasting is one of the most important business activities for the effective management. There has been rapid changes of the growth rate in every respect of the Korean hospitaity industry, especially the hotel industry, before and after the 88 Olympic Games. Therefore, the hoteliers shall be in need of more-than-ever accourate demand forecasting for the more systematic management and control. Under the above circumstances, this study suggested the best forecasting technique and method for the better sales and operations of the hotel rooms. The number of rooms sold is selected as a dependent variable of this study which is regarded as the best representative factor of measuring the growth rate of the rooms division performance of the hotels. The first step was to select the most verifiable independent variable diferently from the other countries or other areas of Korea. As a result, the number of foreign visitors was chosen. Empirical research, i.e. correlation and multiple regression analysis, shows that this independent variable has a strong relationship with the dependent variable told above. The second procedure was to estimate the number of rooms will be sold in 1991 on the basis of the formula calculated through the multiple regression analysis. Time series technique was conducted using the data of the number of foreign visitors by purpose of travel from 1987 to 1990. For the more correct forecasting, however, it would be desirable to adopt the data from 1989 considering the product or the industry life cycle. In addition, deeper analysis for the monthly or seasonal forecasting method is needed as a future research.

• Communications for Statistical Applications and Methods
• /
• v.29 no.6
• /
• pp.629-640
• /
• 2022

Variable selection is one of the most crucial tasks in supervised learning, such as regression and classification. The best subset selection is straightforward and optimal but not practically applicable unless the number of predictors is small. In this article, we propose directly solving the best subset selection via the genetic algorithm (GA), a popular stochastic optimization algorithm based on the principle of Darwinian evolution. To further improve the variable selection performance, we propose to run multiple GA to solve the best subset selection and then synthesize the results, which we call ensemble GA (EGA). The EGA significantly improves variable selection performance. In addition, the proposed method is essentially the best subset selection and hence applicable to a variety of models with different selection criteria. We compare the proposed EGA to existing variable selection methods under various models, including linear regression, Poisson regression, and Cox regression for survival data. Both simulation and real data analysis demonstrate the promising performance of the proposed method.

• Smart Structures and Systems
• /
• v.19 no.2
• /
• pp.203-211
• /
• 2017

This paper provides a numerical solution for multiple confocal elliptic dissimilar cylinders. In the problem, the inner elliptic notch is under the traction free condition. The medium is composed of many confocal elliptic dissimilar cylinders. The transfer matrix method is used to study the continuity condition for the stress and displacement along the interfaces. Two cases, or the infinite matrix case and the finite matrix case, are studied in this paper. In the former case, the remote tension is applied in y- direction. In the latter case, the normal loading is applied along the exterior elliptic contour. For two cases, several numerical results are provided.

• Structural Engineering and Mechanics
• /
• v.84 no.3
• /
• pp.295-310
• /
• 2022

This paper proposes a modified bar simulation method for analyzing the shear lag effect of variable sectional box girder with multiple cells. This theoretical method formulates the equivalent area of stiffening bars and the allocation proportion of shear flows in webs, and re-derives the governing differential equations of bar simulation method. The feasibility of the proposed method is verified by the model test and finite element (FE) analysis of a simply supported multi-cell box girder with constant depth. Subsequently, parametric analysis is conducted to explore the mechanism of shear lag effect of varied sectional cantilever box girder with multiple cells. Results show that the shear lag behavior of variable box-section cantilever box girder is weaker than that of box girder with constant section. It is recommended to make the gradient of shear flow in the web with respect to span length vary as smoothly as possible for eliminating the shear lag effect of box girder. An effective countermeasure for diminishing shear lag effect is to increase the number of box chambers or change the variation manner of bridge depth. The shear lag effect of varied sectional cantilever box girder will get more server when the length of central flanges is shorter than 0.26 or longer than 0.36 times of total width of top flange, as well as the cantilever length exceeds 0.29 times of total length of box's flange. Therefore, the distance between central webs can adjust the shear lag effect of box girder. Especially, the width ratio of cantilever plate with respect to total length of top flange is proposed to be no more 1/3.

• Smart Structures and Systems
• /
• v.21 no.4
• /
• pp.449-467
• /
• 2018

In this paper, a theoretical and numerical study on bridge simultaneous damage detection procedure for identifying both the system parameters and input excitation mass, are presented. This method is called 'Adjoint Variable Method' which is an iterative gradient-based model updating method based on the dynamic response sensitivity. The main advantage of proposed method is inclusion of an analytical method to augment the accuracy and speed of the solution. Moving mass is a model which takes into account the inertia effects of the vehicle. This interaction model is a time varying system and proposed method is capable of detecting damage in this variable system. Robustness of proposed method is illustrated by correctly detection of the location and extension of predetermined single, multiple and random damages in all ranges of speed and mass ratio of moving vehicle. A comparison study of common sensitivity and proposed method confirms its efficiency and performance improvement in sensitivity-based damage detection methods. Various sources of errors including the effects of measurement noise and initial assumption error in stability of method are also discussed.

• Journal of Korean Neurosurgical Society
• /
• v.53 no.2
• /
• pp.102-107
• /
• 2013

Objective : The Leksell Gamma Knife$^{(R)}$ (LGK) is based on a single-fraction high dose treatment strategy. Therefore, independent verification of the Leksell GammaPlan$^{(R)}$ (LGP) is important for ensuring patient safety and minimizing the risk of treatment errors. Although several verification techniques have been previously developed and reported, no method has ever been tested statistically on multiple LGK target treatments. The purpose of this study was to perform and to evaluate the accuracy of a verification method (modified variable ellipsoid modeling technique, MVEMT) for multiple target treatments. Methods : A total of 500 locations in 10 consecutive patients with multiple brain tumor targets were included in this study. We compared the data from an LGP planning system and MVEMT in terms of dose at random points, maximal dose points, and target volumes. All data was analyzed by t-test and the Bland-Altman plot, which are statistical methods used to compare two different measurement techniques. Results : No statistical difference in dose at the 500 random points was observed between LGP and MVEMT. Differences in maximal dose ranged from -2.4% to 6.1%. An average distance of 1.6 mm between the maximal dose points was observed when comparing the two methods. Conclusion : Statistical analyses demonstrated that MVEMT was in excellent agreement with LGP when planning for radiosurgery involving multiple target treatments. MVEMT is a useful, independent tool for planning multiple target treatment that provides statistically identical data to that produced by LGP. Findings from the present study indicate that MVEMT can be used as a reference dose verification system for multiple tumors.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.18 no.1
• /
• pp.139-144
• /
• 2008

Generally, real system that is the object of decision-making is very variable and sometimes it lies situations with uncertainty. To solve these problem, it has used statistical methods as significance level, certainty factor, sensitivity analysis and so on. In this paper, we propose a method for fuzzy decision-making based on MFAC(Multiple Fuzzy Alpha Cut) to improve the definiteness of classification results with similarity evaluation. In the proposed method, MFAC is used for extracting multiple a ${\alpha}$-level with proximity degree at proximity relation between relative Hamming distance and max-min method and for minimizing the number of data which are associated with the partition intervals extracted by MFAC. To determine final alternative of decision-making, we compute the weighted value between extracted data by MFAC From the experimental results, we can see the fact that the proposed method is simpler and more definite than classification performance of the conventional methods and determines an alternative efficiently for decision-maker by testing significance of sample data through statistical method.