• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제5권4호
• /
• pp.297-301
• /
• 2005

In this paper, we propose a method to estimate camera motion parameter based on invariant point features. Typically, feature information of image has drawbacks, it is variable to camera viewpoint, and therefore information quantity increases after time. The LM(Levenberg-Marquardt) method using nonlinear minimum square evaluation for camera extrinsic parameter estimation also has a weak point, which has different iteration number for approaching the minimal point according to the initial values and convergence time increases if the process run into a local minimum. In order to complement these shortfalls, we, first propose constructing feature models using invariant vector of geometry. Secondly, we propose a two-stage calculation method to improve accuracy and convergence by using homography and LM method. In the experiment, we compare and analyze the proposed method with existing method to demonstrate the superiority of the proposed algorithms.

• Journal of Electrical Engineering and Technology
• /
• 제13권1호
• /
• pp.192-200
• /
• 2018

In this study, improvements for error correction, speed, position, and rotation calculation algorithms have been proposed to be used in resolver to digital conversion (RDC) systems. The proposed open-loop system drives the resolver and uses the output signals of the resolver signal to estimate the real time position, the instant speed, and the rotation count with high resolution and accuracy even at high speeds and noise. The proposed solution implements strong features of both closed and open loop based systems while eliminating their weak points. The improvements proposed is resistant to noise owing to digital FIR filter and data averaging techniques. The implementation used for proof of concept is implemented on a hardware using an FPGA and configurable to be used by any resolver.

• 한국정보통신학회논문지
• /
• 제7권7호
• /
• pp.1462-1470
• /
• 2003

정보통신의 발달과 인터넷의 확산으로 인해 정보보안의 필요성이 증대되면서 다양한 암호알고리즘이 개발되어 활용되고 있다. 이와 더불어 암호 공격 기술도 발전하여서, 공격에 강한 알고리즘에 대한 연구가 활발하게 진행되고 있다. Substitution Permutation Networks(SPN)등의 블록 암호알고리즘에서 교환계층의 선형변환행렬이 Maximum Distance Separable(MDS) 코드를 생성하면 차분공격과 선형공격에 강한 특성을 보인다. 본 논문에서는 선형변환행렬이 MDS 코드를 생성하는가를 판단하는 새로운 알고리즘을 제안한다. 선행변환행렬의 입력코드는 GF(2$^n$)상의 원소들로 이들을 변수로 해석할 수 있다. 하나의 변수를 다른 변수들의 대수식으로 변환하고 대입하여 변수를 하나씩 소거한다. 변수가 하나이고 모든 계수가 ‘0’이 아니면 선형변환 행렬은 MDS 코드를 생성한다. 본 논문에서 제안한 알고리즘은 기존의 모든 정방부분행렬이 정칙인지를 판단하는 알고리즘과 비교하여 곱셈 및 역수 연산수를 많이 줄임으로서 수행 시간을 크게 감소 시켰다.

• 방사선산업학회지
• /
• 제17권4호
• /
• pp.345-357
• /
• 2023

Nuclear forensics has been understood as a mendatory component in the international society for nuclear material control and non-proliferation verification. Radiochronometry of nuclear activities for nuclear forensics are decay series characteristics of nuclear materials and the Bateman equation to estimate when nuclear materials were purified and produced. Radiochronometry values have uncertainty of measurement due to the uncertainty factors in the estimation process. These uncertainties should be calculated using appropriate evaluation methods that are representative of the accuracy and reliability. The IAEA, US, and EU have been researched on radiochronometry and uncertainty of measurement, although the uncertainty calculation method using the Bateman equation is limited by the underestimation of the decay constant and the impossibility of estimating the age of more than one generation, so it is necessary to conduct uncertainty calculation research using computer simulation such as Monte Carlo method. This highlights the need for research using computational simulations, such as the Monte Carlo method, to overcome these limitations. In this study, we have analyzed mathematical models and the LHS (Latin Hypercube Sampling) methods to enhance the reliability of radiochronometry which is to develop an uncertainty algorithm for nuclear material radiochronometry using Bateman Equation. We analyzed the LHS method, which can obtain effective statistical results with a small number of samples, and applied it to algorithms that are Monte Carlo methods for uncertainty calculation by computer simulation. This was implemented through the MATLAB computational software. The uncertainty calculation model using mathematical models demonstrated characteristics based on the relationship between sensitivity coefficients and radiative equilibrium. Computational simulation random sampling showed characteristics dependent on random sampling methods, sampling iteration counts, and the probability distribution of uncertainty factors. For validation, we compared models from various international organizations, mathematical models, and the Monte Carlo method. The developed algorithm was found to perform calculations at an equivalent level of accuracy compared to overseas institutions and mathematical model-based methods. To enhance usability, future research and comparisons·validations need to incorporate more complex decay chains and non-homogeneous conditions. The results of this study can serve as foundational technology in the nuclear forensics field, providing tools for the identification of signature nuclides and aiding in the research, development, comparison, and validation of related technologies.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제29권1_2호
• /
• pp.98-113
• /
• 2002

본 논문에서는 블록 정합 알고리즘(BMA: block matching algorithm)인 다단계 연속 제거 알고리즘(MSEA: multi-level successive elimination algorithm)[1]의 연산량을 줄이기 위하여 네 가지 방안을 제안하였다. 첫 번째 제안 방안은 MSEA에서 서브 블록(sub block)의 합 놈(sum norm)에 대한 절대 오차의 합(SAD: sum of absolute difference)을 계산할 때 부분 왜곡 제거(PDE: partial distortion elimination) 기법을 적용하여 연산량을 감소시킨 알고리즘이다. 두 번째 제안 방안인 적응 SAD 계산 알고리즘은 SAD 계산 시 절대 오차가 큰 값에서부터 작은 값의 순으로 SAD를 계산하면 PDE가 빨리 발생하게 되어 연산량을 줄일 수 있는 성질을 이용한 알고리즘이다. 세 번째 제안 방안인 제거 레벨 추정 알고리즘은 탐색점의 제거 레벨을 추정하고 추정된 레벨에서부터 상위 레벨로 다단계 연속 제거 과정을 수행함으로 추정된 제거레벨보다 낮은 레벨들과 연관된 연산량을 감소시킨 알고리즘이다. 제안된 첫 번째, 두 번째, 세 번째 방안은 움직임 추정의 정확도가 전역 탐색 알고리즘(FSA: full search algorithm) 및 MSEA와 동일하면서 MSEA의 연산량을 효과적으로 감소시킨 알고리즘들이다. 네 번째 제안 방안인 나선형 다이아몬드 그물 탐색 알고리즘은 움직임 추정의 정확도가 거의 100%이면서 움직임 추정에 필요한 연산량을 획기적으로 감소시킨 고속 블록 정합 알고리즘이다. 위의 네 가지 제안 방안에 대한 성능을 평가하기 위하여 실험을 수행하였으며 실험에서 제안 방안들의 효율성을 확인하였다.

• 대한조선학회논문집
• /
• 제38권2호
• /
• pp.33-43
• /
• 2001

드릴쉽과 셔틀탱커등 특수선의 건조가 활발해지면서 정해진 위치에서 장시간 작업해야 할 목적으로 동적 자기위치 유지시스템(Dynamic Positioning System)을 장착하는 선박이 늘어나고 있다. 본 논문에서는 DP 시스템의 구성 요소를 소개하고, 수조에서 실험한 결과를 바탕으로 제어이론과 필터이론에 따른 DP성능을 비교하였다. 실험에 사용한 선박은 10만톤급 드릴쉽으로 모델의 길이가 4m이며, 방향이 고정된 3개의 추진기를 사용하였다. 실험 내용은 명령에 따라 주어진 궤적을 이동하는 능력과 파도가 있는 외란조건에서 원하는 선수각을 유지할 수 있는지를 살펴보았다. 추진력을 구하기 위하여 PID 이론을 적용하였고, 제어게인의 변화에 따른 제어특성을 살펴보았다. 선형운동성분과 계측잡음을 제거하기 위하여 칼만 필터와 디지털 필터를 적용하였고, 각각의 필터성능을 비교 검토하였다.

• 전자공학회논문지SC
• /
• 제43권2호
• /
• pp.11-19
• /
• 2006

본 논문에서는 사영과 치환불변 점 특징을 기반으로 카메라의 외부인수를 산출하는 방법을 제안한다. 기존 연구에서의 특징 정보들은 카메라의 뷰 포인트에 따라 변화하기 때문에 대응점 산출이 어렵다. 따라서 본 논문에서는 카메라 위치에 무관한 불변 점 특징을 추출하고 시간 복잡도 감소와 정확한 대응점 산출을 위해 유사도 평가함수와 Graham 탐색 방법을 이용한 새로운 정합방법을 제안한다. 또한 카메라 외부인수 산출단계에서는 LM 알고리즘의 수렴도를 향상시키기 위해 2단계 카메라 동작인수 산출방법을 제안한다. 실험에서는 다양한 실내영상을 이용하여 기존방법과 비교, 분석함으로써 제안한 알고리즘의 우수성을 입증하였다.

• 대한기계학회논문집B
• /
• 제27권6호
• /
• pp.753-765
• /
• 2003

A finite element code for the numerical solution of the Navier-Stokes equation is parallelized by vertex-oriented domain decomposition. To accelerate the convergence of iterative solvers like conjugate gradient method, parallel block ILU, iterative block ILU, and distributed ILU methods are tested as parallel preconditioners. The effectiveness of the algorithms has been investigated when P1P1 finite element discretization is used for the parallel solution of the Navier-Stokes equation. Two-dimensional and three-dimensional Laplace equations are calculated to estimate the speedup of the preconditioners. Calculation domain is partitioned by one- and multi-dimensional partitioning methods in structured grid and by METIS library in unstructured grid. For the domain-decomposed parallel computation of the Navier-Stokes equation, we have solved three-dimensional lid-driven cavity and natural convection problems in a cube as benchmark problems using a parallelized fractional 4-step finite element method. The speedup for each parallel preconditioning method is to be compared using upto 64 processors.

• Structural Engineering and Mechanics
• /
• 제62권6호
• /
• pp.709-717
• /
• 2017

This paper develops a two-stage method for structural damage identification by using modal data. First, the Residual Force Vector (RFV) is introduced to detect any potentially damaged elements of structures. Second, data of the frequency domain are used to build up the objective function, and then the Imperialist Competitive Algorithm (ICA) is utilized to estimate damaged extents. ICA is a heuristic algorithm with simple structure, which is easy to be implemented and it is effective to deal with high-dimension nonlinear optimization problem. The advantages of this present method are: (1) Calculation complexity can be decreased greatly after eliminating many intact elements in the first step. (2) Robustness, ICA ensures the robustness of the proposed method. Various damaged cases and different structures are investigated in numerical simulations. From these results, anyone can point out that the present algorithm is effective and robust for structural damage identification and is also better than many other heuristic algorithms.

• 방사성폐기물학회지
• /
• 제20권2호
• /
• pp.171-183
• /
• 2022

The parent and daughter nuclides in a radioactive decay chain arrive at secular equilibrium once they have a large half-life difference. The characteristics of this equilibrium state can be used to estimate the production time of nuclear materials. In this study, a mathematical model and algorithm that can be applied to radio-chronometry using the radioactive equilibrium relationship were investigated, reviewed, and implemented. A Bateman equation that can analyze the decay of radioactive materials over time was used for the mathematical model. To obtain a differential-based solution of the Bateman equation, an algebraic numerical solution approach and two different matrix exponential functions (Moral and Levy) were implemented. The obtained result was compared with those of commonly used algorithms, such as the Chebyshev rational approximation method and WISE Uranium. The experimental analysis confirmed the similarity of the results. However, the Moral method led to an increasing calculation uncertainty once there was a branching decay, so this aspect must be improved. The time period corresponding to the production of nuclear materials or nuclear activity can be estimated using the proposed algorithm when uranium or its daughter nuclides are included in the target materials for nuclear forensics.