• Korean Chemical Engineering Research
• /
• v.59 no.2
• /
• pp.191-199
• /
• 2021

A distillation column, which is a main facility of the chemical process, separates the desired product from a mixture by using the difference of boiling points. The distillation process requires the optimization and the prediction of operation because it consumes much energy. The target process of this study is difficult to operate efficiently because the composition of feed flow is not steady according to the supplier. To deal with this problem, we could develop a data-driven model to predict operating conditions. However, data preprocessing is essential to improve the predictive performance of the model because the raw data contains outlier and noise. In this study, after optimizing the predictive model based long-short term memory (LSTM) and Random forest (RF), we used a low-pass filter and one-class support vector machine for data preprocessing and compared predictive performance according to the method and range of the preprocessing. The performance of the predictive model and the effect of the preprocessing is compared by using R2 and RMSE. In the case of LSTM, R2 increased from 0.791 to 0.977 by 23.5%, and RMSE decreased from 0.132 to 0.029 by 78.0%. In the case of RF, R2 increased from 0.767 to 0.938 by 22.3%, and RMSE decreased from 0.140 to 0.050 by 64.3%.

• Journal of the Korea Convergence Society
• /
• v.12 no.3
• /
• pp.177-186
• /
• 2021

The 2015 revised curriculum is structured around the core competencies of the 21st century, this is in line with the world's flow of education, such as OECD Education 2030. A future practical leading model was studied to provide a variety of creative teaching and learning experiences to elementary and Secondary students using intelligent information technology to cultivate core competencies such as ICT and computing thinking. In order for this practical model to stably settle the school field, the training was planned and operated to strengthen the creative convergence education capacity required by the teachers at the unit school through various types of the training. In particular, a nationwide administrators training program was operated for three years, reflecting the new curriculum, teaching and learning methods, and evaluation that can lead to future convergence talent training. In this paper, the perception of creative convergence education was investigated and analyzed considering the influence that administrators may have on the school field. Based on this, through the three-year operation results of the training, it was intended to establish a new training method for stable access to future creative convergence education under the post-corona era's social issues.

• Journal of Science Education
• /
• v.47 no.2
• /
• pp.117-126
• /
• 2023

In an elementary school science teacher's guide, the guide in details that can help elementary school teachers with difficulties when preparing and conducting science classes are essential. To gain insights into the development of the 2022 revised curriculum elementary school science teacher's guide, it is necessary to investigate the perceptions and needs of elementary school teacher's guide in details. In this study, we developed a questionnaire by analyzing the components and design of an elementary school science teacher's guide in details to explore elementary school teachers' perceptions and needs. For this purpose, we first investigated elementary school teachers' perceptions of their needs, satisfaction, and the utilization of each component of the current guide in details. Next, we investigated teachers' needs regarding the specific components and design of a guide. The findings were as follows. First, elementary school teachers were delighted with the components that help them prepare and conduct lessons. Second, elementary school teachers wanted an easy-to-read design with a layout that allowed them to see the components they needed for their lessons at a glance. In conclusion, the elementary school science teacher's guide in details to be readable and organized to provide at-a-glance information on lesson preparation, lesson flow, and teaching and learning materials needed for science teaching-learning. Based on the results of this study, new approaches and attempts should be made to develop a textbook that elementary school teachers can utilize in the future.

• Journal of Korea Water Resources Association
• /
• v.55 no.spc1
• /
• pp.1283-1293
• /
• 2022

The purpose of this study is to predict residual chlorine in order to maintain stable residual chlorine concentration in sedimentation basin by using artificial intelligence algorithms in water treatment process employing pre-chlorination. Available water quantity and quality data are collected and analyzed statistically to apply into mathematical multiple regression and artificial intelligence models including multi-layer perceptron neural network, random forest, long short term memory (LSTM) algorithms. Water temperature, turbidity, pH, conductivity, flow rate, alkalinity and pre-chlorination dosage data are used as the input parameters to develop prediction models. As results, it is presented that the random forest algorithm shows the most moderate prediction result among four cases, which are long short term memory, multi-layer perceptron, multiple regression including random forest. Especially, it is result that the multiple regression model can not represent the residual chlorine with the input parameters which varies independently with seasonal change, numerical scale and dimension difference between quantity and quality. For this reason, random forest model is more appropriate for predict water qualities than other algorithms, which is classified into decision tree type algorithm. Also, it is expected that real time prediction by artificial intelligence models can play role of the stable operation of residual chlorine in water treatment plant including pre-chlorination process.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2020.06a
• /
• pp.261-261
• /
• 2020

물관리의 기본이 되는 연속적인 유량 자료 확보를 위해서는 정확도 높은 수위-유량 관계 곡선식 개발이 필수적이다. 수위-유량 관계곡선식은 모든 수문시설 설계의 기초가 되며 홍수, 가뭄 등 물재해 대응을 위해서도 중요한 의미를 가지고 있다. 그러나 일반적으로 유량 측정은 많은 비용과 시간이 들고, 식생성장, 단면변화 등의 통제특성(control)이 변함에 따라 구간분리, 기간분리와 같은 비선형적인 양상이 나타나 자료 해석에 어려움이 존재한다. 특히, 국내 하천의 경우 자연적 및 인위적인 환경 변화가 다양하여 지점 및 기간에 따라 세밀한 분석이 요구된다. 머신러닝(Machine Learning)이란 데이터를 통해 컴퓨터가 스스로 학습하여 모델을 구축하고 성능을 향상시키는 일련의 과정을 뜻한다. 기존의 수위-유량 관계곡선식은 개발자의 판단에 의해 데이터의 종류와 기간 등을 설정하여 회귀식의 파라미터를 산출한다면, 머신러닝은 유효한 전체 데이터를 이용해 스스로 학습하여 자료 간 상관성을 찾아내 모델을 구축하고 성능을 지속적으로 향상 시킬 수 있다. 머신러닝은 충분한 수문자료가 확보되었다는 전제 하에 복잡하고 가변적인 수자원 환경을 반영하여 유량 추정의 정확도를 지속적으로 향상시킬 수 있다는 이점을 가지고 있다. 본 연구는 머신러닝의 대표적인 알고리즘들을 활용하여 유량을 추정하는 모델을 구축하고 성능을 비교·분석하였다. 대상지역은 안정적인 수량을 확보하고 있는 한강수계의 거운교 지점이며, 사용자료는 2010~2018년의 시간, 수위, 유량, 수면폭 등 이다. 프로그램은 파이썬을 기반으로 한 머신러닝 라이브러리인 사이킷런(sklearn)을 사용하였고 알고리즘은 랜덤포레스트 회귀, 의사결정트리, KNN(K-Nearest Neighbor), rgboost을 적용하였다. 학습(train) 데이터는 입력자료 종류별로 조합하여 6개의 세트로 구분하여 모델을 구축하였고, 이를 적용해 검증(test) 데이터를 RMSE(Roog Mean Square Error)로 평가하였다. 그 결과 모델 및 입력 자료의 조합에 따라 3.67~171.46로 다소 넓은 범위의 값이 도출되었다. 그 중 가장 우수한 유형은 수위, 연도, 수면폭 3개의 입력자료를 조합하여 랜덤포레스트 회귀 모델에 적용한 경우이다. 비교를 위해 동일한 검증 데이터를 한국수문조사연보(2018년) 내거운교 지점의 수위별 수위-유량 곡선식을 이용해 유량을 추정한 결과 RMSE가 3.76이 산출되어, 머신러닝이 세분화된 수위-유량 곡선식과 비슷한 수준까지 성능을 내는 것으로 확인되었다. 본 연구는 양질의 유량자료 생산을 위해 기 구축된 수문자료를 기반으로 머신러닝 기법의 적용 가능성을 검토한 기초 연구로써, 국내 효율적인 수문자료 측정 및 수위-유량 곡선 산출에 도움이 될 수 있을 것으로 판단된다. 향후 수자원 환경 및 통제특성에 영향을 미치는 다양한 영향변수를 파악하기 위해 기상자료, 취수량 등의 입력 자료를 적용할 필요가 있으며, 머신러닝 내 비지도학습인 딥러닝과 같은 보다 정교한 모델에 대한 추가적인 연구도 수행되어야 할 것이다.

• Journal of the Institute of Convergence Signal Processing
• /
• v.24 no.1
• /
• pp.42-50
• /
• 2023

The development of technology has achieved digital innovation throughout the media industry, including production techniques and editing technologies, and has brought diversity in the form of consumer viewing through the OTT service and streaming era. The convergence of big data and deep learning networks automatically generated text in format such as news articles, novels, and scripts, but there were insufficient studies that reflected the author's intention and generated story with contextually smooth. In this paper, we describe the flow of pictures in the storyboard with image caption generation techniques, and the automatic generation of story-tailored scenarios through language models. Image caption using CNN and Attention Mechanism, we generate sentences describing pictures on the storyboard, and input the generated sentences into the artificial intelligence natural language processing model KoGPT-2 in order to automatically generate scenarios that meet the planning intention. Through this paper, the author's intention and story customized scenarios are created in large quantities to alleviate the pain of content creation, and artificial intelligence participates in the overall process of digital content production to activate media intelligence.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.7
• /
• pp.33-40
• /
• 2024

The live weight of livestock is important information for managing their health and housing conditions, and it can be used to determine the optimal amount of feed and the timing of shipment. In general, it takes a lot of human resources and time to weigh livestock using a scale, and it is not easy to measure each stage of growth, which prevents effective breeding methods such as feeding amount control from being applied. In this paper, we aims to improve the accuracy of weight measurement of piglets, weaned pigs, nursery pigs, and fattening pigs by collecting, analyzing, learning, and predicting video and image data in animal husbandry and pig farming. For this purpose, we trained using Pytorch, YOLO(you only look once) 5 model, and Scikit Learn library and found that the actual and prediction graphs showed a similar flow with a of RMSE(root mean square error) 0.4%. and MAPE(mean absolute percentage error) 0.2%. It can be utilized in the mammalian pig, weaning pig, nursery pig, and fattening pig sections. The accuracy is expected to be continuously improved based on variously trained image and video data and actual measured weight data. It is expected that efficient breeding management will be possible by predicting the production of pigs by part through video reading in the future.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.4
• /
• pp.693-700
• /
• 2024

As of the end of March 2022, the total area of domestic industrial complexes is 606 km2, which is only about 0.6% of the total land area. However, as of 2018, the annual energy consumption of domestic industrial complexes is 110,866.1 thousand TOE, accounting for 53.5% of the country's total energy consumption and 83.1% of the entire industrial sector energy consumption. In addition, industrial complexes have a significant impact on the environment, accounting for 45.1% of the country's total greenhouse gas emissions and 76.8% of industrial sector greenhouse gas emissions. Under this background, in this study, in order to contribute to the energy efficiency of industrial complexes, a prediction study on energy demand and supply for an industrial complex in Korea using machine learning was conducted. In addition, a simulator UI screen was designed to more efficiently convey information on energy demand/supply prediction results and energy consumption status. Among the machine learning algorithms, Multi-Layer Perceptron (MLP) was used, and Bayesian Optimization was applied as an optimization technique for the prediction model. The energy prediction model for the industrial complex built in this study showed a prediction accuracy of 87.90% for compressed air demand and 99.54% for the flow rate available for the public air compressor.

• Journal of Korean Home Economics Education Association
• /
• v.29 no.4
• /
• pp.161-193
• /
• 2017

The purpose of this study was to develop and evaluate pre-parenthood education program(PPEP) based on Home Economics(HE) subject for high school students. The development and evaluation of PPEP based on HE subject in this study followed ADDIE model except implementation through 4 processes such as analysis, design, development, and evaluation. First, program development directions were set in three aspects such as 'general development', 'contents', and 'teaching and learning methods'. Themes of the program are 11 in total such as '1. Parenting, what is being a parent', '2. Choosing your spouse, happy marital relationship, the best gift to your children', '3. Pregnancy and birth, a moving meeting with a new life', '4. Taking care of a new born infant for 24 hours', '5. Taking care of infants, relationship with my lovely baby, attachment', '6. Taking care of young children, my child from another planet', '7. Parents and children in healthy family', '8. Parent-child relationship, wise parents to make effective interaction with their children', '9. Parents safety manager at home,', '10. Practice to take care of infants', and '11. Practice of community nurturing support service development'. In particular, learning activities of the program have major characteristics such as 1) utilization of cases including practice problems related to parenting, 2) community exchange activities utilizing learned knowledge and techniques, 3) actual life project activities utilizing learning contents related with parenting, 4) activities inducing positive changes in current life of high school students, and 5) practice activities for the necessities of life such as food, clothing and shelter supporting development of children. Second, the program was developed according to the design. Teaching-learning plans and materials for 17 classes were developed according to 11 themes. The developed plans include class flow and teacher's reference. It starts with receiving a class-related message from a virtual child at the introduction stage and ended with replying to the message by summarizing contents of the class and making a promise as a parent-to-be. That is the basic frame of class flow. Learning materials included various plans and reports necessary for learning activities and they are prepared in details so that they can be play the role of textbooks in regular curriculum. Third, evaluation of developed program was executed by a 5 point Likert scale survey on 13 HE experts on two aspects of program development process and program development results. In the evaluation of development process, mean value was 4.61 and index of content validity was 97.4%. For development results, mean value was 4.37 and index of content validity was 86.9%. These values showed that validity in the development process and results in this study was highly secured and confirmed that PPEP based on HE was appropriate and valid to enhance parent qualifications of high school learners.

• Journal of the Korean School Mathematics Society
• /
• v.11 no.1
• /
• pp.31-54
• /
• 2008

The discovery method is known to be the most effective in improving students' mathematical thinking. Recently, the long-term slow thinking(LST) is suggested as a possible method to implement the discovery method into the real classroom. In this concept, we examined whether students can solve such a problem, as appears to be beyond their ability, by themselves(LST) or not. 10 middle school students of the ninth grade were selected for the study, who had no previous experience on the infinite concept of calculus of the high school course. They had tried to solve a problem about the calculus by their LST for three days. Two of students solved the problem by themselves and seven of students solved it with help of hints. This result shows that if students are given the opportunity of LST for rather difficult mathematical problem with appropriate guidance of a teacher, they might solve it by themselves. That is, LST could be a possible method for implementation of the discovery method.