• Journal of Power Electronics
• /
• 제17권2호
• /
• pp.476-489
• /
• 2017

Under partial shading conditions (PSCs), multiple maximums may be exhibited on the P-U curve of string inverter photovoltaic (PV) systems. Under such conditions, heuristic methods are invalid for extracting a global maximum power point (GMPP); intelligent algorithms are time-consuming; and model-based methods are complex and costly. To overcome these shortcomings, a novel hybrid MPPT (MPF-IP&O) based on a model-based peak forecasting (MPF) method and an improved perturbation and observation (IP&O) method is proposed. The MPF considers the influence of temperature and does not require solar radiation measurements. In addition, it can forecast all of the peak values of the PV string without complex computation under PSCs, and it can determine the candidate GMPP after a comparison. Hence, the MPF narrows the searching range tremendously and accelerates the convergence to the GMPP. Additionally, the IP&O with a successive approximation strategy searches for the real GMPP in the neighborhood of the candidate one, which can significantly enhance the tracking efficiency. Finally, simulation and experiment results show that the proposed method has a higher tracking speed and accuracy than the perturbation and observation (P&O) and particle swarm optimization (PSO) methods under PSCs.

• 전기학회논문지
• /
• 제67권5호
• /
• pp.656-662
• /
• 2018

Pattern matching algorithm is widely used in many application fields such as bio-informatics, intrusion detection, etc. Among many string matching algorithms, KMP (Knuth-Morris-Pratt) algorithm is commonly used because of its fast execution time when using large texts. However, the processing speed of KMP algorithm is also limited when the text size increases significantly. In this paper, we propose a high throughput parallel KMP algorithm considering CPU-GPU memory hierarchy based on OpenCL in GPGPU (General Purpose computing on Graphic Processing Unit). We focus on the optimization for the allocation of work-times and work-groups, the local memory copy of the pattern data and the failure table, and the overlapping of the data transfer with the string matching operations. The experimental results show that the execution time of the optimized parallel KMP algorithm is about 3.6 times faster than that of the non-optimized parallel KMP algorithm.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제6권6호
• /
• pp.1627-1648
• /
• 2012

We propose an adaptive method for detecting plagiarized pairs from a large set of source code. This method is adaptive in that it uses an adaptive algorithm and it provides an adaptive threshold for determining plagiarism. Conventional algorithms are based on greedy string tiling or on local alignments of two code strings. However, most of them are not adaptive; they do not consider the characteristics of the program set, thereby causing a problem for a program set in which all the programs are inherently similar. We propose adaptive local alignment-a variant of local alignment that uses an adaptive similarity matrix. Each entry of this matrix is the logarithm of the probabilities of the keywords based on their frequency in a given program set. We also propose an adaptive threshold based on the local outlier factor (LOF), which represents the likelihood of an entity being an outlier. Experimental results indicate that our method is more sensitive than JPlag, which uses greedy string tiling for detecting plagiarism-suspected code pairs. Further, the adaptive threshold based on the LOF is shown to be effective, and the detection performance shows high sensitivity with negligible loss of specificity, compared with that using a fixed threshold.

• 산업공학
• /
• 제25권3호
• /
• pp.290-299
• /
• 2012

Keystroke dynamics refers to a way of typing a string of characters. Since one has his/her own typing behavior, one's keystroke dynamics can be used as a distinctive biometric feature for user authentication. In this paper, two authentication algorithms based on keystroke dynamics of long and free texts are proposed. The first is the K-S score, which is based on the Kolmogorov-Smirnov test, and the second is the 'R-A' measure, which combines 'R' and 'A' measures proposed by Gunetti and Picardi (2005). In order to verify the authentication performance of the proposed algorithms, we collected more than 3,000 key latencies from 34 subjects in Korean and 35 subjects in English. Compared with three benchmark algorithms, we found that the K-S score was outstanding when the reference and test key latencies were not sufficient, while the 'R-A' measure was the best when enough reference and test key latencies were provided.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제2권1호
• /
• pp.53-58
• /
• 2002

In the construction of successful fuzzy models for nonlinear systems, the identification of an optimal fuzzy model system is an important and difficult problem. Traditionally, sGA(simple genetic algorithm) has been used to identify structures and parameters of fuzzy model because it has the ability to search the optimal solution somewhat globally. But SGA optimization process may be the reason of the premature local convergence when the appearance of the superior individual at the population evolution. Therefore, in this paper we propose a new method that can yield a successful fuzzy model using VmGA(virus messy genetic algorithms). The proposed method not only can be the countermeasure of premature convergence through the local information changed in population, but also has more effective and adaptive structure with respect to using changeable length string. In order to demonstrate the superiority and generality of the fuzzy modeling using VmGA, we finally applied the proposed fuzzy modeling methodof a complex nonlinear system.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1993년도 하계학술대회 논문집 B
• /
• pp.923-928
• /
• 1993

Genetic algorithms are powerful optimization methods based on the mechanism of natural genetics and natural selection. Genetic algorithms reduce chance of searching local optima unlike most conventional search algorithms and especially show good performances in complex nonlinear optimization problems because they do not require any information except objective function value. This paper presents a new model based on sexual reproduction in nature. In the proposed Sexual Reproduction model(SR model), individuals consist of the diploid of chromosomes, which are artificially coded as binary string in computer program. The meiosis is modeled to produce the sexual cell(gamete). In the artificial meiosis, crossover between homologous chromosomes plays an essential role for exchanging genetic informations. We apply proposed SR model to optimization of the design parameters of Single-sided Linear Induction Motor(SLIM). Sequential Unconstrained Minimization Technique(SUMT) is used to transform the nonlinear optimization problem with many constraints of SLIM to a simple unconstrained problem, We perform optimal design of SLIM available to FA conveyer systems and discuss its results.

• 정보처리학회논문지A
• /
• 제9A권3호
• /
• pp.387-398
• /
• 2002

최근 활발한 소단위 게놈 프로젝트의 수행으로 많은 생물체의 유전체 전체 서열이 밝혀짐에 따라서 전유전체(whole genome)를 기본 단위로 하여 개별 유전자나 그에 관련된 기능 연구가 매우 활발히 이루어지고 있다. 전유전체의 염기 서열은 수백만 bp(base pairs)에서 수백억 bp(base pairs) 정도의 대용량 텍스트 데이터이기 때문에 단순한 온라인 문자 일치(on-line string matching) 알고리즘으로 분석하는 것은 매우 비효율적이다. 본 논문에서는 대용량의 유전체 서열을 분석하는데 적합한 자료 구조인 스트링 B-트리를 사용하여 유전체 서열의 분석과 가시화를 위한 워크벤치를 개발한 과정을 소개한다. 본 연구에서 개발한 시스템은 크게 질의문 부분과 가시화 부분으로 나뉘어 진다. 질의문 부분에는 유전체 서열에 특정 서열이 나타나는 부분의 위치와 횟수를 알아보거나 k번 나타나는 서열을 조사하는 것과 같은 기본적인 패턴 검색 부분과 k-mer 분석을 위한 질의어가 다양하게 준비되어 있다. 가시화 부분은 전유전체 서열과 주석(annotation)을 보여주거나, 유전체 분석을 용이하도록 여러 가시화 방법, CGR(Chaos Game Representation), k-mer graph, RWP(Random Walk Plot) 등으로 생물학자들이 쉽게 전체 구조와 특성 파악할 수 있도록 도와준다. 본 논문이 제안하는 분석 시스템은 생물체의 진화적 관계를 밝히고, 염색체 내에 아직 알려지지 않은 새로운 유전자나 기능이 밝혀지지 않은 junk DNA들의 기능 등을 연구하는데 사용할 수 있다.

• 인터넷정보학회논문지
• /
• 제21권1호
• /
• pp.45-55
• /
• 2020

정보 통신 기술의 발달로 인해 매년 신종/변종 악성코드가 급격히 증가하고 있으며 최근 사물 인터넷과 클라우드 컴퓨팅 기술의 발전으로 다양한 형태의 악성코드가 확산되고 있는 추세이다. 본 논문에서는 운영체제 환경에 관계없이 활용 가능하며 악성행위와 관련된 라이브러리 호출 정보를 나타내는 문자열 정보를 기반으로 한 악성코드 분석 기법을 제안한다. 공격자는 기존 코드를 활용하거나 자동화된 제작 도구를 사용하여 악성코드를 손쉽게 제작할 수 있으며 생성된 악성코드는 기존 악성코드와 유사한 방식으로 동작하게 된다. 악성 코드에서 추출 할 수 있는 대부분의 문자열은 악성 동작과 밀접한 관련이 있는 정보로 구성되어 있기 때문에 텍스트 마이닝 기반 방식을 활용하여 데이터 특징에 가중치를 부여해 악성코드 분석을 위한 효과적인 Feature로 가공한다. 가공된 데이터를 기반으로 악성여부 탐지와 악성 그룹분류에 대한 실험을 수행하기 위해 다양한 Machine Learning 알고리즘을 이용해 모델을 구축한다. 데이터는 Windows 및 Linux 운영체제에 사용되는 파일 모두에 대해 비교 및 검증하였으며 악성탐지에서는 약93.5%의 정확도와 그룹분류에서는 약 90%의 정확도를 도출하였다. 제안된 기법은 악성 그룹을 분류시 각 그룹에 대한 모델을 구축할 필요가 없기 때문에 단일 모델로서 비교적 간단하고 빠르며 운영체제와 독립적이므로 광범위한 응용 분야를 가진다. 또한 문자열 정보는 정적분석을 통해 추출되므로 코드를 직접 실행하는 분석 방법에 비해 신속하게 처리가능하다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1999년도 하계종합학술대회 논문집
• /
• pp.942-945
• /
• 1999

In this work, we proposed Proper etching algorithm for ultra-large scale integrated circuit device and simulated etching process using the proposed algorithm in the case of ICP (inductive coupled plasma) 〔1〕source. Until now, many algorithms for etching process simulation have been proposed such as Cell remove algorithm, String algorithm and Ray algorithm. These algorithms have several drawbacks due to analytic function; these algorithms are not appropriate for sub 0.1 ${\mu}{\textrm}{m}$ device technologies which should deal with each ion. These algorithms could not present exactly straggle and interaction between Projectile ions and could not consider reflection effects due to interactions among next projectile ions, reflected ions and sputtering ions, simultaneously In order to apply ULSI process simulation, algorithm considering above mentioned interactions at the same time is needed. Proposed algorithm calculates interactions both in plasma source region and in target material region, and uses BCA (binary collision approximation4〕method when ion impact on target material surface. Proposed algorithm considers the interaction between source ions in sheath region (from Quartz region to substrate region). After the collision between target and ion, reflected ion collides next projectile ion or sputtered atoms. In ICP etching, because the main mechanism is sputtering, both SiO$_2$ and Si can be etched. Therefore, to obtain etching profiles, mask thickness and mask composition must be considered. Since we consider both SiO$_2$ etching and Si etching, it is possible to predict the thickness of SiO$_2$ for etching of ULSI.

• 한국CDE학회논문집
• /
• 제5권2호
• /
• pp.127-135
• /
• 2000

Emergent computing paradigms such as genetic algorithms have found increased use in problems in engineering design. These computational tools have been shown to be applicable in the solution of generically difficult design optimization problems characterized by nonconvexities in the design space and the presence of discrete and integer design variables. Another aspect of these computational paradigms that have been lumped under the bread subject category of soft computing, is the domain of artificial intelligence, knowledge-based expert system, and machine learning. The paper explores a machine learning paradigm referred to as teaming classifier systems to construct the high-quality global function approximations between the design variables and a response function for subsequent use in design optimization. A classifier system is a machine teaming system which learns syntactically simple string rules, called classifiers for guiding the system's performance in an arbitrary environment. The capability of a learning classifier system facilitates the adaptive selection of the optimal number of training data according to the noise and multimodality in the design space of interest. The present study used the polynomial based response surface as global function approximation tools and showed its effectiveness in the improvement on the approximation performance.