• Smart Structures and Systems
• /
• 제2권2호
• /
• pp.189-208
• /
• 2006

Monitoring large-scale civil engineering structures such as offshore platforms and high-large buildings requires a large number of sensors of different types. Innovative sensor data information technologies are very extremely important for data transmission, storage and retrieval of large volume sensor data generated from large sensor networks. How to obtain the optimal sensor set and placement is more and more concerned by researchers in vibration-based SHM. In this paper, a method of determining the sensor location which aims to extract the dynamic parameter effectively is presented. The method selects the number and place of sensor being installed on or in structure by through the tolerance domain statistical inference algorithm combined with second order sensitivity technology. The method proposal first finds and determines the sub-set sensors from the theoretic measure point derived from analytical model by the statistical tolerance domain procedure under the principle of modal effective independence. The second step is to judge whether the sorted out measured point set has sensitive to the dynamic change of structure by utilizing second order characteristic value sensitivity analysis. A 76-high-building benchmark mode and an offshore platform structure sensor optimal selection are demonstrated and result shows that the method is available and feasible.

• Geomechanics and Engineering
• /
• 제27권5호
• /
• pp.511-525
• /
• 2021

In the design of underground excavation, the shear strength (SS) is a key characteristic. It describes the way the rock material resists the shear stress-induced deformations. In general, the measurement of the parameters related to rock shear strength is done through laboratory experiments, which are costly, damaging, and time-consuming. Add to this the difficulty of preparing core samples of acceptable quality, particularly in case of highly weathered and fractured rock. This study applies rock index test to the indirect measurement of the SS parameters of shale. For this aim, two efficient artificial intelligence methods, namely (1) adaptive neuro-fuzzy inference system (ANFIS) implemented by subtractive clustering method (SCM) and (2) support vector regression (SVR) optimized by Harmony Search (HS) algorithm, are proposed. Note that, it is the first work that predicts the SS parameters of shale through ANFIS-SCM and SVR-HS hybrid models. In modeling processes of ANFIS-SCM and SVR-HS, the results obtained from the rock index tests were set as inputs, while the SS parameters were set as outputs. By reviewing the obtained results, it was found that both ANFIS-SCM and SVR-HS models can provide acceptable predictions for interlocking and friction angle parameters, however, ANFIS-SCM showed a better generalization capability.

• Nuclear Engineering and Technology
• /
• 제53권12호
• /
• pp.4014-4021
• /
• 2021

If safety injection systems (SISs) do not work in the event of a loss-of-coolant accident (LOCA), the accident can progress to a severe accident in which the reactor core is exposed and the reactor vessel fails. Therefore, it is considered that a technology that provides recoverable maximum time for SIS actuation is necessary to prevent this progression. In this study, the corresponding time was defined as the golden time. To achieve the objective of accurately predicting the golden time, the prediction was performed using the deep fuzzy neural network (DFNN) with rule-dropout. The DFNN with rule-dropout has an architecture in which many of the fuzzy neural networks (FNNs) are connected and is a method in which the fuzzy rule numbers, which are directly related to the number of nodes in the FNN that affect inference performance, are properly adjusted by a genetic algorithm. The golden time prediction performance of the DFNN model with rule-dropout was better than that of the support vector regression model. By using the prediction result through the proposed DFNN with rule-dropout, it is expected to prevent the aggravation of the accidents by providing the maximum remaining time for SIS recovery, which failed in the LOCA situation.

• Imaging Science in Dentistry
• /
• 제52권3호
• /
• pp.275-281
• /
• 2022

Purpose: The aim of this study was to assess the performance of a deep learning system for permanent tooth germ detection on pediatric panoramic radiographs. Materials and Methods: In total, 4518 anonymized panoramic radiographs of children between 5 and 13 years of age were collected. YOLOv4, a convolutional neural network (CNN)-based object detection model, was used to automatically detect permanent tooth germs. Panoramic images of children processed in LabelImg were trained and tested in the YOLOv4 algorithm. True-positive, false-positive, and false-negative rates were calculated. A confusion matrix was used to evaluate the performance of the model. Results: The YOLOv4 model, which detected permanent tooth germs on pediatric panoramic radiographs, provided an average precision value of 94.16% and an F1 value of 0.90, indicating a high level of significance. The average YOLOv4 inference time was 90 ms. Conclusion: The detection of permanent tooth germs on pediatric panoramic X-rays using a deep learning-based approach may facilitate the early diagnosis of tooth deficiency or supernumerary teeth and help dental practitioners find more accurate treatment options while saving time and effort

• International Journal of Korean Welding Society
• /
• 제1권1호
• /
• pp.17-22
• /
• 2001

Because the quantitative relationships between welding parameters and welding result are not yet blown, optimal values of welding parameters for $CO_2$ robotic arc welding is a difficult task. Using the various artificial data processing methods may solve this difficulty. This research aims to develop an expert system for $CO_2$ robotic arc welding to recommend the optimal values of welding parameters. This system has three main functions. First is the recommendation of reasonable values of welding parameters. For such work, the relationships in between the welding parameters are investigated by the use of regression analysis and fuzzy system. The second is the estimation of bead shape by a neural network system. In this study the welding current voltage, speed, weaving width, and root gap are considered as the main parameters influencing a bead shape. The neural network system uses the 3-layer back-propagation model and a generalized delta rule as teaming algorithm. The last is the optimization of the parameters for the correction of undesirable weld bead. The causalities of undesirable weld bead are represented in the form of rules. The inference engine derives conclusions from these rules. The conclusions give the corrected values of the welding parameters. This expert system was developed as a PC-based system of which can be used for the automatic or semi-automatic $CO_2$ fillet welding with 1.2, 1.4, and 1.6mm diameter the solid wires or flux-cored wires.

• 한국컴퓨터정보학회논문지
• /
• 제14권7호
• /
• pp.1-8
• /
• 2009

본 연구에서는 소프트웨어 제품을 개발하여 테스팅을 거친 후 사용자에게 인도하는 시기를 결정하는 방출문제에 대하여 연구하였다. 따라서 최적 소프트웨어 방출 정책은 소프트웨어 요구 신뢰도를 만족시키고 소프트웨어 개발 및 유지 총비용을 최소화 시키는 정책을 수용해야 한다. 본 논문에서는 로그포아송 실행시간모형에 대하여 베이지안 모수 추정법(마코브체인 몬테칼로(MCMC) 기법 중에 하나인 깁스 샘플링과 메트로폴리스 알고리즘을 이용한 근사기법)이 사용되었다. 본 논문의 수치적인 예에서는 Musa의 T1 자료를 적용하여 최우수추정법과 베이지안 모수 추정과의 관계를 빅교하고 또한 최적 방출시기를 추정하였다.

• Structural Engineering and Mechanics
• /
• 제82권4호
• /
• pp.445-458
• /
• 2022

Some of the control systems used in engineering structures that use sensors and decision systems have some time delay reducing efficiency of the control system or even might make it unstable. In this research, in addition to considering the effect of the time delay in vibration control process, predictive control is used to compensate the time delay. A semi-active vibration control approach with the help of magneto-rheological dampers is implemented. In addition to using fuzzy inference system to determine the appropriate control voltage for MR damper, structural behavior prediction system and specifying future responses are also used such that the time delays occurring within control process are overcome. For this purpose, determination of prediction horizon is conducted for one, five, and ten steps ahead for single degree of freedom structures with periods ranging from 0.1 to 4 seconds, subjected to twenty earthquake excitations. The amount of time delay applied to the control system is 0.1 seconds. The obtained results indicate that for 0.1 second time delay, average prediction error values compared to the case without time delay is 3.47 percent. Having 0.1 second time delay in a semi-active control system reduces its efficiency by 11.46 percent; while after providing the control system with structure behavior prediction, the difference in the results for the control system without time delay is just 1.35 percent on average; indicating a 10.11 percent performance improvement for the control system.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2021년도 춘계학술대회
• /
• pp.538-540
• /
• 2021

본 논문에서는 드론의 표적탐지 임무 수행 시 상대운동 정보 제공을 위하여 지정된 표적을 탐지하고 인식하는 스마트 표적탐지 시스템을 제안하였다. 제안된 시스템은 적절한 정확도(i.e. mAP, IoU) 및 높은 실시간성을 동시에 확보할 수 있는 알고리즘을 개발하는데 중점을 두었다. 제안된 시스템은 Google Inception V2 딥러닝 모델의 100k 학습 후 test 결과가 1.0에 가까운 정확성을 보였고 실시간성도 Nvidia GTX 2070 Max-Q를 기반으로 한 고성능 노트북 활용 시에 추론 속도가 약 60-80[Hz]를 기록하였다. 제안된 스마트 표적탐지 시스템은 드론과 같이 운용되어 컴퓨터 영상처리를 활용하여 표적을 자동으로 인식하고 표적을 따라가면서 감시정찰 임무를 성공적으로 수행하는데 도움이 될 것이다.

• 한국정보통신학회논문지
• /
• 제10권7호
• /
• pp.1319-1325
• /
• 2006

요 분석은 임상에서의 기본적인 검사항목으로 숙련된 간호사에 의한 육안검사를 시행한다. 최근에는 분석량의 증가와 분석 시간의 단축을 위하여 자동화된 요 분석 시스템을 이용하여 측정한다. 그러나 이들 시스템은 기기별 로 나타나는 결과에 차이가 발생하고 있다. 따라서 요의 컬러에 따른 정확한 검사를 위하여 새로운 요 컬러 분류 알고리즘이 요구된다. 본 논문은 퍼지 논리와 신경회로망 알고리즘을사용하여 요 분석 시스템의 지능형 컬러 분류기를 제작하였다. 입력 파라미터는 전처리 과정을 거친 RGB 3가지 색상을 사용하였다. 구현된 분류기는 퍼지 논리와 신경회로망 알고리즘을 사용하였으며, 적색, 녹색, 청색의 3 가지 입력 데이터를 사용하여 9 가지 시료에 대한 $3{\sim}7$ 개의 각 단계별 분류를 수행하도록 구현하였다. 실험에 사용된 검체는 표준 시약을 사용하였으며, 요 분석 시스템을 위한 개별 표준시료에 따른 분류기의 성능을 비교하고, 신뢰성 및 임상적용가능성 여부를 검토하였다. 설험 결과 지능형 컬러 분류기는 많은 검사 항목에서 육안검색보다 좋은 결과를 보였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제7권1호
• /
• pp.81-98
• /
• 2013

Since Probabilistic Latent Semantic Analysis (PLSA) and Latent Dirichlet Allocation (LDA) were introduced, many revised or extended topic models have appeared. Due to the intractable likelihood of these models, training any topic model requires to use some approximation algorithm such as variational approximation, Laplace approximation, or Markov chain Monte Carlo (MCMC). Although these approximation algorithms perform well, training a topic model is still computationally expensive given the large amount of data it requires. In this paper, we propose a new method, called non-simultaneous sampling deactivation, for efficient approximation of parameters in a topic model. While each random variable is normally sampled or obtained by a single predefined burn-in period in the traditional approximation algorithms, our new method is based on the observation that the random variable nodes in one topic model have all different periods of convergence. During the iterative approximation process, the proposed method allows each random variable node to be terminated or deactivated when it is converged. Therefore, compared to the traditional approximation ways in which usually every node is deactivated concurrently, the proposed method achieves the inference efficiency in terms of time and memory. We do not propose a new approximation algorithm, but a new process applicable to the existing approximation algorithms. Through experiments, we show the time and memory efficiency of the method, and discuss about the tradeoff between the efficiency of the approximation process and the parameter consistency.