Precision allocation optimization modeling of large-scale CNC hobbing machine based on precision reliability

Zongyan Hu,Shilong Wang,Chi Ma 대한기계학회 2023 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.37 No.2

Reducing manufacturing cost is the main goal of machine tool precision allocation under the condition of meeting precision design requirements. The previous studies generally took "ensuring machining errors completely within the allowable range of design precision" as constraint. Due to the strict constraint, the cost reduction effect is limited. This paper proposes a new method of precision allocation based on precision reliability. The probability that the machining errors are within the design precision allowable range is taken as the measurement index of precision reliability, and the optimization model constraint is relaxed to "that the machining errors are within the allowable range of design precision with predefined precision reliability", so as to obtain lower manufacturing cost under tolerable precision loss risk. The case study of a large-scale gear hobbing machine shows that this method can effectively reduce the manufacturing cost, and the precision allocation is more economical and reasonable.

치과용 3D 프린팅 기술로 제작된 가공물의 정밀성 분석

최성민 대한치과기공학회 2018 대한치과기공학회지 Vol.40 No.4

Purpose: The development of the dental technology industry has digitized the dental process. Definition of Digitalization of Dental Implantation Digitalization is the process of model making and prosthodontic processing using 3D CAD and CAM. Currently, it is becoming popular due to the spread of various devices. However, precision evaluation at this stage is an important factor in precision-based dental procedures. Therefore, in this study, we want to analyze the precision of the processed body made with digital dental machine. Methods: The accuracy of digital dental pore devices was evaluated. Specimens were fabricated with 5 types of 3D printers. The external shape was measured with the prepared specimen. The surface roughness was measured. Results: As a result, precision was excellent in order of EP2 specimen, EP1 specimen, and EA2 specimen. The precision of EP3 specimen and EA1 specimen is not excellent. And the precision of the specimen processed with polymer 3D printer is excellent. The accuracy of LCPS type polymer 3D printers is considered to be excellent. Conclusion : 1. Observation of the shape The 3D printer for LCPS system and the 3D printer for SLM 40μm system were found to be precisely processed. 2. Surface roughness results The LCPS system polymer 3D printer has been shown to have a precise surface.

A Study for Medical Precision Control Machine Using AX-12

Jo, Heung-Kuk The Korea Institute of Information and Commucation 2010 Journal of information and communication convergen Vol.8 No.5

Control devices perform various works for us in many areas. The device is being utilized for precision movement of certain object. In as much as control devices are activated by means of motors, motor control is important.[1][2] Generally, servo motors capable of precision control are more frequently used than DC motors. Use of 3 motors allows 3- way movement. Medical controllers for surgical operation require high precision. [3][4][5][6] AX-12, a servo motor can realize various types of movement. AX-12 can be easily manufactured in the form of a robotic arm and has features that MCU and its peripheral circuits are simple. For precision movement, 3 motors can be controlled by use of a single joystick and 2 buttons, with movement angles being adjusted by having preset values in the program changed.[7][8] By virtue of this study, we have realized small precision robotic arm system utilizing single joystick and 2 buttons. This system can control the robotic arm in the direction desired by the user. The system has been designed such that a joystick controls 2 motors with the remaining motor being controlled by a button. Single MCU is tasked with both control and movement.[9] We have shown precision robotic arm system in the Figure contained in the conclusion part and made reference to results of analysis in there. It has also been demonstrated that the system can be utilized in the industry.[8]

Analysis of the Construction and Effectiveness of Precision- Targeted Classroom Based on Analysis of Students’ Real Learning Situation

Chao Xiong,Xiuyun Yu,Jiaxin Chen 한국수학교육학회 2022 수학교육연구 Vol.25 No.4

In response to the current educational situation of students’ heavy workload, the author constructs the precision-targeted classroom based on Precision Teaching (PT), Network Pharmacology, and Treatment Based on Syndrome Differentiation. The precision-targeted classroom can solve the current problems of PT and the phenomenon of the heavy academic burden on students, achieve the reduction of the burden and increase the efficiency of education. The precision-targeted classroom includes five key points: targeted goals, childlike thinking, precise intervention, intelligent homework, and stereoscopic evaluation, and the implementation process of the precision-targeted classroom is built from three aspects: before, during and after class. In addition, the author applied it to the actual mathematics classroom to test its teaching effect, and the experimental results showed that: the precision-targeted classroom significantly improved students’ academic performance and thinking level; considerably improved students’ classroom learning status, and facilitated teaching personalization and realized homework quantity control and quality improvement.

Mixed Precision 기반 CNN 추론 연산의 구현 및 성능 분석

이종은,장경빈,임승호 한국정보기술학회 2023 한국정보기술학회논문지 Vol.21 No.12

Recently, deep learning technology has increased accuracy due to high-precision models, but resource consumption and processing time required for deep learning operation increase due to the increase in the number of parameters and the depth of complex networks. Such network complexity becomes an overhead for performing deep learning operations in embedded systems such as IoT systems and mobile systems. A lot of research on lightweighting for deep learning models suitable for systems such as quantization has been made. However, since accuracy loss occurs when quantization is applied, mixed precision application that uses mixed parameters is becoming an alternative. In this paper, we implemented an inference operation method of a CNN deep learning model based on Mixed Precision for embedded system, and analyzed the performance of Mixed Precision-based inference of CNN model. As a result of the experiment, it is identified that the inference time can be much reduced with little loss of accuracy with mixed precision when quantization is applied to only few layers with high computational complexity.

딸기 온실에서 점박이응애의 축차표본조사법 개발

최호정,강주완,정효진,최시라,박정준 한국환경생물학회 2017 환경생물 : 환경생물학회지 Vol.35 No.4

두 개의 딸기 온실 (농약살포포장, 천적방사포장)에서 점 박이응애 (Tetranychus urticae Koch)의 고정 정확도 수준에 서 표본조사법 (Fixed-precision sampling plan)을 개발하였 다. 표본추출은 조사구역당 3개엽으로 이루어진 복엽 1줄기 를 기준으로 하여 3줄기를 채취하였다. 각 복엽은 Relative net precision (RNP) 값을 비교하기 위해 3개의 서로 다른 단 위 (1엽, 2엽, 3엽)로 나뉘어졌다. RNP 값 결과 1엽 단위가 다 른 단위들보다 정확도와 효율적인 면에서 우수했다. 공간분 포 분석은 Taylor’s power law (TPL)를 이용하였으며, 각 딸 기 온실별로 계산된 TPL 계수의 동질성검정에는 공분산분 석 (ANCOVA)을 이용하였으며, 분석결과 차이를 보이지 않 았다. 표본추출 정시선을 구하기 위한 TPL 계수는 농약살포 포장과 천적방사포장의 딸기 1엽 단위에서 점박이응애 밀 도 자료를 합한 뒤 재계산하여 사용하였다. 그리고 점박이 응애 발생밀도수준을 3마리와 10마리로 설정하여 방제의사 를 결정하였다. 분석에 사용하지 않은 독립된 자료를 이용 하여 개발된 표본추출법의 유효성을 Resampling Validation for sampling plan (RVSP)으로 확인한 결과 적합한 정확도를 보였다. A fixed-precision-level sampling plan was developed to establish control of the two-spotted spider mite, Tetranychus urticae, in two strawberry greenhouses (conventional plot, natural enemy plot). T. urticae was sampled by taking a three-leaflet leaf (1 stalk) from each plant (3 three-leaflet leaves) from each sampling position. Each leaflet was divided into three different units (1-leaflet, 2-leaflet, and 3-leaflet units) to compare relative net precision (RNP) values for selection of the appropriate sampling unit. The relative net precision values indicated that a 1-leaflet unit was more precise and cost-efficient than other units. The spatial distribution analysis was performed using Taylor’s power law (TPL). Homogeneity of the TPL parameters in each greenhouse was evaluated by using the analysis of covariance (ANCOVA). A fixed-precision-level sequential sampling plan was developed using the parameters of TPL generated from the combined data of the conventional plot and natural enemy plot in a 1-leaflet sampling unit. Sequential classification sampling plans were also developed using the action threshold of 3 and 10 mites for pooled data. Using the results obtained in the independent data, simulated validation of the developed sampling plan by Resampling validation for sampling plan (RVSP) indicated a reasonable level of precision.

골밀도검사의 올바른 질 관리에 따른 임상적용과 해석 - 이중 에너지 방사선 흡수법을 중심으로 : 골밀도검사의 올바른 질 관리에 따른 임상적용과 해석

동경래(Kyung-Rae Dong),김호성(Ho-Sung Kim),정운관(Woon-Kwan Jung) 대한방사선과학회(구 대한방사선기술학회) 2008 방사선기술과학 Vol.31 No.1

목 적: 골밀도검사의 중요한 부분을 차지하고 있는 검사 장비 및 검사자의 정밀도와 정확도는 환경에 따라 차이가 있기 때문에 질 관리가 체계적으로 이루어져야 한다. 골밀도 검사장비의 노화 및 잦은 고장에 의하여 장비의 교체 및 추가 구입으로 인하여, 추적검사를 하는 환자들의 호환성에 문제가 있다. 따라서 장비 교체 및 증설 후 동일한 장비처럼 호환하여 사용해도 환자의 임상적인 골밀도 변화를 정확하고 정밀하게 반영할 수 있는지 알아보고자 한다. 재료 및 방법: 장비 정밀도는 GE Lunar Prodigy Advance 2대의 장비(P1, P2)와 HOLOGIC Spine Phantom (HSP)을 이용하여 각 장비에서 20번씩 스캔하여 팬텀을 이용한 정밀도 데이터를 획득하였고(Group 1), 여성 120명(평균나이 48.78, 20~60세)을 대상으로 각 장비에서 15명씩, 같은 환자가 두 번 촬영을 하여 각 검사자의 정밀도를 측정했다(Group 2), 또한 검사자의 정밀도는 팬텀(ASP)을 이용하여 매일 아침마다 질 관리 시행 후 얻은 데이터를 기준으로, 각각의 장비에서 HSP를 이용하여 각 장비에서 20번씩 스캔 후 데이터를 획득하여 검사자 정밀도 및 교차 보정 데이터를 산출하였고(Group 3), 여성 120명(평균나이 48.78, 20~60세)의 동일 환자를 대상으로 한 장비에서 한 번씩 교차로 측정하여 검사자 정밀도 및 교차보정 데이터를 산출하였다(Group 4). 결 과: Daily Q.C Data는 0.996g/cm², 변동계수(%CV) 0.08로 안정된 장비로서 Group 1에서 Mean±SD 및 %CV값은 ALP(P1: 1.064±0.002g/cm², %CV=0.190g/cm2, P2: 1.061±0.003g/cm², %CV=0.192), Group 2에서 Mean±SD 및 %CV값은 P1: 1.187±0.002g/cm², %CV=0.164g/cm², P2: 1.198±0.002g/cm², %CV= 0.163, Group 3에서의 Mean±2SD 및 %CV는 P1 - (spine: 0.001±0.03g/cm², %CV=0.94, Femur: 0.001±0.019g/cm², %CV=0.96), P2 - (spine: 0.002±0.018g/cm², %CV=0.55, Femur: 0.001±0.013g/cm², %CV=0.48), Group 4에서 Mean±2SD 및 %CV는, r값은 spine: 0.006±0.024g/cm², %CV=0.86, r=0.995, Femur: 0±0.014g/cm², %CV=0.54, r=0.998이였다. 결 론: HOLOGIC Spine Phantom과 LUNAR ASP %CV는 ISCD에서 규정한 정상오차 범위인 ±2% 안에 모두 포함되었고 BMD가 비교적 일정한 값을 유지하며 측정되어 뛰어난 재현성을 보였다. 하지만 Phantom은 환자의 체중이나 체지방 조성의 변화 등 임상적인 부분을 반영하는 데는 한계성을 갖고 있어 mis-calibration을 check하는데 유용할 것으로 판단된다. Group 3과 Group 4의 결과에서 환자를 하나의 장비로 두 번 측정한 값을 보았을 때와 두 대의 장비를 교차하여 측정한 값 모두 2SD값 이내에 포함되었고 선형회귀분석(Regression Analysis) r값이 0.99 이상으로 높은 정밀도와 상관도를 나타냄으로써 두 장비를 호환하여 추적검사를 시행하여도 영향이 없었다. 신뢰있는 BMD 산출을 위해서는 정기적으로 장비 및 검사자의 기능테스트와 이에 대한 적절한 교정행위가 이루어져야 할 것이다. Purpose: Because there is a difference depending on the environment as for an inspection equipment the important part of bone density scan and the precision/accuracy of a tester, the management of quality must be made systematically. The equipment failure caused by overload effect due to the aged equipment and the increase of a patient was made frequently. Thus, the replacement of equipment and additional purchases of new bonedensity equipment caused a compatibility problem in tracking patients. This study wants to know whether the clinical changes of patient"s bonedensity can be accurately and precisely reflected when used it compatiblly like the existing equipment after equipment replacement and expansion. Materials and methods: Two equipments of GE Lunar Prodigy Advance(P1 and P2) and the Phantom HOLOGIC Spine Road(HSP) were used to measure equipment precision. Each device scans 20 times so that precision data was acquired from the phantom(Group 1). The precision of a tester was measured by shooting twice the same patient, every 15 members from each of the target equipment in 120 women(average age 48.78, 20-60 years old)(Group 2). In addition, the measurement of the precision of a tester and the cross-calibration data were made by scanning 20 times in each of the equipment using HSP, based on the data obtained from the management of quality using phantom(ASP) every morning (Group 3). The same patient was shot only once in one equipment alternately to make the measurement of the precision of a tester and the cross-calibration data in 120 women(average age 48.78, 20-60 years old)(Group 4). Results: It is steady equipment according to daily Q.C Data with 0.996g/cm², change value(%CV) 0.08. The mean±SD and a %CV price are ALP in Group 1(P1: 1.064±0.002g/cm², %CV=0.190g/cm², P2: 1.061±0.003g/cm², %CV=0.192). The mean±SD and a %CV price are P1: 1.187±0.002g/cm², %CV=0.164 g/cm², P2: 1.198±0.002g/cm², %CV=0.163 in Group 2. The average error±2SD and %CV are P1 - (spine: 0.001±0.03g/cm², %CV=0.94, Femur: 0.001±0.019g/cm², %CV=0.96), P2 - (spine: 0.002±0.018 g/cm², %CV=0.55, Femur: 0.001±0.013g/cm², %CV=0.48) in Group 3. The average error±2SD, %CV, and r value was spine: 0.006±0.024g/cm², %CV=0.86, r=0.995, Femur: 0±0.014g/cm², %CV=0.54, r=0.998 in Group 4. Conclusion: Both LUNAR ASP CV% and HOLOGIC Spine Phantom are included in the normal range of error of ±2% defined in ISCD. BMD measurement keeps a relatively constant value, so showing excellent repeatability. The Phantom has homogeneous characteristics, but it has limitations to reflect the clinical part including variations in patient"s body weight or body fat. As a result, it is believed that quality control using Phantom will be useful to check mis-calibration of the equipment used. A value measured a patient two times with one equipment, and that of double-crossed two equipment are all included within 2SD Value in the Bland - Altman Graph compared results of Group 3 with Group 4. The r value of 0.99 or higher in Linear regression analysis(Regression Analysis) indicated high precision and correlation. Therefore, it revealed that two compatible equipment did not affect in tracking the patients. Regular testing equipment and capabilities of a tester, then appropriate calibration will have to be achieved in order to calculate confidential BMD.

Analysis of the Dynamic Diameter of Superthin Diamond Blades in the High Speed and Precison Dicing Process

Kenan Li,Mingzhi Wang,Feng Chen,Ning Yan,Qin Zou,Yucheng Zhao,Jianmin Li,Fu Zhao 한국정밀공학회 2019 International Journal of Precision Engineering and Vol.20 No.7

During the planar lightwave circuit (PLC) splitter manufacturing process, high precision is essential. Controlling machining precision is extremely difficult during the high-speed-and-precision dicing process, which is empirically related to the rotational speed, static diameter and elastic modulus of diamond dicing blades and other factors. However, the effect of changes in the outer diameter of blades on machining precision has been disregarded in the PLC manufacturing process. In this research, we proposed a dynamic diameter (Dd) for describing changes in a blade’s outer diameter during the high-speed machining process. Dd is positively correlated with machining precision. Here, we derived a formula for calculating Dd that is related to the rotational speed, static diameter, elastic modulus, Poisson’s ratio, density, cutting length and radial wear rate of diamond dicing blades. Furthermore, a series of experimental Dd values was obtained by changing the rotational speed, static diameter and elastic modulus of diamond dicing blades. These values are highly consistent with the calculated results. Our findings not only provide clues to compensate for diameter during high-speed-and-precision dicing process, but also offer guidelines for a new design route of diamond tools.

The Role of Excuses in Investing into Information Quality

( Byungjin Kwak ),( Hee-yeon Sunwoo ) 한국회계학회 2020 회계학연구 Vol.45 No.2

We analyze a manager's choice of precision of the information in a repeated cheap talk game under which the manager can excuse his lie by blaming the poor quality of the information system. In this setting, the manager can protect his reputation for the future period if the precision of information is imperfect, providing the manager with incentives to choose the poor quality of information system. However, the analyses show that even though the manager can excuse his lie relying on the imperfect precision, he finally selects perfect precision in most cases. The only exception is when he cannot publically determine the precision and he knows that his interest is not aligned with investor's. Even in this case, he randomizes his choice of information quality between perfect and imperfect precision. Otherwise, the manager chooses the perfect precision of information system. Our theoretical model provides important insights into why managers provide truthful or precise disclosures in most cases even if they are able to lie. In addition, our findings explain why they sometimes deviate from their normal practice and provide inaccurate, biased, or imprecise forecasts.

Are Currently Available Wearable Devices for Activity Tracking and Heart Rate Monitoring Accurate, Precise, and Medically Beneficial?

Fatema El-Amrawy,Mohamed Ismail Nounou 대한의료정보학회 2015 Healthcare Informatics Research Vol.21 No.4

The new wave of wireless technologies, fitness trackers, and body sensor devices can have great impact on healthcare systems and the quality of life. However, there have not been enough studies to prove the accuracy and precision of these trackers. The objective of this study was to evaluate the accuracy, precision, and overall performance of seventeen wearable devices currently available compared with direct observation of step counts and heart rate monitoring. Methods: Each participant in this study used three accelerometers at a time, running the three corresponding applications of each tracker on an Android or iOS device simultaneously. Each participant was instructed to walk 200, 500, and 1,000 steps. Each set was repeated 40 times. Data was recorded after each trial, and the mean step count, standard deviation, accuracy, and precision were estimated for each tracker. Heart rate was measured by all trackers (if applicable), which support heart rate monitoring, and compared to a positive control, the Onyx Vantage 9590 professional clinical pulse oximeter. Results: The accuracy of the tested products ranged between 79.8% and 99.1%, while the coefficient of variation (precision) ranged between 4% and 17.5%. MisFit Shine showed the highest accuracy and precision (along with Qualcomm Toq), while Samsung Gear 2 showed the lowest accuracy, and Jawbone UP showed the lowest precision. However, Xiaomi Mi band showed the best package compared to its price. Conclusions: The accuracy and precision of the selected fitness trackers are reasonable and can indicate the average level of activity and thus average energy expenditure.