• 한국철도학회:학술대회논문집
• /
• 한국철도학회 2011년도 정기총회 및 추계학술대회 논문집
• /
• pp.303-310
• /
• 2011

This paper is focused on development of models for measuring lateral and vertical track irregularities from corresponding accelerometers of an in-service high-speed train. Generally, the track irregularity was measured by a special railway inspection vehicle or system with contact or non-contact sensors. However, the sensors are very expensive and vulnerable to a harsh environment. Displacement estimation from an inertial measurement unit and its wave-band filtering was already developed in the previous study, and it was found that their results included not only the track irregularities but also other information such as phase delay of the applied filters, and suspension and conicity of the wheel. To identify the track irregularities from those results, a compensation filtering method was proposed. Each directional compensation filter was derived by using a system identification method with the estimated directional displacement as input and the corresponding track irregularities as output. In this paper, they are integrated into a model for each direction and applied to the measured lateral and vertical acceleration data from the axle-box and bogie of an in-service high-speed train. Their results are compared with the data from the track geometry measurement system. From the comparison, the proposed models are a useful tool for the measurement of the track irregularities using accelerometers of in-service high-speed trains.

• Communications for Statistical Applications and Methods
• /
• 제23권2호
• /
• pp.131-146
• /
• 2016

In research on behavioral studies, significant attention has been paid to the stage-sequential process for longitudinal data. Latent class profile analysis (LCPA) is an useful method to study sequential patterns of the behavioral development by the two-step identification process: identifying a small number of latent classes at each measurement occasion and two or more homogeneous subgroups in which individuals exhibit a similar sequence of latent class membership over time. Maximum likelihood (ML) estimates for LCPA are easily obtained by expectation-maximization (EM) algorithm, and Bayesian inference can be implemented via Markov chain Monte Carlo (MCMC). However, unusual properties in the likelihood of LCPA can cause difficulties in ML and Bayesian inference as well as estimation in small samples. This article describes and addresses erratic problems that involve conventional ML and Bayesian estimates for LCPA with small samples. We argue that these problems can be alleviated with a small amount of prior input. This study evaluates the performance of likelihood and MCMC-based estimates with the proposed prior in drawing inference over repeated sampling. Our simulation shows that estimates from the proposed methods perform better than those from the conventional ML and Bayesian method.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2002년도 하계학술대회 논문집 D
• /
• pp.2821-2823
• /
• 2002

In this paper we propose the Multi-FNN (Fuzzy-Neural Networks) for optimal identification modeling of complex system. The proposed Multi-FNNs is based on a concept of FNNs and exploit linear inference being treated as generic inference mechanisms. In the networks learning, backpropagation(BP) algorithm of neural networks is used to updata the parameters of the network in order to control of nonlinear process with complexity and uncertainty of data, proposed model use a HCM(Hard C-Means)clustering algorithm which carry out the input-output dat a preprocessing function and Genetic Algorithm which carry out optimization of model The HCM clustering method is utilized to determine the structure of Multi-FNNs. The parameters of Multi-FNN model such as apexes of membership function, learning rates, and momentum coefficients are adjusted using genetic algorithms. An aggregate performance index with a weighting factor is proposed in order to achieve a sound balance between approximation and generalization abilities of the model. NOx emission process data of gas turbine power plant is simulated in order to confirm the efficiency and feasibility of the proposed approach in this paper.

• 한국통신학회논문지
• /
• 제30권2C호
• /
• pp.31-40
• /
• 2005

오류 역전파 알고리즘의 문제점과 ART 신경회로망의 문제점을 개선하기 위해 Jacobs가 제안한 delta-bar-delta 방법과 신경회로망을 결합한 자가 생성 지도 학습 알고리듬을 제안한다. 입력층과 은닉층에서는 ART-1과 ART-2 알고리듬을 이용하고, winner-take-all 방식은 완전 연결 구조이나 연결된 가중치만을 조정하도록 채택하였다. 실험을 위해 학생증, 주민등록증, 컨테이너의 영상으로 추출한 패턴을 신경회로망의 은닉층 노드에 대해 실험하였고, 실험결과 제안된 자기 생성 지도 학습알고리듬이 지역최소화, 학습 속도, 정체 현상이 기존의 방법보다 성능이 개선된 것을 확인하였다.

• 조명전기설비학회논문지
• /
• 제13권1호
• /
• pp.60-68
• /
• 1999

본 논문은 미지이거나 변화하는 산업용 로봇 매니퓰레이터의 추종문제를 다룬다. 제시된 방법은 간접적응 제어기 구조를 근거로 퍼지추론에 의한 스위칭 구조를 통해 다중모델 적응제어기를 설계하고자 한다. 매니퓰레이터 한 개의 식별기를 이용한 다중 식별기들은 미지 관성 파라미터와 부하무게의 초기 추정치를 제외하고는 동일하다. 관절 액츄에이터에 가해진 토크입력은 매순간 로봇 동적시스템의 가장 근접한 식별기 모델에 의해 결정된다. 제시된 방법을 이용하여 용접라인 및 부품조립 자동화 설비 등의 적용 예로서 산업용 로봇 매니퓰레이터의 추종성능이 개선됨을 컴퓨터 모의 실험을 통해 보인다.

• Nuclear Engineering and Technology
• /
• 제43권6호
• /
• pp.573-582
• /
• 2011

The in-cabinet response spectrum is used to define the input motion in the seismic qualification of instruments and devices mounted inside an electrical cabinet. This paper presents a procedure for generating the in-cabinet response spectrum for electrical equipment based on in-situ testing by an impact hammer. The proposed procedure includes an algorithm to build the relationship between the impact forces and the measured acceleration responses of cabinet structures by estimating the state-space model. This model is used to predict seismic responses to the equivalent earthquake forces. Three types of structural model are analyzed for numerical verification of the proposed method. A comparison of predicted and simulated response spectra shows good convergence, demonstrating the potential of the proposed method to predict the response spectra for real cabinet structures using vibration tests. The presented procedure eliminates the uncertainty associated with constructing an analytical model of the electrical cabinet, which has complex mass distribution and stiffness.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2001년도 추계학술대회논문집 II
• /
• pp.1166-1171
• /
• 2001

The method for output disturbance estimation is proposed. In this method, output disturbance is estimated from the closed loop system dynamics using the output and control input signals. In the closed-loop output-disturbance estimator, precise system identification is required to reduce estimation error. The realization of estimator was done by the DSP board (DSPl103), and disturbance estimation in various environments was performed： change of rotation speed, media feature and spindle motor with (or without) auto-ball balancing system (ABS). From these experiments, the disturbance characteristics of ODD under various conditions are analyzed, and the desirable servo loop configuration based these results is proposed.

• 대한방사선기술학회지:방사선기술과학
• /
• 제42권2호
• /
• pp.119-130
• /
• 2019

Positron emission tomography (PET) is widely used imaging modality for effective and accurate functional testing and medical diagnosis using radioactive isotopes. However, PET has difficulties in acquiring images with high image quality due to constraints such as the amount of radioactive isotopes injected into the patient, the detection time, the characteristics of the detector, and the patient's motion. In order to overcome this problem, we have succeeded to improve the image quality by using the dynamic image reconstruction method based on singular value decomposition. However, there is still some question about the characteristics of the proposed technique. In this study, the characteristics of reconstruction method based on singular value decomposition was estimated over computational simulation. As a result, we confirmed that the singular value decomposition based reconstruction technique distinguishes the images well when the signal - to - noise ratio of the input image is more than 20 decibels and the feature vector angle is more than 60 degrees. In addition, the proposed methode to estimate the characteristics of reconstruction technique can be applied to other spatio-temporal feature based dynamic image reconstruction techniques. The deduced conclusion of this study can be useful guideline to apply medical image into SVD based dynamic image reconstruction technique to improve the accuracy of medical diagnosis.

• Steel and Composite Structures
• /
• 제49권4호
• /
• pp.407-417
• /
• 2023

After a disaster like the catastrophic earthquake, the government have to use rapid assessment of the condition (or damage) of bridges, buildings and other infrastructures is mandatory for rapid feedbacks, rescue and post-event management. Many domain schemes based on the measured vibration computations, including least squares estimation and neural fuzzy logic control, have been studied and found to be effective for online/offline monitoring of structural damage. Traditional strategies require all external stimulus data (input data) which have been measured available, but this may not be the generalized for all structures. In this article, a new method with unknown inputs (excitations) is provided to identify structural matrix such as stiffness, mass, damping and other nonlinear parts, unknown disturbances for example. An analytical solution is thus constructed and presented because the solution in the existing literature has not been available. The goals of this paper are towards access to adequate, safe and affordable housing and basic services, promotion of inclusive and sustainable urbanization and participation, implementation of sustainable and disaster-resilient buildings, sustainable human settlement planning and manage. Simulation results of linear and nonlinear structures show that the proposed method is able to identify structural parameters and their changes due to damage and unknown excitations. Therefore, the goal is believed to achieved in the near future by the ongoing development of AI and control theory.

• 정보보호학회논문지
• /
• 제21권1호
• /
• pp.135-141
• /
• 2011

최근 다양한 기능을 탑재한 모바일 장치가 보급되고 개인정보를 다루는 각종 응용들이 등장하면서 사용자에 대한 인증이 중요한 이슈가 되고 있다. 본 논문에서는 기존의 PIN 방식과 더불어 터치스크린 상에서 입력을 하였을 때 얻을 수 있는 터치 위치데이터를 인증에 이용하는 터치 위치기반 인증방법을 제안한다. 본 연구에서 제안하는 방식은 키 패드를 이용하는 기존의 PIN 입력 방식과 동일한 인터페이스를 이용하므로 호환성을 제공하며, 기존 PIN 입력 방식의 안전성을 높이는 방법 중 하나인 행위 분석 방식에 비해 사용자 등록 단계가 간소화되어 편의성을 제공한다. 사용자 인증 실험 결과에 의하면 서로 다른 사용자가 같은 PIN 및 전화번호를 입력한다고 가정할 때 4자리, 6자리 PIN 및 11자리 전화번호에 대해 각각 8.1%, 6.2%, 8.1%의 EER을 나타내었으며, 이를 사용자마다 다른 PIN 및 전화번호를 사용하는 상황에 적용하면 매우 높은 사용자 인식 성능을 보장할 수 있다. 또한 기존 PIN 입력 방식과 동일한 크기의 패스워드(PIN) 탐색 공간을 갖도록 파라미터를 설정한 후 수행한 공격 실험에 의하면 같은 안전성을 가지는 기존의 PIN 입력 방식에 비해 제안한 방식이 매우 높은 안전성을 가짐을 확인할 수 있었다.