• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2003.10a
• /
• pp.271-276
• /
• 2003

In this study, we compared backpropagation neural network(BPNN) with probabilistic neural network(PNN) as shape recognition algorithm of welding flaws. For this purpose, variables are applied the same to two algorithm. Here, feature variable is composed of time domain signal itself and frequency domain signal itself, Through this process, we comfirmed advantages/disadvantages of two algorithms and identified application methods of two algorithms.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 1999.10a
• /
• pp.432-439
• /
• 1999

In this paper, an off-line programming(OLP) system is presented as the three dimensional graphic simulator and one of the human-robot interface systems for industrial robots. The OLP system has been especially developed to testify robot programs visually using three dimensional geometric modeling and graphics technologies in personal computes. A special feature is its capability of collision detection and of comparing performance of control algorithms. This paper places the focus on the structure and major characteristic of OLP system.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2004.10a
• /
• pp.280-285
• /
• 2004

In this study, we made a comparative study of backpropagation neural network and probabilistic neural network and bayesian classifier and perceptron as shape recognition algorithm of welding flaws. For this purpose, variables are applied the same to four algorithms. Here, feature variable is composed of time domain signal itself and frequency domain signal itself. Through this process, we comfirmed advantages/disadvantages of four algorithms and identified application methods of four algorithms.

• Proceedings of the KIEE Conference
• /
• 2003.07d
• /
• pp.2549-2551
• /
• 2003

We propose character recognition system for Gerber files. The Gerber file is the vector-formatted drawing file for PCB manufacturing. To consider the special vector format and rotated characters, we develop segmentation and feature extraction method. The modularized neural network is then applied to the recognition algorithm. Finally, comparative simulation results are presented to verify the usefulness of the proposed method.

• Journal of the Korean Society for Precision Engineering
• /
• v.30 no.12
• /
• pp.1303-1312
• /
• 2013

We report an ultra-precision lathe designed to machine micron-scale features on a large-area roll mold. The lathe can machine rolls up to 600 mm in diameter and 2,500 mm in length. All axes use hydrostatic oil bearings to exploit the high-precision, stiffness, and damping characteristics. The headstock spindle and rotary tooling table are driven by frameless direct drive motors, while coreless linear motors are used for the two linear axes. Finite element method modeling reveals that the effects of structural deformation on the machining accuracy are less than $1{\mu}m$. The results of thermal testing show that the maximum temperature rise at the spindle outer surface is approximately $0.5^{\circ}C$. Finally, performance evaluations of the error motion, micro-positioning capability, and fine-pitch machining demonstrate that the lathe is capable of producing optical-quality surfaces with micron-scale patterns with feature sizes as small as $20{\mu}m$ on a large-area roll mold.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.12
• /
• pp.3330-3344
• /
• 2023

This paper proposes an antenna performance prediction model in the autonomous driving radar manufacturing process. Our research work is based upon a challenge dataset, Driving Radar Manufacturing Process Dataset, and a typical AutoML machine learning workflow engine, Pycaret open-source Python library. Note that the dataset contains the total 70 data-items, out of which 54 used as input features and 16 used as output features, and the dataset is properly built into resolving the multi-output regression problem. During the data regression analysis and preprocessing phase, we identified several input features having similar correlations and so detached some of those input features, which may become a serious cause of the multicollinearity problem that affect the overall model performance. In the training phase, we train each of output-feature regression models by using the AutoML approach. Next, we selected the top 5 models showing the higher performances in the AutoML result reports and applied the ensemble method so as for the selected models' performances to be improved. In performing the experimental performance evaluation of the regression prediction model, we particularly used two metrics, MAE and RMSE, and the results of which were 0.6928 and 1.2065, respectively. Additionally, we carried out a series of experiments to verify the proposed model's performance by comparing with other existing models' performances. In conclusion, we enhance accuracy for safer autonomous vehicles, reduces manufacturing costs through AutoML-Pycaret and machine learning ensembled model, and prevents the production of faulty radar systems, conserving resources. Ultimately, the proposed model holds significant promise not only for antenna performance but also for improving manufacturing quality and advancing radar systems in autonomous vehicles.

• Journal of Korea Multimedia Society
• /
• v.13 no.8
• /
• pp.1115-1127
• /
• 2010

In order to increase the productivity in manufacturing TFT-LCD(thin film transistor-liquid crystal display), it is essential to classify defects that occur during the production and make an appropriate decision on whether the product with defects is scrapped or not. The decision mainly depends on classifying the defects accurately. In this paper, we present an effective classification method for film defects acquired in the panel production line by analyzing the intensity distribution and shape feature of the defects. We first generate a binary image for each defect by separating defect regions from background (non-defect) regions. Then, we extract various features from the defect regions such as the linearity of the defect, the intensity distribution, and the shape characteristics considering intensity, and construct a referential image database that stores those feature values. Finally, we determine the type of a defect by matching a defect image with a referential image in the database through the matching cost function between the two images. To verify the effectiveness of our method, we conducted a classification experiment using defect images acquired from real TFT-LCD production lines. Experimental results show that our method has achieved highly effective classification enough to be used in the production line.

• Journal of Korea Spatial Information System Society
• /
• v.5 no.2 s.10
• /
• pp.23-31
• /
• 2003

As the results of the information project, manufacturing of the digital map and various thematic maps of the national land has been completed. Furthermore, it is necessary to organize a systemized management system, which can control and utilize all artificial features (e.g. buildings, roads, bridges etc.) as well as natural geographic features (e.g. rivers etc). Howener, it has difficulties in managing NGIS as a unified system, because of the discordance of DB of each geographic management organizations. Therefore, this study has been conducted to apply to the system and method providing geographic UFID that can be a key in order to managing, searching and utilizing geographic and artificial features and that makes it be able to estimate the location with the only identifier. Moreover, the system and method, providing geographic UFID, applies to systemized management of NGIS DB as well as consistency of information.

• KIEE International Transaction on Electrical Machinery and Energy Conversion Systems
• /
• v.11B no.3
• /
• pp.112-118
• /
• 2001

In the near future, the voltage of power system for passenger vehicle will be changed to 42V from existing 14V./ Because of increasing power and voltage ratings used in the vehicle the motor drive system has high switching dv/dt and it generates electromagnetic interference (EMI) To solve these problems multi-level inverter system may be used The feature of multi-level inverter is the output voltage to be synthesized from several levels of voltage Because of this feature high switching dv/dt and EMI can be reduced in the multi-level inverter system But as the number of level is increased manufacturing cost is getting expensive and system size is getting large. Because of these disadvantages the application of multi-level inverter has been restricted only to high power drives. The method to reduce manufacturing cost and system size is to integrate circuit of multi-level inverter into a few chips But isolated power supply and signal isolation circuit using transformer or opto-coupler for drive circuit are obstacles to implement the integrated circuit (IC) In this paper a drive circuit of 4-level inverter suitable for integration to hybrid or one chip is proposed In the proposed drive circuit DC link voltage is used directly as the power source of each gate drive circuit NPN transistors and PNP transistors are used to isolate to transfer the control signals. So the proposed drive circuit needs no transformers and opto-couplers for electrical isolation of drive circuit and is constructed only using components to be implemented on a silicon wafer With th e proposed drive circuit 4- level inverter system will be possible to be implemented through integrated circuit technology Using the proposed drive circuit 4- level inverter system is constructed and the validity and characteristics of the proposed drive circuit are proved through the experiments.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.17 no.2
• /
• pp.83-88
• /
• 2018

Additive manufacturing (AM) technologies have attracted wide attention as key technologies for the next industrial revolution. Among AM technologies using various materials, powder bed fusion (PBF) processes and direct energy deposition (DED) are representative of the metal 3-D printing process. Both of these processes have a common feature that the laser is used as a heat source to fabricate the 3-D shape through melting of the metal powder and solidification. However, the material properties of the deposited metals differ when produced by different process conditions and methods. 17-4 precipitation-hardening stainless steel (17-4PH SS) is widely used in the field of aircraft, chemical, and nuclear industries because of its good mechanical properties and excellent corrosion resistance. In this study, we investigated the differences in microstructure and mechanical properties of deposited 17-4PH SS by PBF and DED processes, including the heat treatment effect.