• 소음진동
• /
• 제11권2호
• /
• pp.336-345
• /
• 2001

This research presents a nonlinear model to analyze the ball bearing nitration due to the waviness in a rigid rotor supported by multi-row ball bearings. The waviness of a ball and each races is modeled by the superposition of sinusoidal function, and the position vectors of inner and outer groove radius center are defined with respect to the mass center of the rotor in order to consider five degrees of freedom of a general rotor-bearing system. The waviness of a ball bearing is introduced to these position vectors to use the Hertzian contact theory in order to calculate the elastic deflection and nonlinear contact force resulting from the waviness while the rotor has translational and angular motion. They can be determined by solving the nonlinear equations of motion with five degrees of freedom by using the Runge-Kutta-Fehlberg algorithm. Numerical results of this research are validated with those of prior researchers. The proposed model can calculate the translational displacement as well as the angular displacement of the rotor supported by the multi-row ball bearings with waviness. It also characterizes the nitration frequencies resulting from the various kinds of waviness in rolling elements, the harmonic frequencies resulting from the nonlinear load-deflection characteristics of ball bearing. and the sideband frequencies resulting from the waviness interaction.

• 한국감성과학회:학술대회논문집
• /
• 한국감성과학회 2002년도 춘계학술대회 논문집
• /
• pp.195-200
• /
• 2002

ART (Adaptive Resonance Theory [1]) neural network and its variations perform non-hierarchical clustering by unsupervised learning. We propose a scheme "arboART" for hierarchical clustering by using several ART1.5-SSS networks. It classifies multidimensional vectors as a cluster tree, and finds features of clusters. The Basic idea of arboART is to use the prototype formed in an ART network as an input to other ART network that has looser distance criteria (Ishihara, et al., [2,3]). By sending prototype vectors made by ART to one after another, many small categories are combined into larger and more generalized categories. We can draw a dendrogram using classification records of sample and categories. We have confirmed its ability using standard test data commonly used in pattern recognition community. The clustering result is better than traditional computing methods, on separation of outliers, smaller error (diameter) of clusters and causes no chaining. This methodology is applied to Kansei evaluation experiment data analysis.

• Journal of Information Science Theory and Practice
• /
• 제12권3호
• /
• pp.36-48
• /
• 2024

The information retrieval (IR) process often encounters a challenge known as query-document vocabulary mismatch, where user queries do not align with document content, impacting search effectiveness. Automatic query expansion (AQE) techniques aim to mitigate this issue by augmenting user queries with related terms or synonyms. Word embedding, particularly Word2Vec, has gained prominence for AQE due to its ability to represent words as real-number vectors. However, AQE methods typically expand individual query terms, potentially leading to query drift if not carefully selected. To address this, researchers propose utilizing median vectors derived from deep median networks to capture query similarity comprehensively. Integrating median vectors into candidate term generation and combining them with the BM25 probabilistic model and two IR strategies (EQE1 and V2Q) yields promising results, outperforming baseline methods in experimental settings.

• International Journal of Control, Automation, and Systems
• /
• 제5권2호
• /
• pp.111-116
• /
• 2007

The design problem of partial pole assignment (PPA) in two classes of frequency-domain MIMO models by constant gain feedback is investigated in this paper. Its aim is to design a constant gain feedback which changes only a subset of the open-loop eigenvalues, while the rest of them are kept unchanged in the closed-loop system. A near general parametric expression for the feedback gain matrix in term of a set of design parameter vectors and the set of the closed-loop poles, and a simple parametric approach for solving the proposed problem are presented. The set of poles do not need to be previously prescribed, and can be set undetermined and treated together with the set of parametric vectors as degrees of design freedom provided by the approach. An illustrative example shows that the proposed parametric method is simple and effective.

• 한국생물정보학회:학술대회논문집
• /
• 한국생물정보시스템생물학회 2005년도 BIOINFO 2005
• /
• pp.23-28
• /
• 2005

The high-dimensionality and insufficiency of gene expression profiles and proteomic profiles makes feature selection become a critical step in efficiently building accurate models for cancer problems based on such data sets. In this paper, we use a method, called Discrete Function Learning algorithm, to find discriminatory feature vectors based on information theory. The target feature vectors contain all or most information (in terms of entropy) of the class attribute. Two data sets are selected to validate our approach, one leukemia subtype gene expression data set and one ovarian cancer proteomic data set. The experimental results show that the our method generalizes well when applied to these insufficient and high-dimensional data sets. Furthermore, the obtained classifiers are highly understandable and accurate.

• Journal of Power Electronics
• /
• 제19권2호
• /
• pp.443-453
• /
• 2019

As the core component of transmission systems, converters are very prone to failure. To improve the accuracy of fault diagnosis for wind power converters, a fault feature extraction method combined with a wavelet transform and compressed sensing theory is proposed. In addition, an improved AdaBoost-SVM is used to diagnose wind power converters. The three-phase output current signal is selected as the research object and is processed by the wavelet transform to reduce the signal noise. The wavelet approximation coefficients are dimensionality reduced to obtain measurement signals based on the theory of compressive sensing. A sparse vector is obtained by the orthogonal matching pursuit algorithm, and then the fault feature vector is extracted. The fault feature vectors are input to the improved AdaBoost-SVM classifier to realize fault diagnosis. Simulation results show that this method can effectively realize the fault diagnosis of the power transistors in converters and improve the precision of fault diagnosis.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2005년도 학술대회 논문집 정보 및 제어부문
• /
• pp.632-634
• /
• 2005

In this paper, The Frenet-Serret formula of classical geometric curve theory with the concept of a missile pointing velocity vector are used to analyze and design a missile guidance law. The capture capability of this guidance law is qualitatively studied by comparing the rotations of the velocity vectors of missile and target relative to the line of sight vector. when fuzzy Table look-up theory applied in target-missile distance & angle displacement, this research. It's performance is better then classical research.

• 대한기계학회논문집A
• /
• 제29권10호
• /
• pp.1321-1328
• /
• 2005

Based upon the Floquet theory, the complex modal solution for general rotor systems with periodically time-varying parameters is newly derived. The complete modal response can be obtained from the orthonormality condition between the time-variant eigenvectors and the corresponding adjoint vectors. The harmonic solutions such as the response and directional special a patterns are then derived in terms of harmonic modes whose coefficients are obtained from the modal analysis. The stability analysis by the Floquet's transition matrix and the eigen-analysis is also performed.

• Structural Engineering and Mechanics
• /
• 제74권1호
• /
• pp.33-54
• /
• 2020

The aim of this study is to calculate natural frequencies and harmonic responses of cracked frames with general boundary conditions by using transfer matrix method (TMM). The TMM is a straightforward technique to obtain harmonic responses and natural frequencies of frame structures as the method is based on constructing a relationship between state vectors of two ends of structure by a chain multiplication procedure. A single variable shear deformation theory (SVSDT) is applied, as well as, Timoshenko beam theory (TBT) and Euler-Bernoulli beam theory (EBT) for comparison purposes. Firstly, free vibration analysis of intact and cracked frames are performed for different crack ratios using TMM. The crack is modelled by means of a linear rotational spring that divides frame members into segments. The results are verified by experimental data and finite element method (FEM) solutions. The harmonic response curves that represent resonant and anti-resonant frequencies directly are plotted for various crack lengths. It is seen that the TMM can be used effectively for harmonic response analysis of cracked frames as well as natural frequencies calculation. The results imply that the SVSDT is an efficient alternative for investigation of cracked frame vibrations especially with thick frame members. Moreover, EBT results can easily be obtained by ignoring shear deformation related terms from governing equation of motion of SVSDT.

• 한국광학회:학술대회논문집
• /
• 한국광학회 2003년도 제14회 정기총회 및 03년 동계학술발표회
• /
• pp.22-23
• /
• 2003

The new method of x-ray photon correlation spectroscopy (XPCS) [1] is developed recently for probing the dynamics of surface height fluctuations as a function of lateral length scale. Measurements were performed on supported polystyrene (PS) films of thicknesses varying from 84 to 333 nm at temperatures above the PS glass transition temperature. Within a range of wave vectors spanning $10^{-3}$ to $10^{-2}$ $nm^{-1}$, good agreement is found between the measured surface dynamics and the theory of overdamped thermal capillary waves on thin films. Quantitatively, the data can be accounted for using the viscosity of bulk PS. (omitted)