• Journal of the Korean Society for Precision Engineering
• /
• v.15 no.7
• /
• pp.44-51
• /
• 1998

Two-step neural network is designed for determining electrical discharge machining parameters in low erosion. The first neural network, which is used as a classification network, checks whether the current conditions are appropriate to electrical discharge machining in low tool erosion. If the conditions are appropriate to EDM in low erosion, suitable EDM parameters are generated by the second neural network. Theoretically known EDM conditions are produced and also utilized for training the second neural network. The trained neural network is tested how well suitable EDM machining conditions are generated under unknown machining situations Experimental result shows that the proposed two-step neural network approach could be effectively used for determining EDM parameters in low tool erosion. The results also have a practical contribution to EDM area in that it could be applied for maintaining low tool wear as well as obtaining maximum machining rates simultaneously.

• Electronics and Telecommunications Trends
• /
• v.36 no.4
• /
• pp.34-47
• /
• 2021

To realize remote surgery from hundreds of kilometers away, a new communication environment with ultra-low latency and high-precision features is required. Thus, ultra-high precision networking technology that guarantees the maximum latency and jitter of end-to-end traffic on an Internet-scale wide area network is in development as part of 6G network research. This paper describes the current status of various networking technologies in ITU-T, ETSI, IEEE, and IETF to ensure ultra-low latency and high precision in wired networks.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2002.05a
• /
• pp.424-427
• /
• 2002

Pneumatic control systems have the potential to provide high output power to weight and size ratios at a relatively low cost. However, they are mainly employed in open-loop control applications where positioning repeatability is not of great importance. This paper presents precision positioning control of pneumatic servo cylinder with on-off valve, Pneumatic low-friction cylinder with servo valve and DC servo motor under parameter variations. Basically positioning control uses PID controller, where needs a linearized model. A neural network is added to a PID controller to compensator nonlinearity of the system and an influence of friction force is consider as disturbance. The performances of the proposed algorithms were compared by experiments with them of PID controller. From those experiments is was shown that the proposed algorithms are more efficient about settling time, steady 7tate error and overshoot than PID control algorithm.

• Journal of Positioning, Navigation, and Timing
• /
• v.10 no.1
• /
• pp.67-73
• /
• 2021

The need for precise location data is growing across numerous markets, and so is the number of affordable high-precision GPS receivers. In this paper, we validated the performance of RTAP2U, a low-cost high-precision RTK receiver that was recently released. Two positioning modes were tested: static and driving. The static test conducted Zero-Baseline Single-RTK and Network-RTK survey for 57 hours and 51 hours, respectively. For the driving test, Network-RTK survey was conducted using VRS services provided by NGII based on Trimble PIVOT and Geo++ GNSMART. The static test showed about 1 cm horizontal and vertical accuracies, which is very stable considering the test duration longer than 50 hours. The integer ambiguity FIX rate marked a solid 100%. The driving test result also reached a 100% FIX rate. Horizontal and vertical accuracies were better than 2 cm and 3 cm, respectively. Researchers can refer to this paper when considering affordable high-precision GPS receivers as an option.

• Journal of the Korean Society for Precision Engineering
• /
• v.13 no.10
• /
• pp.62-70
• /
• 1996

We experimented on cutting characteristics-cutting force, behavior of cutting temperature, surface roughness. chip thickness under low temperature, which generated by liquid nitrogen(77K). The work-pieces were freezed to-195 .deg. C and liquid nitrogen was also sprinkled on cutting area in order to decrease an experimental error of machining in low temperature. The workpiece was became to -195 .deg. C in5 minutes. In cooled condition surface roughness of workpiece was better than normal condition. In addition, we investigated the possibility that surface roughness of workpiece and cutting force can be predicted analyzing cutting conditions by the trained neural network.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2010.05a
• /
• pp.801-802
• /
• 2010

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1994.10a
• /
• pp.32-37
• /
• 1994

We experimented on cutting characteristics - cutting force, behavior of cutting temperature, surface roughness, behavior of chips-under low temperature,which generated by liquid nitrogen(77K). The workpieces were freezed to -195 .deg. C and liquid nitrogen was also sprinkled on cutting area in order to increase the efficiency of machining in low temperature. The workpiece was became to -195 .deg. C in 5 minutes. In cooled condition(CC) surface roughness of workpiece was better than normal condition(NC). In addition, we investigated the possibility that surface roughness of workpiece and shear angle can be predicted analyzing cutting condititions by the trained neural network.

• International Journal of Precision Engineering and Manufacturing
• /
• v.8 no.4
• /
• pp.27-32
• /
• 2007

Gears, crucial components in modern precision machinery for power transmission mechanisms, are required to have low contacting noise with high torque transmission, which makes the use of gear-tooth profile modifications and gear-tooth surface crowning extremely efficient and valuable. Due to the shortcomings of current techniques, such as manual rectification, mechanical modification, and numerically controlled rectification, we propose a novel electrochemical gear-tooth profile modification method based on an artificial neural network control technique. The fundamentals of electrochemical tooth-profile modifications based on real-time control and a mathematical model of the process are discussed in detail. Due to the complex and uncertain relationships among the machining parameters of electrochemical tooth-profile modification processes, we used an artificial neural network to determine the required processing electric current as the tooth-profile modification requirements were supplied. The system was implemented and a practical example was used to demonstrate that this technology is feasible and has potential applications in the production of precision machinery.

• Journal of the Korean Society for Precision Engineering
• /
• v.22 no.4
• /
• pp.60-67
• /
• 2005

Compared with conventional welding, laser spot welding offers a unique combination of high speed, precision and low heat distortion. This combination of advantages is attractive for manufacturing industries including automotive and electronics companies. In this paper, a real time monitoring scheme fur a pulsed Nd:YAG laser spot welding was suggested. Acoustic emission (AE) signals were collected during welding and analyzed for given process conditions such as laser power and pulse duration. A back propagation artificial neural network, with AE frequency content inputs, was used to predict the weldability of stainless steel sheets.

• Journal of Broadcast Engineering
• /
• v.26 no.1
• /
• pp.14-25
• /
• 2021

Change detection, which finds changes in remote sensing images of the same location captured at different times, is very important because it is used in various applications. However, registration errors, building displacement errors, and shadow errors cause false positives. To solve these problems, we propose a novle deep convolutional network called CADNet (Change Attention Dense Siamese Network). CADNet uses FPN (Feature Pyramid Network) to detect multi-scale changes, applies a Change Attention Module that attends to the changes, and uses DenseNet as a feature extractor to use feature maps that contain both low-level and high-level features for change detection. CADNet performance measured from the Precision, Recall, F1 side is 98.44%, 98.47%, 98.46% for WHU datasets and 90.72%, 91.89%, 91.30% for LEVIR-CD datasets. The results of this experiment show that CADNet can offer better performance than any other traditional change detection method.