• 한국전자통신학회논문지
• /
• 제15권1호
• /
• pp.101-106
• /
• 2020

본 연구는 다양한 어셈블리 부품의 용접부 내부결함을 검사하기 위한 초음파 탐상 장비 개발에 관한 것이다. 본 연구에서는 초음파 탐상을 위하여 시스템의 모션제어 S/W, 초음파 송수신기 제어, 결함 판정 기준 설정 등의 계측 S/W 등이 설계되었으며, 양품과 불량품의 비교분석을 하기 위하여 용접결함 불량품 샘플워크 등도 제작되었다. 이와 같은 구성으로 이루어진 시스템을 통하여 어셈블리 부품 용접부의 결함 위치 및 깊이에 대한 자동검사 기능을 수행할 수 있었으며, 종전에 전문가에 의해 이루어졌던 용접부의 내부결함에 대한 판단을 시스템이 수행하도록 하였다.

• 한국정밀공학회지
• /
• 제19권8호
• /
• pp.116-125
• /
• 2002

Semi-Solid Diecasters usually carry out the Semi-Solid diecasting experiments before producing new casts. At the Semi-Solid diecasting stages, the runner-gate part has been always repeatedly corrected, which leads to a tedious processing time and increased processing cost. A large amount of experience is essential in manual assessment and if the design is defective, much time and a great deal of efforts will be wasted in the modification of the die. In this study, design system has been developed based on design database. In addition, gate experiment for gating system design has been carried out to append the database. It is possible for engineers to make efficient gating system design of Semi-Solid diecasting and it will result in the reduction of expenses and time to be required. The detailed contents of the research are described in the followings.

• 한국컴퓨터산업학회논문지
• /
• 제3권9호
• /
• pp.1177-1188
• /
• 2002

본 논문에서는 생산현장에서 생산하는 Lead Pin을 생산근로자의 시각으로 품질을 검사하는 방법을 개선하기 위하여, 영상처리 알고리즘을 사용하여 자동으로 품질의 정ㆍ오판별을 검사하기 위한 실시간 영상처리방법을 제안한다. 먼저 영상정보의 실시간 취득을 위하여 C.C.D와 영상취득기(Image frame grabber : DT3153)를 사용하여 초당 30프레임(30 Frame/second)으로 영상을 취득할 수 있는 실시간 영상취득시스템을 구성하였으며, 이를 사용하여 Lead Pin의 영상을 취득하여 Lead pin의 형상을 나타내는 형상프로파일의 영상처리 알고리즘을 사용하여 연구를 수행하였다. Lead Pin의 정품과 비정품을 평가하기 위해 숙련된 작업자에 의해 판별된 정품 100개, 비정품 100개, 전체 200개의 인식대상 물체를 판별한 수행결과 정품을 정품으로 인식하여 판별한 경우는 97%, 비정품을 비정품으로 판별한 경우는 95%로 전체 인식률은 96%의 인식결과를 나타내었으며, 전체 오분류률은 4%를 나타냄을 알 수 있었다.

• 소성∙가공
• /
• 제27권5호
• /
• pp.289-295
• /
• 2018

The profile ring rolling process can realize various ring shapes unlike conventional rectangular cross-sectional ring products. In this paper, the defective groove in the bottom surface of L-shaped ring products was analyzed. Grooves are generated by non-uniform external forces due to profile main roll and initial blank shape. Process parameters such as the motion of dies and working temperature were determined. Mechanism of groove formation was analyzed by FE simulation on the basis of local external forces acting on the blank. Analysis results were similar to the groove actually occurring in the production line. Based on results of the analysis, two solutions were proposed for the groove. The position of the base plate supporting the blank was adjusted and edge length of the main roll was extended to suppress growth of grooves. It has been verified that groove was improved by applying two proposed methods in the shop-floor.

• 대한기계학회논문집A
• /
• 제31권8호
• /
• pp.833-838
• /
• 2007

In tube hydroforming process, several defective products could be obtained such as bursting, wrinkling, folding, buckling. Because, especially, bursting is most frequently occurred failure among the well known failures, it is mostly important to predict the onset of bursting failure on tube hydroforming process. For most sheet metal forming processes, strain based forming limit diagram(FLD) is used often as a criteria to estimate the possibility of onset of the failures proposed above. However, FLD has a shortcoming that it is dependent on strain path while stress based diagram is independent on strain history. Generally, tube hydroforming consists of three main processes such as pre-bending, pre-forming, and hydroforming and it means that the strain histories of final products are nonlinear. Therefore, forming limit stress diagram(FLSD) is more suitable to predict forming limit for hydroforming parts. In this study, FLSD is applied to estimate bursting failure for an engine cradle of an automobile part. Consequently, it is proved that application of FLSD to predict forming limit is available for tube hydroforming parts.

• Design & Manufacturing
• /
• 제12권1호
• /
• pp.18-25
• /
• 2018

Both injection and injection molding dies have evolved into advanced technology. Product quality is also evolving day after day. Therefore, the conditions of the injection mold and the injection conditions are becoming important. In order to improve the quality of the product, the Hardware part of the mold has developed as an advanced technology, and the Software part has also developed with advanced technology. This study deals with the cooling part, which is part of the hardware. In addition to fluid cooling, which is commonly used in the industry, by using gas cooling identify the phenomena that appear on the surface of the product and the critical point strain of the product to find the optimal cooling. Electronic parts and automobile parts whose surface condition is important, the cooling process is important to such a degree that they are divided with good products and defective products according to the cooling process at the time of injection. By controlling this important cooling and reducing the injection time with additional cooling, the product quality can be increased to the highest production efficiency. In addition, high efficiency can be achieved without additional investment costs. This study was conducted to apply these various advantages in the field.

• 한국자동차공학회논문집
• /
• 제24권1호
• /
• pp.90-98
• /
• 2016

Currently government certified test method for an automobile remanufactured products is insufficient. Thus many automotive parts in the remanufacturing market are lacking proper evaluation criteria and production of defective products are causing customer dissatisfaction. In this paper a power steering oil pump, which requires stringent manufacturing standards, is studied by the failure mode and effect analysis approach. The research suggested that the test criteria such as discharge flow characteristic test, tightness test, pulley run-out test, pressure switch operation test, low temperature test and rotation pressure durability test should be performed to evaluate the reliability of remanufactured power steering oil pumps. As a result of tests, the performance of remanufactured power steering oil pump satisfied the evaluation criteria of pressure switch operation test and low temperature test. However, the remanufactured power steering oil pump failed to satisfy the evaluation criteria on discharge performance test, tightness test and pulley run-out test. These performance evaluation tests proved the necessity of standard process for the remanufactured power steering oil pump.

• Design & Manufacturing
• /
• 제15권3호
• /
• pp.1-6
• /
• 2021

An artificial neural network model based on a deep learning algorithm is known to be more accurate than humans in image classification, but there is still a limit in the sense that there needs to be a lot of training data that can be called big data. Therefore, various techniques are being studied to build an artificial neural network model with high precision, even with small data. The transfer learning technique is assessed as an excellent alternative. As a result, the purpose of this study is to develop an artificial neural network system that can classify burr images of light guide plate products with 99% accuracy using transfer learning technique. Specifically, for the light guide plate product, 150 images of the normal product and the burr were taken at various angles, heights, positions, etc., respectively. Then, after the preprocessing of images such as thresholding and image augmentation, for a total of 3,300 images were generated. 2,970 images were separated for training, while the remaining 330 images were separated for model accuracy testing. For the transfer learning, a base model was developed using the NASNet-Large model that pre-trained 14 million ImageNet data. According to the final model accuracy test, the 99% accuracy in the image classification for training and test images was confirmed. Consequently, based on the results of this study, it is expected to help develop an integrated AI production management system by training not only the burr but also various defective images.

• 산업경영시스템학회지
• /
• 제45권4호
• /
• pp.86-98
• /
• 2022

Recently, many studies have been conducted to improve quality by applying machine learning models to semiconductor manufacturing process data. However, in the semiconductor manufacturing process, the ratio of good products is much higher than that of defective products, so the problem of data imbalance is serious in terms of machine learning. In addition, since the number of features of data used in machine learning is very large, it is very important to perform machine learning by extracting only important features from among them to increase accuracy and utilization. This study proposes an anomaly detection methodology that can learn excellently despite data imbalance and high-dimensional characteristics of semiconductor process data. The anomaly detection methodology applies the LIME algorithm after applying the SMOTE method and the RFECV method. The proposed methodology analyzes the classification result of the anomaly classification model, detects the cause of the anomaly, and derives a semiconductor process requiring action. The proposed methodology confirmed applicability and feasibility through application of cases.

• 인터넷정보학회논문지
• /
• 제25권2호
• /
• pp.21-28
• /
• 2024

레이다는 자율주행 차에 있어 필수적인 센서 부품으로, 레이다가 활용되는 시장은 점차 커지고 있으며 제품 종류도 다양해지고 있다. 본 연구에서는 평가 공정에서부터 레이다의 불량 여부를 예측해 자율주행의 안정성과 효율성을 높일 수 있도록 성능 예측 모델을 구축하고 평가하였다. 레이더 공정 과정의 39607개 입력 데이터로 모델을 학습하였으며, 결과적으로 17개 모델을 스태킹 앙상블했을 때 Meta Ridge 모델이 가장 높은 학습률을 나타내는 것을 확인하였다. 이러한 연구 결과가 제품의 불량을 공정 단계에서 우선 예측해 수율을 극대화하고 불량으로 인한 제품 폐기 비용을 감축하는 데 도움이 될 것으로 기대 한다.