• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제10권9호
• /
• pp.339-348
• /
• 2021

최근 소프트웨어 결함 예측 연구는 교차 프로젝트 간의 결함 예측뿐만 아니라 교차 버전 프로젝트 간의 결함 예측 또한 이루어지고 있다. 종래의 교차 버전 결함 예측 연구들은 WP(Within-Project)로 가정한다. 하지만, CV(Cross-Version) 환경에서는 프로젝트 버전 간의 분포 차이의 중요성을 고려한 연구들이 없다. 본 연구에서는 다른 버전 간의 분포 차이까지 고려하는 자동화된 베이지안 최적화 프레임워크를 제안한다. 이를 통해 분포차이에 따라 전이 학습(Transfer Learning) 수행 여부를 자동으로 선택하여 준다. 해당 프레임워크는 버전 간의 분포 차이, 전이 학습과 분류기(Classifier)의 하이퍼파라미터를 최적화하는 기법이다. 실험을 통해 전이 학습 수행 여부를 분포차 기준으로 자동으로 선택하는 방법이 효과적이라는 것을 알 수 있다. 그리고 최적화를 이용하는 것이 성능 향상에 효과가 있으며 이러한 결과 소프트웨어 인스펙션 노력을 감소할 수 있다는 것을 확인할 수 있다. 이를 통해 교차 버전 프로젝트 환경에서 신규 버전 프로젝트에 대하여 효과적인 품질 보증 활동 수행을 지원할 것으로 기대된다.

• Advances in Automotive Engineering
• /
• 제1권1호
• /
• pp.41-59
• /
• 2018

In the study, investigation of fiber- reinforced composite materials that can be an alternative to conventional steel was performed by finite element analysis with the help of software. Steel and composite materials have been studied on a four axle truck chassis model. Three-dimensional finite element model was created with software, and then analyzes were performed. The analyses were performed for static and dynamic/fatigue cases. Fatigue cases are formed with the help of design spectra model and fatigue analyses were performed as static analyses with this design spectra. First, analyses were performed for steel and after that optimization analyses were made for the AS4-PEEK carbon fiber composite and Eglass-Epoxy fiber composite materials. Optimization of composite material analyzes include determining the total laminate thickness, thickness of each ply, orientation of each ply and ply stacking sequence. Analyzes were made according to macro mechanical properties of composite, micromechanics case has not been considered. Improvements in weight reduction up to %50 provided at the end of the composite optimization analyzes with satisfying stiffness performance of chassis. Fatigue strength of the composite structure depends on various factors such as, fiber orientation, ply thickness, ply stack sequence, fiber ductility, ductility of the matrix, loading angle. Therefore, the accuracy of theoretical calculations and analyzes should be correlated by testing.

• Journal of Advanced Marine Engineering and Technology
• /
• 제40권10호
• /
• pp.935-941
• /
• 2016

Predicting and evaluating ship manoeuvring characteristics are very important not only for the design stage, but also for the existing vessels. There are several ways to predict ship's manoeuvrability and most of them are highly connected with the estimation of hydrodynamic coefficients. This paper presents a new estimation method using the system identification with mathematical algorithms for estimating hydrodynamic coefficient in the ship's mathematical model. Specifically a double ended ferry which equips four azimuth propulsion systems were chosen as benchmark ship and a set of benchmark data which is generated in the fast time simulation software was provided to conduct mathematical optimization process. Also the initial values for the optimization were borrowed from the empirical regression formulas of the simulation software of Rheinmetall Defence ship simulator. Therefore the newly suggested mathematical optimization algorithm gave a successful result for estimation hydrodynamic coefficients. Proper optimization conditions of the objective function and constraints were also verified during the study.

• International Journal of Aeronautical and Space Sciences
• /
• 제15권4호
• /
• pp.383-395
• /
• 2014

According to the requirement of wing weight estimation and frequent adjustments during aircraft conceptual design, a wing weight estimation method considering the constraints of structural strength and stiffness is proposed to help designers make wing weight estimations rapidly and accurately. This method implements weight predictions on the basis of structure weight optimization with stiffness constraints and strength constraints, which include achievement of wing shape parametric modeling, rapid structure layout, finite element (FE) model automated generation, load calculation, structure analysis, weight optimization, and weight computed based on modeling. A software tool is developed with this wing weight estimation method. This software can realize the whole process of wing weight estimation with the method and the workload of wing weight estimation is reduced because much of the work can be completed by the software. Finally, an example is given to illustrate that this weight estimation method is effective.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권12호
• /
• pp.4081-4098
• /
• 2022

With the aim of tackling the contradiction between computation intensive industrial applications and resource-weak Edge Devices (EDs) in Industrial Internet of Things (IIoT), a novel computation task offloading scheme in SDIN-enabled MEC based IIoT is proposed in this paper. With the aim of reducing the task accomplished latency and energy consumption of EDs, a joint optimization method is proposed for optimizing the local CPU-cycle frequency, offloading decision, and wireless and computation resources allocation jointly. Based on the optimization, the task offloading problem is formulated into a Mixed Integer Nonlinear Programming (MINLP) problem which is a large-scale NP-hard problem. In order to solve this problem in an accessible time complexity, a sub-optimal algorithm GPCOA, which is based on hybrid evolutionary computation, is proposed. Outcomes of emulation revel that the proposed method outperforms other baseline methods, and the optimization result shows that the latency-related weight is efficient for reducing the task execution delay and improving the energy efficiency.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2001년도 춘계학술대회논문집C
• /
• pp.565-571
• /
• 2001

As a method of structural optimization, a practical algorithm based on the Taguchi method is developed. The Taguchi method is applied iteratively updating the level values of design variables. The design region is translated or reduced during optimization and by appropriate choice of reduction factor and initial level intervals, a near-optimum solution can be found very efficiently. To treat inequality constraints, a variable penalty method is utilized. A software system named 'DS/Taguchi' is developed by integrating the proposed algorithm and commercial finite element analysis codes on the parametric CAD platform. Two examples are taken to examine the performance of the proposed algorithm and the developed software system.

• ETRI Journal
• /
• 제32권5호
• /
• pp.795-800
• /
• 2010

Embedding a large vocabulary speech recognition system in mobile devices requires a reduced acoustic model obtained by eliminating redundant model parameters. In conventional optimization methods based on the minimum description length (MDL) criterion, a binary Gaussian tree is built at each state of a hidden Markov model by iteratively finding and merging similar mixture components. An optimal subset of the tree nodes is then selected to generate a downsized acoustic model. To obtain a better binary Gaussian tree by improving the process of finding the most similar Gaussian components, this paper proposes a new distance measure that exploits the difference in likelihood values for cases before and after two components are combined. The mixture weight of Gaussian components is also introduced in the component merging step. Experimental results show that the proposed method outperforms MDL-based optimization using either a Kullback-Leibler (KL) divergence or weighted KL divergence measure. The proposed method could also reduce the acoustic model size by 50% with less than a 1.5% increase in error rate compared to a baseline system.

• Industrial Engineering and Management Systems
• /
• 제14권1호
• /
• pp.11-21
• /
• 2015

The purpose of this research is to develop algorithms using the Differential Evolution Algorithm (DE) to solve a multi-objective, sources and stages location-allocation problem. The development process starts from the design of a standard DE, then modifies the recombination process of the DE in order improve the efficiency of the standard DE. The modified algorithm is called modified DE. The proposed algorithms have been tested with one real case study (large size problem) and 2 randomly selected data sets (small and medium size problems). The computational results show that the modified DE gives better solutions and uses less computational time than the standard DE. The proposed heuristics can find solutions 0 to 3.56% different from the optimal solution in small test instances, while differences are 1.4-3.5% higher than that of the lower bound generated by optimization software in medium and large test instances, while using more than 99% less computational time than the optimization software.

• ETRI Journal
• /
• 제31권2호
• /
• pp.121-128
• /
• 2009

In this paper, we describe a fixed-threshold sequential minimal optimization (FSMO) for structured SVM problems. FSMO is conceptually simple, easy to implement, and faster than the standard support vector machine (SVM) training algorithms for structured SVM problems. Because FSMO uses the fact that the formulation of structured SVM has no bias (that is, the threshold b is fixed at zero), FSMO breaks down the quadratic programming (QP) problems of structured SVM into a series of smallest QP problems, each involving only one variable. By involving only one variable, FSMO is advantageous in that each QP sub-problem does not need subset selection. For the various test sets, FSMO is as accurate as an existing structured SVM implementation (SVM-Struct) but is much faster on large data sets. The training time of FSMO empirically scales between O(n) and O($n^{1.2}$), while SVM-Struct scales between O($n^{1.5}$) and O($n^{1.8}$).

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제18권3호
• /
• pp.633-652
• /
• 2024

Software testing is an important phase as it ensures the software quality. The software testing process comprises of three steps: generation, execution, and evaluation of test cases. Literature claims the usage of single and multiple 'Unified Modeling Language' (UML) diagrams to generate test cases. Using multiple UML diagrams increases test case coverage. However, the existing approaches show limitations in test case generation from UML diagrams. Therefore, in this research study, we propose an approach to generate the test cases using UML State Chart Diagram (SCD), Activity Diagram (AD), and Sequence Diagram (SD). The proposed approach transforms UML diagrams into intermediate forms: SCD Graph, AD Graph, and SD Graph respectively. Furthermore, by integrating these three graphs, a System Testing Graph (STG) is formed. Finally, test cases are identified from STG by using a traversal algorithm such as Depth First Search (DFS) that is an optimization method. The results show that the proposed approach is better compared to existing approaches in terms of coverage and performance. Moreover, the generated test cases have the ability to detect faults at the unit level, integration, and system level testing.