• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38C no.2
• /
• pp.213-218
• /
• 2013

One of the most popular approaches to parameter adaptation in hidden Markov model (HMM) based systems is the maximum likelihood linear regression (MLLR) technique. In our previous study, we proposed factored MLLR (FMLLR) where each MLLR parameter is defined as a function of a control vector. We presented a method to train the FMLLR parameters based on a general framework of the expectation-maximization (EM) algorithm. Using the proposed algorithm, supplementary information which cannot be included in the models is effectively reflected in the adaptation process. In this paper, we apply the FMLLR algorithm to a pitch sequence as well as spectrum parameters. In a series of experiments on artificial generation of expressive speech, we evaluate the performance of the FMLLR technique and also compare with other approaches to parameter adaptation in HMM-based speech synthesis.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.25 no.4
• /
• pp.20-27
• /
• 2021

The present study presents a nonlinear finite element analysis (FEA) approach using the general program of Abaqus to evaluate the seismic response of unreinforced masonry walls strengthened with the steel bar truss system developed in the previous investigation. For finite element models of masonry walls, the concrete damaged plasticity (CDP) and meso-scale methods were considered on the basis of the stress-strain relationships under compression and tension and shear friction-slip relationship of masonry prisms proposed by Yang et al. in order to formulate the interface characteristics between brick elements and mortars. The predictions obtained from the FEA approach were compared with test results under different design parameters; as a result, a good agreement could be observed with respect to the crack propagation, failure mode, rocking strength, peak strength, and lateral load-displacement relationship of masonry walls. Thus, it can be stated that the proposed FEA approach shows a good potential for designing the seismic strengthening of masonry walls.

• The Plant Pathology Journal
• /
• v.35 no.3
• /
• pp.189-199
• /
• 2019

In this research, the relationships among the 31 microsatellite markers with charcoal rot disease resistance related indices in 130 different soybean cultivars and lines were evaluated using association analysis based on the general linear model (GLM) and the mixed linear model (MLM) by the Structure and Tassel software. The results of microsatellite markers showed that the genetic structure of the studied population has three subpopulations (K=3) which the results of bar plat also confirmed it. In association analysis based on GLM and MLM models, 31 and 35 loci showed significant relationships with the evaluated traits, respectively, and confirmed considerable variation of the studied traits. The identified markers related to some of the studied traits were the same which can probably be due to pleiotropic effects or tight linkage among the genomic regions controlling these traits. Some of these relationships were including, the relationship between Sat_252 marker with amount of charcoal rot disease, Satt359, Satt190 and Sat_169 markers with number of microsclerota in stem, amount of charcoal rot disease and severity of charcoal rot disease, Sat_416 marker with number of microsclerota in stem and amount of charcoal rot disease and the Satt460 marker with number of microsclerota in stem and severity of charcoal rot disease. The results of this research and the linked microsatellite markers with the charcoal rot disease-related characteristics can be used to identify the suitable parents and to improve the soybean population in future breeding programs.

• Journal of the Architectural Institute of Korea Structure & Construction
• /
• v.36 no.3
• /
• pp.121-128
• /
• 2020

In this paper, the ASCE 7 equivalent static approach for seismic design of non-structural elements is critically evaluated based on the measured floor acceleration data, theory of structural dynamics, and linear/nonlinear dynamic analysis of three-dimensional building models. The analysis of this study on the up-to-date database of the instrumented buildings in California clearly reveals that the measured database does not well corroborate the magnitude and the profile of the floor acceleration as proposed by ASCE 7. The basic flaws in the equivalent static approach are illustrated using elementary structural dynamics. Based on the linear and nonlinear dynamic analyses of three-dimensional case study buildings, it is shown that the magnitude and distribution of the PFA (peak floor acceleration) can significantly be affected by the supporting structural characteristics such as fundamental period, higher modes, structural nonlinearity, and torsional irregularity. In general, the equivalent static approach yields more conservative acceleration demand as building period becomes longer, and the PFA distribution in long-period buildings tend to become constant along the building height due to the higher mode effect. Structural nonlinearity was generally shown to reduce floor acceleration because of its period-lengthening effect. Torsional floor amplification as high as 250% was observed in the building model of significant torsional irregularity, indicating the need for inclusion of the torsional amplification to the equivalent static approach when building torsion is severe. All these results lead to the conclusion that, if permitted, dynamic methods which can account for supporting structural characteristics, should be preferred for rational seismic design of non-structural elements.

• Smart Structures and Systems
• /
• v.18 no.6
• /
• pp.1145-1167
• /
• 2016

This study presents a new method to computes analytical fragility curves of a structure subject to tsunami waves. The method uses dynamic analysis at each stage of the computation. First, the smooth particle hydrodynamics (SPH) model simulates the propagation of the tsunami waves from shallow water to their impact on the target structure. The advantage of SPH over mesh based methods is its capability to model wave surface interaction when large deformations are involved, such as the impact of water on a structure. Although SPH is computationally more expensive than mesh based method, nowadays the advent of parallel computing on general purpose graphic processing unit overcome this limitation. Then, the impact force is applied to a finite element model of the structure and its dynamic non-linear response is computed. When a data-set of tsunami waves is used analytical fragility curves can be computed. This study proves it is possible to obtain the response of a structure to a tsunami wave using state of the art dynamic models in every stage of the computation at an affordable cost.

• Steel and Composite Structures
• /
• v.23 no.5
• /
• pp.597-610
• /
• 2017

This paper presents a mathematical model for fixed-end moments of I-sections with straight haunches for the general case (symmetrical and/or non-symmetrical) subjected to a concentrated load localized anywhere on beam taking into account the bending deformations and shear, which is the novelty of this research. The properties of the cross section of the beam vary along its axis "x", i.e., the flange width "b", the flange thickness "t", the web thickness "e" are constant and the height "d" varies along of the beam, this variation is linear type. The compatibility equations and equilibrium are used to solve such problems, and the deformations anywhere of beam are found by the virtual work principle through exact integrations using the software "Derive" to obtain some results. The traditional model takes into account only bending deformations, and others authors present tables considering the bending deformations and shear, but are restricted. A comparison between the traditional model and the proposed model is made to observe differences, and an example of structural analysis of a continuous highway bridge under live load is resolved. Besides the effectiveness and accuracy of the developed models, a significant advantage is that fixed-end moments are calculated for any cross section of the beam "I" using the mathematical formulas.

• Annals of Clinical Neurophysiology
• /
• v.6 no.2
• /
• pp.75-79
• /
• 2004

Backgrounds: Electrodiagnostic tests have been developed to estimate the degree of facial nerve injury during the acute phase. Side-to-side amplitude comparison with the affected side expressed as a percentage of the nonaffected side has been one of the most valuable electrophysiologic methods of assessing facial nerve functioning. This study was designed to know whether there is any difference in the side-to-side comparison of amplitudes and terminal latencies of the compound muscle action potentials (CMAP) of the facial muscles in the patients with Bell's palsy. Methods: Electroneurographic recordings with surface electrodes on the frontalis, orbicularis oculi, nasalis, and orbicularis oris muscles were made within 2 weeks post-onset (mean, day 7) in 39 patients. Results: Of the 39 Bell's palsy patients, 38 patients (97.4%) recovered satisfactorily within 6 months. The amplitude of CMAP in all patients was not reduced to 10% or less of that of the contralateral healthy muscle. The correlation of amplitude change between four facial muscles was relatively strong, but the correlation of latency change was weak. When the electroneurographic values were compared in the four muscle groups, the general linear models procedure did not show any significant difference for CMAP amplitude and latency changes (p=0.62-0.63). Conclusions: This study did not show any significant clinical advantage of electroneurographic recordings in more than one facial muscle at the early stage of Bell's palsy.

• The Korean Journal of Health Service Management
• /
• v.10 no.3
• /
• pp.123-133
• /
• 2016

Objectives : This study analyzed the effect of health-related characteristics of middle-aged and elderly people on the number of existing permanent teeth by considering each age group through the improvement of health-related characteristics. This study contributes by offering fundamental information on the prevention of dental diseases and on developing an education program that can be used by dental health businesses to improve dental health. Methods : This study used primitive data of the sixth term of the first year and the second year and selected a total of 4,393 people as the final research subject. In order to determine the factors that influence the number of existing natural teeth, confounding factors were considered, and regression analysis using general linear models was done. Results : According to the results of this research, the middle-aged and elderly groups have more existing permanent teeth in the following three cases: 1. Talking is comfortable when daily brushing frequency is 3 times or more. 2. Self-recognition of dental health condition is good. 3. Number of dental caries present in permanent teeth is low. Conclusions : For a and medical system, a dental health care system that takes into consideration the lifespans of people must be legislated and implemented.

• International Journal of Aeronautical and Space Sciences
• /
• v.13 no.2
• /
• pp.127-153
• /
• 2012

This paper presents an advanced computational method for the prediction of the responses in the frequency domain of general linear dissipative structural-acoustic and fluid-structure systems, in the low-and medium-frequency domains and this includes uncertainty quantification. The system under consideration is constituted of a deformable dissipative structure that is coupled with an internal dissipative acoustic fluid. This includes wall acoustic impedances and it is surrounded by an infinite acoustic fluid. The system is submitted to given internal and external acoustic sources and to the prescribed mechanical forces. An efficient reduced-order computational model is constructed by using a finite element discretization for the structure and an internal acoustic fluid. The external acoustic fluid is treated by using an appropriate boundary element method in the frequency domain. All the required modeling aspects for the analysis of the medium-frequency domain have been introduced namely, a viscoelastic behavior for the structure, an appropriate dissipative model for the internal acoustic fluid that includes wall acoustic impedance and a model of uncertainty in particular for the modeling errors. This advanced computational formulation, corresponding to new extensions and complements with respect to the state-of-the-art are well adapted for the development of a new generation of software, in particular for parallel computers.

• Journal of Korean Society on Water Environment
• /
• v.28 no.4
• /
• pp.538-550
• /
• 2012

This study suggests a general methodology which is designed for assessing RDOC behavior at the catchment scale by coupling properly a series of steam flow and water quality simulation models and actual monitoring data set. The modified TANK model in which a river routing function is incorporated to the conventional one is applied to simulate the long-term daily stream flow data, and the simulated stream flow data is combined with the 7-parameter log-linear model coupled to the minimum variance unbiased estimator to simulate the long-term daily water quality (BOD, COD and TOC) loads. Finally, the regression analysis between the usually monitored water quality data (BOD, COD and TOC) and RDOC is combined with the simulated water quality data to manifest the spatio-temporal variability of RDOC flux behavior at the Korean TMDL catchment scale.