• 환경생물
• /
• 제39권4호
• /
• pp.581-589
• /
• 2021

본 연구는 메꽃과 6종의 식물에 대해 신속하고 비파괴적으로 분류하기 위해 근적외선(Vis-NIR) 스펙트럼을 이용하였고 데이터의 전처리와 머신러닝 기술을 적용하였다. 전국적으로 분포하는 메꽃과 6종에 대해 야외에서 휴대용 분광기를 이용하여 판별하였다. 식물의 잎의 표면에서 400~1,075 nm의 근적외선 스펙트럼(1.5 nm)을 수집하였다. 수집된 스펙트럼 데이터는 3가지의 전처리와 raw데이터를 이용하였고 4종류의 머신러닝 모델을 적용하여 높은 판별 정확도를 확인하였다. 전처리와 머신러닝 모델의 조합을 통해 분석된 판별의 정확도는 43~99%의 범위로 분석되었고, standard normal variate 전처리와 support vector machine 머신러닝 모델의 조합에서 판별 정확도가 98.6%로 가장 높게 나타났다. 본 연구에서 수집된 스펙트럼은 식물의 성장단계, 다양한 측정 지역 및 잎에서의 측정 위치 등과 같은 요인과 더불어 데이터 분석을 위한 조건으로 최적의 전처리와 머신러닝 기술을 적용한다면 메꽃과 식물의 야외에서의 정확한 분류가 가능하고 이들 식물의 효과적인 관리와 모니터링에 활용할 수 있을 것으로 판단되었다.

• Journal of Multimedia Information System
• /
• 제6권2호
• /
• pp.49-60
• /
• 2019

Plants are very crucial for life on Earth. There is a wide variety of plant species available, and the number is increasing every year. Species knowledge is a necessity of various groups of society like foresters, farmers, environmentalists, educators for different work areas. This makes species identification an interdisciplinary interest. This, however, requires expert knowledge and becomes a tedious and challenging task for the non-experts who have very little or no knowledge of the typical botanical terms. However, the advancements in the fields of machine learning and computer vision can help make this task comparatively easier. There is still not a system so developed that can identify all the plant species, but some efforts have been made. In this study, we also have made such an attempt. Plant identification usually involves four steps, i.e. image acquisition, pre-processing, feature extraction, and classification. In this study, images from Swedish leaf dataset have been used, which contains 1,125 images of 15 different species. This is followed by pre-processing using Gaussian filtering mechanism and then texture and color features have been extracted. Finally, classification has been done using Multiclass-support vector machine, which achieved accuracy of nearly 93.26%, which we aim to enhance further.

• Nuclear Engineering and Technology
• /
• 제55권2호
• /
• pp.493-505
• /
• 2023

Comprehensive condition monitoring of large industry systems such as nuclear power plants (NPPs) is essential for safety and maintenance. In this study, we developed novel system-scale diagnostic technology based on deep-learning and IR thermography that can efficiently and cost-effectively classify system conditions using compact Raspberry Pi and IR sensors. This diagnostic technology can identify the presence of an abnormality or accident in whole system, and when an accident occurs, the type of accident and the location of the abnormality can be identified in real-time. For technology development, the experiment for the thermal image measurement and performance validation of major components at each accident condition of NPPs was conducted using a thermal-hydraulic integral effect test facility with compact infrared sensor modules. These thermal images were used for training of deep-learning model, convolutional neural networks (CNN), which is effective for image processing. As a result, a proposed novel diagnostic was developed that can perform diagnosis of components, whole system and accident classification using thermal images. The optimal model was derived based on the modern CNN model and performed prompt and accurate condition monitoring of component and whole system diagnosis, and accident classification. This diagnostic technology is expected to be applied to comprehensive condition monitoring of nuclear power plants for safety.

• 한국초등과학교육학회지:초등과학교육
• /
• 제39권2호
• /
• pp.284-295
• /
• 2020

The purpose of this study is to explore the possibility of artificial intelligence convergence education (AICE) in elementary biological classification learning. First, the possibility of AICE was analyzed in the field of 2015 revised elementary life science curriculum. The artificial intelligence biological classification (AIBC) education program targeted plant life. The possibility of AICE in the elementary life science curriculum was suggested through the consultation process of three elementary science education experts. The AIBC education program was developed through the review process of elementary education experts. The results of this study are as follows. First, 8(32%) achievement standards were available for AICE in elementary life science. Second, 18(86%) of the 21 items reviewed by the experts for the AIBC education program developed in this study were positively evaluated. Third, in this study, through the analysis of the possibility of AIBC in the elementary life field and the review of the experts, the AIBC education program including teaching and learning models, strategies, and guidance was developed. The results of this study were based on the review of the experts, and as a follow-up study, applied research to elementary students is needed. It is also hoped that various studies on AICE will be conducted not only in the life field but also in science and other fields. Finally, we expect that the results of this study will be applied to bio-classification learning to help students improve classification capabilities and generate classification knowledge.

• 한국원자력학회:학술대회논문집
• /
• 한국원자력학회 1995년도 추계학술발표회논문집(1)
• /
• pp.293-299
• /
• 1995

A methodology is applied to identify tile learning trend related to the safety and availability of U.S. commercial nuclear power plants. The application is intended to aid in reducing likelihood of human errors. To assure that tile methodology ran be easily adapted to various types of classification schemes of operation data, a data bank classified by the Transient Analysis Classification and Evaluation(TRACE) scheme is selected for the methodology. The significance criteria for human-initiated events affecting tile systems and for events caused by human deficiencies were used. Clustering analysis was used to identify the learning trend in multi-dimensional histograms. A computer rode is developed based on tile K-Means algorithm and applied to find the learning period in which error rates are monotonously decreasing with plant age.

• Nuclear Engineering and Technology
• /
• 제55권6호
• /
• pp.2026-2033
• /
• 2023

Due to Thailand's nuclear energy public acceptance problem, the understanding of nuclear energy public perception was the key factor affecting to re-consideration of the nuclear energy program. Thailand Institute of Nuclear Technology and its alliances together developed the classification model for the nuclear energy public perception from the big data comments on social media using Facebook using deep transfer learning. The objective was to insight into the Thailand nuclear energy public perception on Facebook social media platform using sentiment analysis. The supervised learning was used to generate up-to-date classification model with more than 80% accuracy to classify the public perception on nuclear power plant news on Facebook from 2009 to 2022. The majority of neutral sentiments (80%) represented the opportunity for Thailand to convince people to receive a better nuclear perception. Negative sentiments (14%) showed support for other alternative energies due to nuclear accident concerns while positive sentiments (6%) expressed support for innovative nuclear technologies.

• Nuclear Engineering and Technology
• /
• 제52권10호
• /
• pp.2262-2273
• /
• 2020

Digital twin technology can provide significant value for the prognostics and health management (PHM) of critical plant components by improving insight into system design and operating conditions. Digital twinning of systems can be utilized for anomaly detection, diagnosis and the estimation of the system's remaining useful life in order to optimize operations and maintenance processes in a nuclear plant. In this regard, a conceptual framework for the application of digital twin technology for the prognosis of Control Element Drive Mechanism (CEDM), and a data-driven approach to anomaly detection using coil current profile are presented in this study. Health management of plant components can capitalize on the data and signals that are already recorded as part of the monitored parameters of the plant's instrumentation and control systems. This work is focused on the development of machine learning algorithm and workflow for the analysis of the CEDM using the recorded coil current data. The workflow involves features extraction from the coil-current profile and consequently performing both clustering and classification algorithms. This approach provides an opportunity for health monitoring in support of condition-based predictive maintenance optimization and in the development of the CEDM digital twin model for improved plant safety and availability.

• Nuclear Engineering and Technology
• /
• 제55권2호
• /
• pp.506-515
• /
• 2023

This study is to find a method to learn engine sound after the start-up of a diesel generator installed in nuclear power plant with an unsupervised deep learning algorithm (CNN autoencoder) and a new method to predict the failure of a diesel generator using it. In order to learn the sound of a diesel generator with a deep learning algorithm, sound data recorded before and after the start-up of two diesel generators was used. The sound data of 20 min and 2 h were cut into 7 s, and the split sound was converted into a spectrogram image. 1200 and 7200 spectrogram images were created from sound data of 20 min and 2 h, respectively. Using two different deep learning algorithms (CNN autoencoder and binary classification), it was investigated whether the diesel generator post-start sounds were learned as normal. It was possible to accurately determine the post-start sounds as normal and the pre-start sounds as abnormal. It was also confirmed that the deep learning algorithm could detect the virtual abnormal sounds created by mixing the unusual sounds with the post-start sounds. This study showed that the unsupervised anomaly detection algorithm has a good accuracy increased about 3% with comparing to the binary classification algorithm.

• 인간식물환경학회지
• /
• 제24권5호
• /
• pp.485-498
• /
• 2021

Background and objective: This study was conducted to find out the status of plant utilization in the current textbooks by analyzing the plants by grade and subject in the national textbooks for all elementary school grades in the 2015 revised curriculum in Korea. Methods: The data collected was analyzed using Microsoft Office Excel to obtain the frequency and ratio of collected plant data and SPSS for Windows 26.0 to determine learning content areas by grade and the R program was used to visualize the learning content areas. Results: A total of 232 species of plants were presented 1,047 times in the national textbooks. Based on an analysis of the plants presented by grade, the species that continued to increase in the lower grades tended to decrease in the fifth and sixth grades, the upper grades of elementary school. As for the number and frequency of plant species by subject, Korean Language had the highest number and frequency of plant species. The types of presentation of plants in textbooks were mainly text, followed by illustrations and photos of plants, which were largely used in first grade textbooks. In addition, as for the area of learning contents in which plants are used, in the lower grades, plants were used in the linguistic domain, and in the upper grades, in the botanical and environmental domains of the natural sciences. Herbaceous plants were presented more than woody plants, and according to an analysis of the plants based on the classification of crops, horticultural crops were presented the most, followed by food crops. Out of horticultural crops, flowering plants were found the most diversity with 63 species, but the plants that appeared most frequently were fruit trees that are commonly encountered in real life. Conclusion: As a result of this study, various plant species were included in elementary school textbooks, but most of them were horticultural crops encountered in real life depending on their use. Nevertheless, plant species with high frequency have continued a similar trend of frequency from the previous curriculums. Therefore, in the next curriculum, plant learning materials should be reflected according to social changes and students' preference for plants.

• 한국산학기술학회논문지
• /
• 제22권5호
• /
• pp.7-14
• /
• 2021

토마토 작물은 병해충의 영향을 많이 받기 때문에 이를 예방하지 않으면 농업 경제에 막대한 손실을 초래할 수 있다. 따라서 토마토의 다양한 병해충의 진단을 빠르고 정확하게 진단하는 시스템이 요구된다. 본 논문에서는 ImageNet 데이터 셋 상에서 다양하게 사전 학습된 딥러닝 기반 CNN 모델을 적용하여 토마토의 9가지 병해충 및 정상인 경우의 클래스를 분류하는 시스템을 제안한다. PlantVillage 데이터 셋으로부터 발췌한 토마토 잎의 이미지 셋을 3가지 딥러닝 기반 CNN 구조를 갖는 ResNet, Xception, DenseNet의 입력으로 사용한다. 기본 CNN 모델 위에 톱-레벨 분류기를 추가하여 제안 모델을 구성하였으며, 훈련 데이터 셋에 대해 5-fold 교차검증 기법을 적용하여 학습시켰다. 3가지 제안 모델의 학습은 모두 기본 CNN 모델의 계층을 동결하여 학습시키는 전이 학습과 동결을 해제한 후 학습률을 매우 작은 수로 설정하여 학습시키는 미세 조정 학습 두 단계로 진행하였다. 모델 최적화 알고리즘으로는 SGD, RMSprop, Adam을 적용하였다. 실험 결과는 RMSprop 알고리즘이 적용된 DenseNet CNN 모델이 98.63%의 정확도로 가장 우수한 결과를 보였다.