• Communications for Statistical Applications and Methods
• /
• 제19권6호
• /
• pp.761-770
• /
• 2012

We frequently encounter outcomes of count that have extra variation. This paper considers several alternative models for overdispersed count responses such as a quasi-Poisson model, zero-inflated Poisson model and a negative binomial model with a special focus on a generalized linear mixed model. We also explain various goodness-of-fit criteria by discussing their appropriateness of applicability and cautions on misuses according to the patterns of response categories. The overdispersion models for counts data have been explained through two examples with different response patterns.

• Animal Bioscience
• /
• 제36권8호
• /
• pp.1263-1273
• /
• 2023

Objective: This study was conducted to evaluate the effects of crude protein (CP) levels on the physiological response, reproductive performance, blood profiles, milk composition and odor emission in gestating sows. Methods: Seventy-two multiparous sows (Yorkshire×Landrace) of average body weight (BW), backfat thickness, and parity were assigned to one of six treatments with 10 or 11 sows per treatment in a completely randomized design. Experimental diets with different CP levels were as follows: i) CP11, corn-soybean-based diet containing 11% CP; ii) CP12, corn-soybean-based diet containing 12% CP; iii) CP13, corn-soybean-based diet containing 13% CP; iv) CP14, corn-soybean-based diet containing 14% CP; v) CP15, corn-soybean-based diet containing 15% CP; and vi) CP16: corn-soybean-based diet containing 16% CP. Results: There was no significant difference in the performance of sow or piglet growth when sows were fed different dietary protein levels. Milk fat (linear, p = 0.05) and total solids (linear, p = 0.04) decreased as dietary CP levels increased. Increasing dietary CP levels in the gestation diet caused a significant increase in creatinine at days 35 and 110 of gestation (linear, p = 0.01; linear, p = 0.01). The total protein in sows also increased as dietary CP levels increased during the gestation period and 24 hours postpartum (linear, p = 0.01; linear, p = 0.01). During the whole experimental period, an increase in urea in sows was observed when sows were fed increasing levels of dietary CP (linear, p = 0.01), and increasing blood urea nitrogen (BUN) concentrations were observed as well. In the blood parameters of piglets, there were linear improvements in creatinine (linear, p = 0.01), total protein (linear, p = 0.01), urea (linear, p = 0.01), and BUN (linear, p = 0.01) with increasing levels of dietary CP as measured 24 hours postpartum. At two measurement points (days 35 and 110) of gestation, the odor gas concentration, including amine, ammonia, and hydrogen sulfide, increased linearly when sows fed diets with increasing levels of dietary CP (linear, p = 0.01). Moreover, as dietary CP levels increased to 16%, the odor gas concentration was increased with a quadratic response (quadratic, p = 0.01). Conclusion: Reducing dietary CP levels from 16% to 11% in a gestating diet did not exert detrimental effects on sow body condition or piglet performance. Moreover, a low protein diet (11% CP) may improve dietary protein utilization and metabolism to reduce odor gas emissions in manure and urine in gestating sows.

• 한국강구조학회 논문집
• /
• 제20권1호
• /
• pp.21-31
• /
• 2008

응답면기법(RSM, Response surface method)은 복잡한 구조물의 매우 작은 발생확률이나 신뢰성해 석에서 폭넓게 사용된다. MCS(Monte-Carlo Simulation)방법은 어떤 시스템의 평가에서도 사용될 수 있으나 해석시간이 파괴확률의 역수에 비례하게 되어 발생확률이 매우 희박한 시스템의 평가에 불리하다. 확률유한요소해석법은 이러한 MCS의 한계점을 해결해 줄 수 있는 대안이 될 수 있다. 그러나 이 방법은 평균과 표준편차 등이 모델링 (내부 프로그래밍)된 특별한 프로그램에서만 적용 가능하며 임의의 범용소프트웨어의 응답을 모델링하거나 임의의 프로그램의 특성을 이용할 수가 없다. RSM방법은 복잡한 구조시스템에서 응답에 대한 회귀모델을 구성하여 효율적인 해석단계를 통해 시간과 노력을 획기적으로 절감시킬 수 있다. 그러나 RSM의 정확도는 한계상태방정식의 선형성과 축점간의 거리에 영향을 받게 된다. 이런 단점을 해결하기 위해 한계상태방정식의 선형성과 무관하게 정확한 수렴해를 구하기 위한 개선된 적응적 응답면기법을 개발하고 선형과 2차형식의 응답면방정식에 대한 2가지 예를 들어 검증하였다. 검증결과 가장 효율적인 RSM기법을 결정하였다. 개발된 선형적응적가중응답 면기법 (linear adaptive weighted response surface method, LAW-RSM)은 비적응적이거나 비가중형식의 2차 RSM기법에 비해서 정해의 신뢰성지수에 가장 근접한 정확성과 수렴성을 나타낸다.

• 제어로봇시스템학회논문지
• /
• 제19권4호
• /
• pp.285-289
• /
• 2013

In switching from the track-seeking or track-jumping control mode to track-following control mode in hard disk drives or optical disk drives, initial values of the feedback controller are tuned to improve the transient response. In general, all the initial values of the controller have been compensated for this purpose. In this paper, by partially compensating initial values of digital controllers, we achieve a good performance of the transient response. In the proposed method for IVC (Intial Value Compensation), LMIs (Linear Matrix Inequalities) are used, which includes conditions for improving the performance of the transient response such as reducing a tracking error and control efforts. We obtain optimal initial values of the controller by solving an optimization problem with constraints represented by only one LMI. Although initial values of the controller are partially compensated, we can show that not only a sufficient performance of the transient response is obtained but also control efforts are diminished. The feasibility of the method is verified by simulation studies.

• 전산구조공학
• /
• 제10권3호
• /
• pp.125-131
• /
• 1997

대부분의 동적계는 기진력 및 계 인자들에 있어서 다양한 불확정 특성을 갖고 있다. 본 연구에서는 기진력의 불확정성과 계 인자들의 불확정성을 모두 갖는 선형 동적계에 대한 응답해석 과정을 제안하였다. 확률특성을 갖는 계 인자와 응답은 섭동법에 의해 모델링되었으며, 응답해석은 불규칙 진동 이론에 의하여 정식화 되었다. 또한 제안된 응답 모델에 의해 계산되기 어려운 응답의 평균에 대한 해석은 확률유한요소법을 사용하였다. 적용 예로서 정상 백색잡음 기진력을 받으며 불확정 질량과 스프링 상수를 갖는 1자유도계 문제에 대하여 과도응답을 계산하고, 그 결과를 수치 시뮬레이션 결과와 비교하여 그 타당성을 검토하였다.

• 한국지반공학회:학술대회논문집
• /
• 한국지반공학회 2005년도 춘계 학술발표회 논문집
• /
• pp.294-301
• /
• 2005

Ground response analysis is one of the most important and most commonly encountered problems in geotechnical earthquake engineering. It is a prerequisite step for liquefaction assessment of saturated soil or the pseudo-static and dynamic analysis of geotechnical structures. A number of techniques have been developed for ground response analysis. In this study, ground response analyses were performed using the computer programs that are currently being used domestically. From these analyses, the analysis techniques applied to the programs were compared and analyzed. The results of ground response analyses were compared as follows: 1) 1-dimensional analysis vs. 2-dimensional analysis; 2) equivalent linear analysis vs. nonlinear analysis.

• 대한기계학회논문집
• /
• 제19권1호
• /
• pp.45-60
• /
• 1995

It is attempted to estimate excitation force of a linear vibratory system using measured vibration responses. The excitation force is estimated from the relationship between the vibration response and system characteristic matrices which are extracted from both the mathematical model of the system and actual response in contrast to the usual approach of inverting the frequency response matrices. This extraction scheme is based on the fact that the vibration response can be expressed in term of linear combination of frequency domain modal vectors defined as mutually orthonormal basis vectors in frequency domain. The extracted frequency domain basis vectors are very stable in computational manipulation. It is found that the estimated excitation force is in good agreement with actually measured force except at the natural frequencies the structure, which is the common feature still to be overcome by the research efforts in this area. From the results of this paper, this disagreement is considered to come from the discrepancy between the model and actual value of the mass, damping and stiffness of the structure.

• Wind and Structures
• /
• 제2권3호
• /
• pp.151-171
• /
• 1999

In its 90 years of life, the Tuned Mass Damper have found application in many fields of engineering as a vibration reducing device. The evolution of the theory of TMDs is briefly outlined in the paper. A generalised mathematical linear model for the analysis of the response of line-like structures with TMDs is presented. The system matrices of the system including the TMDs are written in the state space as a function of the mean wind speed. The stability of the system can be analysed and the Power Spectral Density Function of any response parameter calculated, taking into account an arbitrary number of modes of vibration as well as an arbitrary number of TMDs, for any given PSDF of the excitation. The procedure can be used to optimise the number, position and mechanical properties of the damping devices, with respect to any response parameter. Due to the stationarity of the excitation, the method is well suited to structures subjected to the wind action. In particular the procedure allows the calculation of the onset galloping wind speed and the response to buffeting, and a linearisation of the aeroelastic behaviour allows its use also for the evaluation of the response to vortex shedding. Finally three examples illustrate the suggested procedure.

• Earthquakes and Structures
• /
• 제22권6호
• /
• pp.585-595
• /
• 2022

Using the nonlinear static procedures has become very common in seismic codes to achieve the nonlinear response of the structure during an earthquake. The capacity spectrum method (CSM) adopted in ATC-40 is considered as one of the most known and useful procedures. For this procedure the seismic demand can be approximated from the maximum deformation of an equivalent linear elastic Single-Degree-of-Freedom system (SDOF) that has an equivalent damping ratio and period by using an iterative procedure. Data from the results of this procedure are plotted in acceleration- displacement response spectrum (ADRS) format. Different improvements have been made in order to have more accurate results compared to the Non Linear Time History Analysis (NL-THA). A new procedure is presented in this paper where the iteration process shall not be required. This will be done by estimation the ductility demand response spectrum (DDRS) and the corresponding effective damping of the bilinear system based on a new parameter of control, called normalized yield strength coefficient (η), while retaining the attraction of graphical implementation of the improved procedure of the FEMA-440. The proposed procedure accuracy should be verified with the NL-THA analysis results as a first implementation. The comparison shows that the new procedure provided a good estimation of the nonlinear response of the structure compared with those obtained when using the NL-THA analysis.

• 한국지반공학회논문집
• /
• 제24권12호
• /
• pp.113-120
• /
• 2008

1차원 지반응답해석은 지반에 의한 지진동의 증폭현상을 모사하는 데 널리 사용되고 있다. 가장 널리 사용되고 있는 등가선형 지반응답해석기법은 유효변형률에 상응하는 전단탄성계수와 감쇠비를 전 주파수 영역에 일률적으로 적용하며, 진동주파수에 대한 지반 응답의 영향을 고려하지 않는다. 지반의 비선형 거동을 주파수영역에서 보다 정확하게 모사하기 위하여 진동주파수와 지반거동의 상관관계를 주파수-변형률 곡선으로써 모의하는 등가선형해석기법이 개발되었으며, 이 방법은 기존의 등가선형해석기법에 비하여 정확성을 향상시킨다고 알려져 있다. 이제까지 다양한 형상의 주파수변형률 곡선이 제안되었으며, 이들은 한결같이 해석의 정확성을 증가시킨다고 주장된 바 있다. 본 연구에서는 기존의 연구에서 제안되었건 두 개의 주파수-변형률 관계 완화곡선과 본 연구에서 새롭게 제안된 세 개의 곡선을 이용하여 주파수 의존성을 고려한 등가선형해석기법의 정확성을 검증하였다. 검증에는 세 개의 부지에서 기록된 6개의 지진파가 사용되었다. 해석결과, 완화곡선은 지반응답에 지배적인 영향을 미치는 것으로 나타났으며, 주파수 의존성을 고려한 해석은 정확성을 향상시킬수 있는 것으로 나타났다. 하지만, 모든 사례에서 가장 정확하게 응답을 예측하는 완화곡선은 없는 것으로 나타났으며, 해석사례별로 최적의 완화곡선이 다른 것으로 나타났다. 따라서, FDEL 해석 수행 시, 다양한 완화곡선을 사용해야 할 것으로 판단되며 등가선형해석과 병행해서 수행되어야 한다고 판단된다.