• 한국시뮬레이션학회논문지
• /
• 제9권1호
• /
• pp.93-108
• /
• 2000

The stochastic manufacturing system has one or more random variables as inputs that lead to random outputs. Since the outputs are random, they can be considered only as estimates of the true characteristics of the system. These estimates could greatly differ from the corresponding real characteristics for the system. Multiple replications are necessary to get reliable information on the system and output data should be analyzed to get optimal solution. It requires too much computation time practically, In this paper a GA method, named Stochastic Genetic Algorithm(SGA) is proposed and tested to find the optimal solution fast and efficiently by reducing the number of replications.

• 소성∙가공
• /
• 제13권1호
• /
• pp.27-32
• /
• 2004

Thermomechanical behavior of Al-Mg-Si alloys was studied to investigate the effect of microstructural features such as pre-existing substructure and distribution of particles on the deformation characteristics. The controlled compression tests were carried out to get the information on how the alloy responds to temperature, strain amount and strain rate. Then hot forging of Al-Mg-Si alloys carried out and analyzed by the comparison with the compression tests. Microstructural features after forging were discussed in terms of the thermomechanical response of Al-Mg-Si alloys. As already well mentioned, we found that the deformation of Al-Mg-Si at the elevated temperature brought the recovered structure on most conditions. In a certain time, however, abnormally large grains were found as a result of deformation assisted grain growth, which means that hot forging of Al-Mg-Si alloys could lead to a undesirable microstructural variation and the consequent mechanical properties such as fatigue strength.

• 아태비즈니스연구
• /
• 제2권2호
• /
• pp.1-35
• /
• 2011

Facing information overload in today's complex environments, managers look to a concise set of marketing metrics to provide direction for marketing decision making. While there have been several papers dealing with the theoretical aspects of dashboard creation, no research creates and tests a dashboard using scientific techniques. This study develops and demonstrates an empirical approach to dashboard metric selection. In a fast moving consumer goods category, this research selects leading indicators for national-brand and store-brand sales and revenue premium performance from 99 brand-specific and relative-to-competition variables including price, brand equity, usage occasions, and multiple measures of awareness, trial/usage, purchase intent, and liking/satisfaction. Plotting impact size and wear-in time reveals that different kinds of variables predict sales at distinct lead times, which implies that managerial action may be taken to turn the metrics around before performance itself declines.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2006년도 춘계학술대회 논문집
• /
• pp.335-336
• /
• 2006

Compound semiconductor manufacturing environment also has been emerged as mass customization and open foundry service so integrated manufacturing system is needed. In this study, we design data acquisition system of compound semiconductor fabrication that has monitoring and control of process. The developed DAS is consisted of key-in system inputted by operator and automatic acquisition system by GEM protocol. And we implemented them in the actual compound semiconductor. It is expected that using developed system would offer precise process information to buyer, reduce a lead-time, and obey a due-dates and so on.

• Asian Journal of Innovation and Policy
• /
• 제2권1호
• /
• pp.20-36
• /
• 2013

In this practice article, we present the results of a scenario planning approach that is a hybrid of the three main schools of thought. Our research objective was to study the future of Interactive Digital Media applications such as online music, on-demand television and massively multi-player online role-playing games. Our approach, while essentially qualitative in nature, nevertheless draws from the rigors of the quantitative school in identifying and then tracking the significant dimensions of analysis that emerge over time as strands of events leading to plausible scenarios. Our empirical analysis revealed mapping strands to three themes - ownership, distribution and innovation - which we used in an expert validation exercise to formulate scenarios. We present and discuss the major findings and implications of this empirical investigation. In a nutshell, we conjecture that an open, competitive IDM marketplace with performance safeguards may serve both and lead to a win-win scenario. While there are differences among IDM sectors, a unified approach to regulation and policy would be effective.

• 한국IT서비스학회지
• /
• 제12권2호
• /
• pp.85-101
• /
• 2013

The wide spread of smartphones and the growth of the number of internet users are reshaping the SNS (Social Network Service). Merging into other service areas and creating new business models with high profits, SNS is no more a service simply providing personal connections. Now SNS has positioned itself as a service platform. SNG (Social Network Game) is a new outcome utilizing SNS as a game platform and showing a rapid growth rate in a short period of time. The number of SNG users is expected to increase steadily. In this study, we examine whether SNG user motivations lead to flow experience and future intention to use. For that purpose, we conducted a survey with smartphone users. The results indicate that flow experience functions as a mediator and user motivations indirectly affect intention to use through flow experience. This paper concludes with discussions on findings and suggestions for future research.

• 산업경영시스템학회지
• /
• 제31권2호
• /
• pp.80-93
• /
• 2008

This paper improves algorithms for an assembly-type flowshop scheduling problem in which each job is to assemble two types of components and makespan is the objective measure. For the assembly, one type of the components is outsourced with job-dependent lead time but the other type is fabricated in-house. When both components for a job are prepared, the assembly operation for the job can be started. This problem had been proved to be NP-Complete, so branch-and-bound (B&B) and heuristic algorithms have already been developed. In this paper, we suggest other dominance rules, lowerbound and heutistic algorithms. Also, we develop a new B&B algorithm using these improved bound and dominance rules. The suggested heuristics and B&B algorithm are compared with existing algorithms on randomly-generated test problems.

• 산업경영시스템학회지
• /
• 제22권51호
• /
• pp.189-197
• /
• 1999

Assembly line balancing problem is known as one of difficult combinatorial optimization problems. This problem has been solved with linear programming, dynamic programming approaches. but unfortunately these approaches do not lead to efficient algorithms. Recently, genetic algorithm has been recognized as an efficient procedure for solving hard combinatorial optimization problems, but has a defect that requires long-run time and computational complexties to find the solution. For this reason, we adapt a new method called the Critical Node Method that is intuitive, easy to understand, simple for implementation. Fuzzy set theory is frequently used to represent uncertainty of information. In this paper, to treat the data of real world problems we use a fuzzy number to represent the duration and Critical Node Method based heuristic procedure is developed for solving fuzzy assembly line balancing problem.

• 전자공학회논문지B
• /
• 제29B권1호
• /
• pp.94-101
• /
• 1992

The multilayer perceptron, trained by the error back-propagation learning rule, has been known as a mapping network which can represent arbitrary functions. However depending on the complexity of a function and the initial weights of the multilayer perceptron, the error back-propagation learning may fall into a local minimum or a flat area which may require a long learning time or lead to unsuccessful learning. To solve such difficulties in training the multilayer perceptron by standard error back-propagation learning rule, the paper proposes a learning method which progressively enlarges the learning domain from a small area to the entire region. The proposed method is devised from the investigation on the roles of hidden nodes and connection weights in the multilayer perceptron which approximates a function of one variable. The validity of the proposed method was illustrated through simulations for a function of one variable and a function of two variable with many extremal points.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2001년도 추계학술대회 논문집 전기기기 및 에너지변환시스템부문
• /
• pp.321-324
• /
• 2001

The accidents in a railway level crossing where passengers or vehicles pass on the track went through by the train with a inherent high rate on incidents lead to a vast economic damage on people and material resources. So, many protection equipments are placed on the railway crossing to prevent those kind of accidents. But there is a limit to control driver's rough activity, passenger's careless or an intentional accident. This study is trying to ensure the safety in railway crossing by providing the information on train's speed and expected passing time on the railroad to a road traffic control system or car driver. Additionally, this system has been studying the railway level crossing system concerning ITS to support harmonious traffic flow.