• Transactions of the Korean Society of Automotive Engineers
• /
• v.20 no.4
• /
• pp.103-108
• /
• 2012

Fuel economy is an important factor in a vehicle owing to recent energy supply and environmental problems. This paper deals with fuel cell hybrid vehicles (FCHVs) and introduces a fuel economy evaluation method. The fuel economy of an FCHV depends on its power management strategy. Two rule-based power management strategies are applied to this paper and their fuel economy is evaluated based on the optimal control theory. The concept of the optimal line is also applied to this paper, which is used to compare the fuel consumption of a power management strategy to the optimal result. The two rule-based strategies are also compared to each other.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.9 no.3
• /
• pp.676-683
• /
• 2005

In this paper, we proposed a method that optimal values of fuzzy control rule base and quantization of membership function are searched by genetic algorithm. Proposed method searched the optimal values of membership function and control rules using genetic algorithm by off-line. Then fuzzy controller operates using these values by on-line. Proposed fuzzy control system is optimized the control rule base and membership function by genetic algorithm without expert's knowledge. We investigated proposed method through simulation and experiment using DC motor and one link manipulator, and confirmed the following usefulness.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2010.04a
• /
• pp.969-972
• /
• 2010

소프트웨어 제품의 품질을 보장하기 위해서는 제품을 개발하는 단계에 미리 결함율을 예측하여 원하는 수준의 품질을 확보하는 것이 중요하다. 결함은 사용자의 요구사항이 제품으로 구현되고 기능에 대한 테스트가 수행되는 단계에 가장 객관적이며 정량적으로 관리될 수 있다. 따라서 본 논문에서는 통합테스트에 대한 계획을 수립하는 단계에 제품에 대한 결함율을 미리 예측하여 제품 결함율이 조직의 관리범위에 들어올 수 있도록 통제하는 결함예측모델을 제안한다. 조직의 제품 결함율 베이스라인을 설정하고 통합테스트 결함율에 영향을 미치는 변수들과의 회귀분석을 통해 통합테스트 결함예측모형을 구축한다. 또한 제품 결함율에 영향을 미치는 변수들과의 회귀분석을 통해 제품 결함예측모형을 구축하고 결함예측모형을 활용해 제품 결함율을 분석 및 통제한다. 본 논문에서 제안한 결함예측모델은 실제 프로젝트에 적용하여 실효성을 검증하였으며 제품이 완성되기 전에 결함율을 예측하여 통제할 수 있게 함으로써 소프트웨어 품질을 향상한다.

• Annual Conference on Human and Language Technology
• /
• 2022.10a
• /
• pp.205-209
• /
• 2022

페르소나 대화 시스템이 상대방의 개인화된 정보에 일관된 응답을 생성하는 것은 상당히 중요하며, 이를 해결하기 위해 최근에 많은 연구들이 활발히 이루어지고 있다. 그 중, PersonaChat 데이터셋에 대해 수반/중립/모순 관계를 라벨링한 DialoguNLI 데이터셋이 제안되었으며, 일관성 측정, 페르소나 속성 추론 태스크 등 여러 분야에 활용되고 있다. 그러나, 공개적으로 이용가능한 한국어로 된 대화 추론 데이터셋은 없다. 본 연구에서는 한국어로 번역된 페르소나 대화 데이터셋과 한국어 자연어 추론 데이터셋에 학습된 모델을 이용하여 한국어 대화 추론 데이터셋(KorDialogueNLI)를 구축한다. 또한, 사전학습된 언어모델을 학습하여 한국어 대화 추론 모델 베이스라인도 구축한다. 실험을 통해 정확도 및 F1 점수 평가 지표에서 KLUE-RoBERTa 모델을 미세조정(fine-tuning)시킨 모델이 가장 높은 성능을 달성하였다. 코드 및 데이터셋은 https://github.com/passing2961/KorDialogueNLI에 공개한다.

• Annual Conference on Human and Language Technology
• /
• 2022.10a
• /
• pp.473-476
• /
• 2022

문장 임베딩은 문장의 의미를 고려하여 모델이 적절하게 의미적인 벡터 공간에 표상하는 것이다. 문장 임베딩을 위해 다양한 방법들이 제안되었지만, 최근 가장 높은 성능을 보이는 방법은 대조적 학습 방법이다. 대조적 학습을 이용한 문장 임베딩은 문장의 의미가 의미적으로 유사하면 가까운 공간에 배치하고, 그렇지 않으면 멀게 배치하도록 학습하는 방법이다. 이러한 대조적 학습은 비지도와 지도 학습 방법이 존재하는데, 본 논문에서는 효과적인 비지도 학습방법을 제안한다. 기존의 비지도 학습 방법은 문장 표현을 학습하는 언어모델이 자체적인 정보를 활용하여 문장의 의미를 구별한다. 그러나, 하나의 모델이 판단하는 정보로만 문장 표현을 학습하는 것은 편향적으로 학습될 수 있기 때문에 한계가 존재한다. 따라서 본 논문에서는 Cross-Encoder의 Re-Ranker를 통한 의미 검색으로부터 추천된 문장 쌍을 학습하여 기존 모델의 성능을 개선한다. 결과적으로, STS 테스크에서 베이스라인보다 2% 정도 더 높은 성능을 보여준다.

• KIPS Transactions on Software and Data Engineering
• /
• v.2 no.12
• /
• pp.873-880
• /
• 2013

This paper approaches the problem of resolving Korean zero pronouns using Centering Theory modeling local coherence. Centering Theory has been widely used to resolve English pronouns. However, it is much difficult to apply the centering framework for zero pronoun resolution in languages such as Japanese and Korean. Since in particular the use of non-anaphoric zero pronouns without explicit antecedents is not considered in the Centering Theory of Grosz et al., the presence of non-anaphoric cases negatively affects the performance of the resolution system based on Centering Theory. To overcome this, this paper presents a method which determines the intra-sentential anaphoricity of zero pronouns in subject position by using relationships between clauses, and then identifies antecedents of zero subjects. In our experiments, the proposed method outperforms the baseline method relying solely on Centering Theory.

• Journal of the Korean Society for information Management
• /
• v.29 no.4
• /
• pp.43-60
• /
• 2012

Identifying citing sentences from article full-text is a prerequisite for creating a variety of future academic information services such as citation-based automatic summarization, automatic generation of review articles, sentiment analysis of citing statements, information retrieval based on citation contexts, etc. However, finding citing sentences is not easy due to the existence of implicit citing sentences which do not have explicit citation markers. While several methods have been proposed to attack this problem for English, it is difficult to find such automatic methods for Korean academic literature. This article presents a rule-based approach to identifying Korean citing sentences. Experiments show that the proposed method could find 30% of implicit citing sentences in our test data in nearly 70% precision.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.22 no.2
• /
• pp.261-269
• /
• 2018

This paper presents a convenient method to estimate the internal parameters of plenoptic camera using CCD(charge-coupled device) camera model. The images used for plenoptic camera calibration generally use the checkerboard pattern used in CCD camera calibration. Based on the CCD camera model, the determinant of the plenoptic camera model can be derived through the relationship with the plenoptic camera model. We formulate four equations that express the focal length, the principal point, the baseline, and distance between the virtual camera and the object. By performing a nonlinear optimization technique, we solve the equations to estimate the parameters. We compare the estimation results with the actual parameters and evaluate the reprojection error. Experimental results show that the MSE(mean square error) is 0.309 and estimation values are very close to actual values.

• Journal of Digital Contents Society
• /
• v.17 no.6
• /
• pp.593-601
• /
• 2016

This paper is related to the Region-based Spectral Correlation Estimation(RSCE) coding method that makes it possible to achieve the high-compression ratio by estimating color component images from luminance image. The proposed method is composed of three steps. First, Y/C bit-plane summation image is defined using normalized chrominance summation image and luminance image, and then the Y/C bit-plane summation image is segmented for extracting the shape information of the regions. Secondly, the scale factor and the offset factor minimizing the approximation square errors between luminance image and R, B images by the each region are calculated. Finally, the scale factor and the offset factor for the each region are encoded into bit stream. Referring to the results of computer simulation, the proposed method provides more than two or three times higher compression ratio than JPEG/Baseline or JPEG2000/EBCOT algorithm in terms of bpp needed for encoding two color component images with the same PSNR.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.4 no.3
• /
• pp.705-711
• /
• 2000

A thin film oxide semiconductor micro gas sensor array which shows only 60 mW of power consumption at an operating temperature of $300^{\circ}C$ has been fabricated using microfabrication and micromachining techniques. Excellent thermal insulation of the membrane is achieved by the use of a double-layer structure of $0.1\mum\; thick\; Si_3N_4 \;and\; 1 \mum$ thick phosphosilicate glass (PSG) prepared by low-pressure chemical-vapor deposition (LPCVD) and atmospheric-pressure chemical-vapor deposition (APCVD), respectively. The sensor array consists of such thin film oxide semiconductor sensing materials as 1 wt.% Pd-doped $SnO_2,\; 6 wt.% A1_2O_3-doped\; ZnO,\; WO_3$/ and ZnO. Baseline resistances of the four sensing materials were found to be stable after the aging for three days at $300^{\circ}C$. The thin film oxide semiconductor micro gas sensor array exhibited resistance changes usable for subsequent data processing upon exposure to various gases and the sensitivity strongly depended on the sensing layer materials.