• Journal of Internet Computing and Services
• /
• v.25 no.4
• /
• pp.7-21
• /
• 2024

There are various machine learning techniques such as Reinforcement Learning, Deep Learning, Neural Network Learning, and so on. In recent, Large Language Models (LLMs) are popularly used for Generative AI based on Reinforcement Learning. It makes decisions with the most optimal rewards through the fine tuning process in a particular situation. Unfortunately, LLMs can not provide any explanation for how they reach the goal because the training is based on learning of black-box AI. Reinforcement Learning as black-box AI is based on graph-evolving structure for deriving enhanced solution through adjustment by human feedback or reinforced data. In this research, for mutually exclusive decision-making, Mutually Exclusive Learning (MEL) is proposed to provide explanations of the chosen goals that are achieved by a decision on both ends with specified conditions. In MEL, decision-making process is based on the tree-based structure that can provide processes of pruning branches that are used as explanations of how to achieve the goals. The goal can be reached by trade-off among mutually exclusive alternatives according to the specific contextual conditions. Therefore, the tree-based structure is adopted to provide feasible solutions with the explanations based on the pruning branches. The sequence of pruning processes can be used to provide the explanations of the inferences and ways to reach the goals, as Explainable AI (XAI). The learning process is based on the pruning branches according to the multi-dimensional contextual conditions. To deep-dive the search, they are composed of time window to determine the temporal perspective, depth of phases for lookahead and decision criteria to prune branches. The goal depends on the policy of the pruning branches, which can be dynamically changed by configured situation with the specific multi-dimensional contextual conditions at a particular moment. The explanation is represented by the chosen episode among the decision alternatives according to configured situations. In this research, MEL adopts the tree-based learning model to provide explanation for the goal derived with specific conditions. Therefore, as an example of mutually exclusive problems, employment process is proposed to demonstrate the decision-making process of how to reach the goal and explanation by the pruning branches. Finally, further study is discussed to verify the effectiveness of MEL with experiments.

• Korea-Australia Rheology Journal
• /
• v.20 no.2
• /
• pp.79-88
• /
• 2008

In this study, we focus on the improvement of the filling efficiency in injection molding by application of ultrasonic vibration. While studies about the filling efficiency of typical filling processes in the injection molding have been widely performed, there have been only few studies about the filling efficiency of an ultrasonic process. The effect of the ultrasonic vibration is an important process condition, which influences the flow characteristics of polymer melt. This new condition even affects well-known injection conditions such as cavity pressure, injection temperature and mold temperature. For this study, we carried out a numerical analysis by appropriate modeling and analysis of the ultrasonic process in the filling process. To verify this numerical analysis, we compared the numerical results with the experimental data. Also, we analyzed the filling process in a thin cavity using this numerical analysis. To understand the flow characteristics of polymer melt in the ultrasonic process, we substituted real and complex vibration conditions with simplified and classified conditions according to the position of vibrating cavity surfaces and the phase difference between two opposing cavity surfaces. We also introduced MFR (melt flow ratio) as a new index to estimate the filling efficiency in the ultrasonic process.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.19 no.9
• /
• pp.53-58
• /
• 2020

This study analyzed the surface roughness of the fine particle spraying process in the plane and the surface roughness by the factors in the fine particle spraying process on the helical surface is analyzed. Finally, the surface fine particle spraying process and the helical curved surface fine particle Analyze the difference in injection conditions of the injection process. Key process variables are particle type, nozzle diameter, and pressure. The remaining conditions are fixed values of. A total of 32 experiments were conducted, each with different process variables. Rectangular and cylindrical specimens were fabricated and a corresponding jig was prepared for use in the experiment. Analyses conducted by using ANOVA enabled comparisons of the effects of each process variable on the experiment.

• Transactions of Materials Processing
• /
• v.12 no.2
• /
• pp.102-109
• /
• 2003

In the manufacturing process of metal matrix composites parts, thixoforging is one of the most effective forming processes. The major purpose of the current study is to provide the proper conditions such as the die shape, the forging velocity, the forging time, the forging pressure and reinforcement injection velocity and pressure on various defects in thixoforged cylinder liner, filling tests were performed by MAGMA S/W. In order to evaluate the effectiveness of the calculated conditions which is given by computer aided engineering, A357, A380 and SiC$_{p}$/A380 cylind~5$mu extrm{m}$r liner were fabricated under the calculated conditions. SiC$_{p}$/A380 composite billets were fabricated by both the mechanical stirring and electrical magnetic stirring process. Incase fo SiC$_{p}$/A380 composite cylinder liner, reinforcement distribution and effect of reinforcement(SiC$_{p}$) content(10~20 vol. %)and size(5.5~14${\mu}{\textrm}{m}$) on the mechanical properties were investigatedstigated.

• Journal of Computational Design and Engineering
• /
• v.3 no.1
• /
• pp.71-90
• /
• 2016

Steel nitriding is a thermo-chemical process leading to surface hardening and improvement in fatigue properties. The process is strongly influenced by many different variables such as steel composition, nitrogen potential, temperature, time, and quenching media. In the present study, the influence of such parameters affecting physic-chemical and mechanical properties of nitride steels was evaluated. The aim was to streamline the process by numerical-experimental analysis allowing defining the optimal conditions for the success of the process. Input parameters-output results correlations were calculated through the employment of a multi-objective optimization software, modeFRONTIER (Esteco). The mechanical and microstructural results belonging to the nitriding process, performed with different processing conditions for various steels, are presented. The data were employed to obtain the analytical equations describing nitriding behavior as a function of nitriding parameters and steel composition. The obtained model was validated, through control designs, and optimized by taking into account physical and processing conditions.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.9 no.4
• /
• pp.8-13
• /
• 2010

The constant velocity internal grinding is a popular process for studying axle machine design and process in automobile industry. In this study, The program which gives the data of wheel size and truing diameter of ball groove is developed. As a result I obtained the data of grinding conditions makes good surface roughness get a grinding conditions. The grinding characteristics and conditions of constant velocity joint were investigated with respect to grinding feed, cutting depth, grinding time. At machine failure, the results were suddenly increased and the detailed surfaces were extremely obtained. Grinding condition was big more affected by grinding depth, grinding speed and grinding time.

• Journal of the Korean Society of Visualization
• /
• v.13 no.2
• /
• pp.16-21
• /
• 2015

Concentration polarization(CP) phenomena show great potential on desalination technology. As we can control the dimension of silicon based nanochannel system, it can be used to model ion exchange membrane. In this study, to investigate the effect of nanochannel number and working conditions on the CP based desalination process, nanochannel based microfluidic system is fabricated. First, we optimized nanochannel number and working conditions to conduct visualization study on CP based desalination process. Second, we visualized the desalination process with fluorescent dye in the desalination chip. We also visualized that the particles also can be removed by the CP based desalination process.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2001.10a
• /
• pp.634-639
• /
• 2001

In this paper, we discuss optimum process conditions of Hemispherical Grained Silicon formation for high density DRAM'S capacitor. In optimum process renditions, the phosphorous concentration, storage polysilicon deposition temperature and thickness of hemispherical grain silicon are in the range of 3.0-4.0E19atoms/㎤, 53$0^{\circ}C$ and 40(equation omitted), respectively. in the 64M bit DRAM capacitor using optimum process conditions, limit thickness of nitride is about 65(equation omitted). The results obtained in this study are applicable to process control and HSG-Si formation for high reliability and high density DRAM's capacitor.

• Transactions of Materials Processing
• /
• v.21 no.1
• /
• pp.42-48
• /
• 2012

Hydroforming has attracted the attention of manufacturing industries for vehicles and transportation systems. A wide range of products such as subframes, camshafts, radiator frames, axles and crankshafts are made by the hydroforming process. Hydroformed parts often need to be structurally joined to other components during assembly. Therefore it is useful if the hydroformed automotive parts can be attached with a localized flange. In this study, a hydroforming process to produce a rectangular shape flange is proposed. FE analysis to form the flanged rectangular shape was performed by Dynaform 5.5. The hydroforming characteristics at various die aspect ratios and feeding conditions were analyzed and optimal process conditions which can avoid defects are suggested. For validation purposes, hydroforming experiments to form the flange were conducted with the optimized conditions. The results show that the flanged parts can be successfully formed with a hydroforming process without additional processing steps.

• Journal of the Korean Society for Precision Engineering
• /
• v.13 no.11
• /
• pp.82-91
• /
• 1996

In this study, we put our object to carry out adaptive modeling of cutting process in turning system, and to find out the optimal cutting conditions to maximize material removal rate under some constraints. We used a back-propagation neural network to model the cutting process adaptively and a genetic algorithm to find out optimal cutting conditions. The experimental results show that a back-propagation neural network could model the cutting process effciently, and optimized cutting conditions for maximizing the material removal rate were obtained through the adaptive process model and genetic algorithms. Therefore, the proposed approach can be applied to the real machining system.