• 동력기계공학회지
• /
• 제6권2호
• /
• pp.60-67
• /
• 2002

연삭가공은 나노스케일(Nano-scale)의 미소한 입자 절삭날을 이용한 가공으로, 공작물의 표면을 경면(Mirror surface)으로 가공할 수 있어 제품의 최종 마무리공정으로 사용되어 왔다. 그러나 연삭공정에 있어서는 공구(연삭숫돌)의 수명이 다하거나 가공계(Machining system)가 불안정해지면 채터진동과 연삭버닝 등의 현상이 발생하여 가공물의 표면품위를 저하시키는 요인으로 작용하고 있다. 따라서 본 연구는 원통플른지 연삭공정을 대상으로 공작물에서 발생하는 음향방출 신호와 연삭기 주축 모터의 동력 신호를 연삭가공 중에 검출하고, 이를 신경회로망에 적용하여 연삭가공 상태를 진단하는 시스템을 구축하고, 그 성능을 평가하였다.

• 동력기계공학회지
• /
• 제4권2호
• /
• pp.25-30
• /
• 2000

The grinding process is very complex and relates many parameters to control the process. As this reason, a theoretical analysis and a quantitative estimation of the grinding process has not been well established. In this study, the in-process monitoring system was suggested by applying the neural network for monitoring and shooting the malfunction of cylindrical plunge grinding process. This system used the power signals from the electric power meter. This neural network was composed of processing elements [4-(5-5)-3] with 4 identified power parameters. Because sensitivity is blunted some minute vibration components, the simulation result of this system has appeared about 10% erroneous recognition in the uncertain pattern and the average success rate of the trouble recognition was about 90%. Consequently, the developed system, which applied to the power signals, can be recognize enough to monitor the grinding process as in-process.

• 한국정밀공학회지
• /
• 제13권9호
• /
• pp.149-155
• /
• 1996

Recent researches in the trouble monitoring system of grinding process have emphasized the use of deep knowledge. Such works include the monitoring and diagnostic systems for cylindrical grinding using sensors on chatter vibration and grinding burn during the process. But, since grinding operations are especially related with a lalrge amount of ambique parameters, it is effectively difficult to detect the grinding troubles occuring during the grinding process. In this paper, monitoring system for grinding utilizes the neural networks based on grinding power signatures. The monitoring system of grinding operations, which makes use of PDP neural networks, is presented. Then, the implementation results by computer simulations and experimental data with respect to chatter vibration and grinding burn are compared.

• 한국기계가공학회지
• /
• 제1권1호
• /
• pp.29-37
• /
• 2002

The wavelet transform is a popular tool for studying intermittent and localized phenomena in signals. In this study the wavelet transform of cutting force signals was conducted for the diagnosis of grinding conditions in grinding process. We used the Daubechies wavelet analyzing function to detect a sudden change in cutting signal level. STD11 workpiece was 85 times of machined pieces cut by the WA wheel and a tool dynamometer obtained cutting force signals. From the results of the wavelet transform, the obtained signals were divided into approximation terms and detailed terms. At dressing time, the approximation signals were slowly increased and 45 machined times noticed dressing time.

• 한국공작기계학회:학술대회논문집
• /
• 한국공작기계학회 2000년도 춘계학술대회논문집 - 한국공작기계학회
• /
• pp.20-27
• /
• 2000

A chatter vibration and a workpiece burn are the main phenomena to be monitored in modern grinding processes. This study describes a trouble diagnosis of the cylindrical plunge grinding process using the power and acoustic emission (AE) signals. The raw signals of the power and the AE occurred during the grinding operation were sampled and analyzed to determine the relationship between each fault and change of signals. A neural network that has a high success rate of the fault detection was used. Furthermore, an analysis on the influence of parameters to the chatter vibration and the grinding burn was conducted.

• 한국생산제조학회지
• /
• 제9권4호
• /
• pp.108-116
• /
• 2000

Undesired trouble such as chatter vibration and burning on the ground surface appears frequently in the cylindrical plunge grinding process. Establishment of a credible fault diagnostic system for the grinding process is the major purpose of this study. Power signals generated during the grinding operation were sampled and analyzed to determine the relationship between grinding troubles and behavior of signal changes. In addition, a neural network was optimized with a momentum coefficient a learning rate, and a structure of the hidden layer through the iterative learning process. Based on the established system, success rates of the trouble recognition were verified.