• 한국재난정보학회 논문집
• /
• 제19권4호
• /
• pp.897-907
• /
• 2023

연구목적: 본 연구에서는 단일 24GHz FMCW레이더를 사용하여 수집된 적은 양의 학습데이터로 학습된 AI 모델을 사용하여 학습되지 않은 사람의 3가지 자세를 구분하고자 한다. 연구방법: 실내에서 학습 대상자들의 3가지 자세(서기, 앉기, 눕기)에 대한 FFT데이터를 수집하여 2D 이미지로 변환시킨 후 제안하는 2D CNN 모델로 학습시켜 학습에 사용되지 않은 새로운 대상자들의 자세를 잘 구분할 수 있는지 실험을 통해 정확도를 분석하였다. 연구결과: 제안하는 기법을 통해 3가지 자세의 평균 정확도가 89.99%임을 보였고, 기존의 1D CNN이나 SVM 보다 성능이 향상되었다. 결론: 실내에서 재난이 발생하는 경우 단일 FMCW 레이더와 AI 기법을 통해 요구조자의 자세를 추정하고자 하였으며, 학습되지 않은 대상자의 자세도 높은 정확도로 추정이 가능함을 실험을 통해 확인하였다.

• 지능정보연구
• /
• 제24권1호
• /
• pp.253-266
• /
• 2018

본 연구는 질의 응답(QA) 시스템에서 사용하는 개체명 인식(NER)의 성능을 향상시키기 위하여 시퀀스 태깅 방법론을 적용한 새로운 방법론을 제안한다. 사용자의 질의를 입력 받아 데이터베이스에 저장된 정답을 추출하기 위해서는 사람의 언어를 컴퓨터가 알아들을 수 있도록 구조화 질의어(SQL)와 같은 데이터베이스의 언어로 전환하는 과정이 필요한데, 개체명 인식은 사용자의 질의에서 데이터베이스에 포함된 클래스나 데이터 명을 식별하는 과정이다. 기존의 데이터베이스에서 질의에 포함된 단어를 검색하여 개체명을 인식하는 방식은 동음이의어와 문장성분 구를 문맥을 고려하여 식별하지 못한다. 다수의 검색 결과가 존재하면 그들 모두를 결과로 반환하기 때문에 질의에 대한 해석이 여러 가지가 나올 수 있고, 계산을 위한 시간복잡도가 커진다. 본 연구에서는 이러한 단점을 극복하기 위해 신경망 기반의 방법론을 사용하여 질의가 가지는 문맥적 의미를 반영함으로써 이러한 문제를 해결하고자 했고 신경망 기반의 방법론의 문제점인 학습되지 않은 단어에 대해서도 문맥을 통해 식별을 하고자 하였다. Sequence Tagging 분야에서 최신 기술인 Bidirectional LSTM-CRF 모델을 도입함으로써 신경망 모델이 가진 단점을 해결하였고, 학습되지 않은 단어에 대해서는 온톨로지 기반 특성치를 활용하여 문맥을 반영한 추론을 사용하였다. 음악 도메인의 온톨로지(Ontology) 지식베이스를 대상으로 실험을 진행하고 그 성능을 평가하였다. 본 연구에서 제안한 방법론인 L-Bidirectional LSTM-CRF의 성능을 정확하게 평가하기 위하여 학습에 포함된 단어들뿐만 아니라 학습에 포함되지 않은 단어들도 포함한 질의를 평가에 사용하였다. 그 결과 L-Bidirectional LSTM-CRF 모형을 재학습 시키지 않아도 학습에 포함되지 않은 단어를 포함한 질의에 대한 개체명 인식이 가능함을 확인하였고, 전체적으로 개체명 인식의 성능이 향상됨을 확인할 수 있었다.

• 한국지반공학회지:지반
• /
• 제12권6호
• /
• pp.101-114
• /
• 1996

Coulomb토압이론은 실무에서 주로 사용되고 있음에도 불구하고 점성토로 뒤채움된 옹벽에 대한 일반적인 조건에 적용할 수 있는 토압이론식이 존재하지 않는다. 여기서는 점성토로 뒤채움한 중력식 옹벽에 대해서 배수 및 비배수 해석으로 인장균열이 부시된 경우와 고려된 경우에 각각 토압이론식을 유도하였다. 그리고 상기의 조건들에 대해서 전주동토압과 인장균열 깊이를 구하여 설계도표를 작성하였다.

• Transactions on Control, Automation and Systems Engineering
• /
• 제2권1호
• /
• pp.40-46
• /
• 2000

In this paper, a transient identification based on a Hidden Markov Model (HMM) has been suggested and evaluated experimentally for the classification of transients in the dynamic process. The transient can be identified by its unique time dependent patterns related to the principal variables. The HMM, a double stochastic process, can be applied to transient identification which is a spatial and temporal classification problem under a statistical pattern recognition framework. The HMM is created for each transient from a set of training data by the maximum-likelihood estimation method. The transient identification is determined by calculating which model has the highest probability for the given test data. Several experimental tests have been performed with normalization methods, clustering algorithms, and a number of states in HMM. Several experimental tests have been performed including superimposing random noise, adding systematic error, and untrained transients. The proposed real-time transient identification system has many advantages, however, there are still a lot of problems that should be solved to apply to a real dynamic process. Further efforts are being made to improve the system performance and robustness to demonstrate reliability and accuracy to the required level.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2005년도 춘계 학술발표회 논문집
• /
• pp.367-374
• /
• 2005

The use of system identification approaches for damage detection has been expanded in recent years. Soft computing techniques such as neural networks have been utilized increasingly. Damage assessment using neural networks is presented in this study. Data set for training neural networks are acceleration response of simple beam under the various damage states ,which are the inputs. The outputs are the damage locations and extents. Not only the trained damages but also untrained damages are. detected accuratelyintheassessmentstage.

• 한국식생활문화학회지
• /
• 제19권3호
• /
• pp.267-272
• /
• 2004

This study was conducted to develop functional muffin in which grape seed extract(GSE) was added at 0, 0.5, 1 or 2% levels and to evaluate the quality characteristics. The volume of muffin decreased with the increasing GSE content, but no significant differences were observed. The weight and height of muffins increased with the increasing GSE content. The incorporation of GSE in the product lowered lightness (L value) and increased redness(a value). The mechanical texture parameters including hardness, springiness, gumminess and brittleness increased with the increasing GSE levels. Sensory analysis showed that muffin prepared with 0.5% GSE was considered to be as acceptable as control, based on hedonic scale ratings given by untrained taste panel.

• 한국공작기계학회:학술대회논문집
• /
• 한국공작기계학회 2004년도 춘계학술대회 논문집
• /
• pp.241-246
• /
• 2004

This paper proposed trajectory tracking control of mobile robot. Trajectory tracking control scheme are real coding genetic-algorithm and back-propergation algorithm. Control scheme ability experience proposed simulation. Stable tracking control problem of mobile robots have been studied in recent years. These studios have guaranteed stability of controller, but the performance of transient state has not been guaranteed. In some situations, constant gain controller shows overshoots and oscillations. So we introduce better control scheme using Real coding Genetic Algorithm(RCGA) and neural network. Using RCGA, we can find proper gains in several situations and these gains are generalized by neural network. The generalization power of neural network will give proper gain in untrained situation. Performance of proposed controller will verify numerical simulations and the results show better performance than constant gain controller.

• 한국공작기계학회논문집
• /
• 제13권3호
• /
• pp.126-132
• /
• 2004

This paper proposed trajectory tracking control based on genetic algorithm. Trajectory tracking control scheme are real coding genetic algorithm(RCGA) and back-propagation algorithm(BPA). Control scheme ability experience proposed simulation. Stable tracking control problem of mobile robots have been studied in recent years. These studies have guaranteed stability of controller, but the performance of transient state has not been guaranteed. In some situations, constant gain controller shows overshoots and oscillations. So we introduce better control scheme using real coding genetic algorithm and neural network. Using RCGA, we can find proper gains in several situations and these gains are generalized by neural network. The generalization power of neural network will give proper gain in untrained situation. Performance of proposed controller will verity numerical simulations and the results show better performance than constant gain controller.

• 한국정밀공학회지
• /
• 제16권12호
• /
• pp.78-86
• /
• 1999

The improvement of the water level control is important since it will prevent the steam generator trip so that improve the reliability and credibility of operation system. In this paper, the closed loop system identification is performed which can be used for the system monitoring and prediction of the system response. The model also can be used for the prediction control. Irving model is used as a steam generator model. The plant is an open loop unstable and non-minimum phase system. Fuzzy controller stabilize the system and the stable controller stabilize the system and the stable closed loop system is identified using neural networks. The obtained neural network model is validated using the untrained input and output. The results of computer simulation show the obtained Neural Network model represents the closed loop system well.

• 한국감성과학회:학술대회논문집
• /
• 한국감성과학회 1999년도 춘계학술발표논문집 논문집
• /
• pp.3-6
• /
• 1999

The study was performed to evaluate the sense of equilibrium by means of rotary chair system in middle school girls with dance learning. Control of equilibrium, which is the most important function in dancing, is maintained by vestibular, visual, and proprioceptive inputs. Experimental groups were divided into trained group with dance (experimental group, 13~14 ys, n=15) and untrained group (control group, 13~14 ys, n=15). experimental group was trained by programmed dancing for 12 weeks, 45 min/day and 5 day/week. Gain of eye movement was measured for vestibulocular reflex (VOR), visual vestibular stimulation (VVOR), optokinetic stimulation(OKN), and visual fixation (VFX). The gain of eye movement at higher than 0.08 hz was significant decrease in experimental group, and VVOR showed the similar fashion to VOR. The gain in OKN and VFX was not significant difference between experimental and control groups. These results suggest that programmed training of dancing improves control of balance by adaptation of the vestibular function, and rotary chair system would be useful to evaluate the sense of equilibrium ability.