• 한국농림기상학회지
• /
• 제20권2호
• /
• pp.205-213
• /
• 2018

작물 모델의 개발과 검증에 사용되는 생체중 자료는 파괴적 샘플링을 통해 얻어져 왔다. 파괴적 샘플링이 가지는 단점을 보완하기 위해 저가형 3D 센서인 Kinect 센서와 무료 공개 소프트웨어들을 사용하여 생체중을 추정하는 기법을 개발하였다. 특히, 많은 작물모델들이 개발되어 있지 않은 배추를 대상으로 입체이미지를 생성하여, 그로부터 얻어진 부피와 생체중 추정치의 신뢰도를 분석하고자 하였다. 크기가 다른 배추 결구 부위를 스캔하기 위해 Kinect 센서와, Microsoft가 무상으로 제공하는 Software Development Kit 내 Kinect Fusion Explorer 프로그램을 사용하였다. 개별 배추의 입체이미지를 생성하기 위해 3D 그래픽 편집 소프트웨어인 Meshlab을 활용하여 배경과 불필요한 물체를 수동으로 제거하였다. 또한, 불완전한 입체모델로부터 생체중 추정을 위해 3D 프린터 소프트웨어인 Makerbot Desktop 을 사용하여 배추를 생성하기 위해 필요한 플라스틱 필라멘트 소모량을 추정하였다. 입체모델 편집 프로그램인 Blender를 사용하여 부피를 추정하였을 때, 실제 부피에 비해 17.6%에서 2160.6% 범위의 상당한 오차가 있었다. 반면, 필라멘트 소요량은 실제 배추 생체중 변이의 98.7%를 설명할 수 있었다. 또한, 이들의 상관관계는 5% 수준에서 유의하였다. 이러한 결과들은 직접적인 부피 계산 절차를 제외하더라도 간편하게 생체중을 추정할 수 있음을 확인하였다. Kinect 센서를 사용하여 배추의 생체중 추정이 가능하다는 것이 확인 되었으나, 기존의 고가형 3D 센서에 비해 낮은 해상도와 주간에 활용이 어려운 점이 있다. 그럼에도 불구하고, 배추 생육 모델의 시계열적 검증 자료를 Kinect 센서를 이용하여 간편하고 신속하게 획득할 수 있어 모델의 불확도를 감소하는 데에 기여할 수 있을 것으로 판단된다. 따라서, 후속 연구에서 보다 저렴한 가격대의 3D 센서들을 대상으로 야외 및 주간조건애서 작물의 생체중 측정 가능성에 대해 검토하고 작물 모형 개발 및 개선을 위한 기술개발이 이루어져야 할 것으로 사료된다.

• 한국콘텐츠학회논문지
• /
• 제10권12호
• /
• pp.34-42
• /
• 2010

이동통신 환경에서 매개변수와 비매개변수추정 알고리즘을 토대로 DOA와 적응 빔성형 알고리즘을 연구하고 분석하였다. 매개변수추정 알고리즘에서는 배열안테나로 수신된 신호의 상관행렬로부터 신호성분과 잡음성분에 대한 고유치를 구하고, 그 고유치들로부터 전력스펙트럼을 판별하였다. 반면에 비매개변수추정 알고리즘에서는 입사각의 함수로써 신호에너지를 추정하기 위하여 고분해능과 잡음억압을 도출하도록 비이차놈을 사용하여 규칙화 목적함수를 최소화하였다. 그리고 나서, 신호와 잡음공간 조정벡터로부터 DOA를 추정하였고, 공간벡터에 의하여 도출된 가중치를 적용하여 적응 빔성형 패턴을 개선하였다. 따라서 희소강제조정을 갖는 개선된 방향성 추정 알고리즘은 다른 알고리즘들과 비교하여 잡음 억제와 고분해능을 갖는다.

• 응용통계연구
• /
• 제36권4호
• /
• pp.323-335
• /
• 2023

고차원 데이터의 인과 추론에서 고차원 공변량의 차원을 축소하고 적절히 변형하여 처리와 잠재 결과에 영향을 줄 수 있는 교란을 통제하는 것은 중요한 문제이다. 평균 처리 효과(average treatment effect; ATE) 추정에 있어서, 성향점수와 결과 모형 추정을 이용한 확장된 역확률 가중치 방법이 주로 사용된다. 고차원 데이터의 분석시 모든 공변량을 포함한 모수 모형을 이용하여 성향 점수와 결과 모형 추정을 할 경우, ATE 추정량이 일치성을 갖지 않거나 추정량의 분산이 큰 값을 가질 수 있다. 이런 이유로 고차원 데이터에 대한 적절한 차원 축소 방법과 준모수 모형을 이용한 ATE 방법이 주목 받고 있다. 이와 관련된 연구로는 차원 축소부분에 준모수 모형과 희소 충분 차원 축소 방법을 활용한 연구가 있다. 최근에는 성향점수와 결과 모형을 추정하지 않고, 차원 축소 후 매칭을 활용한 ATE 추정 방법도 제시되었다. 고차원 데이터의 ATE 추정 방법연구 중 최근에 제시된 네 가지 연구에 대해 소개하고, 추정치 해석시 유의할 점에 대하여 논하기로 한다.

• Journal of Advanced Marine Engineering and Technology
• /
• 제38권2호
• /
• pp.188-194
• /
• 2014

Recently, indoor localization systems based on wireless sensor networks have received a great deal of attention because they help achieve high accuracy in position determination by using various algorithms. In order to minimize the error in the estimated azimuth that can occur owing to sensor drift and recursive calculation in these algorithms, we propose a novel relative azimuth estimation algorithm. The advantages of the proposed technique in an indoor environment are that an improved weight average filter is used to effectively reduce impulse noise from the raw data acquired from nodes with inherent errors and a rotational displacement algorithm is applied to obtain a precise relative azimuth without using additional sensors, which can be affected by electromagnetic noise. Results from simulations show that the proposed filter reduces the impulse noise, and the acquired estimation error does not accumulate with time by using proposed algorithm.

• 한국지반공학회:학술대회논문집
• /
• 한국지반공학회 2009년도 세계 도시지반공학 심포지엄
• /
• pp.1065-1078
• /
• 2009

The shear strength has been managed as an important factor in soil mechanics. The shear strength estimation model was developed to evaluate the shear strength using only a few soil properties by the linear regression analysis model which is one of the statistical methods. The shear strength is divided into two part; one is the internal friction angle ($\Phi$) and the other is the cohesion (c). Therefore, some valid soil factors among the results of soil tests are selected through the correlation analysis using SPSS and then the model are formulated by the linear regression analysis based on the relationship between factors. Also, the developed model is compared with the result of direct shear test to prove the rationality of model. As the results of analysis about relationship between soil properties and shear strength, the internal friction angle is highly influenced by the void ratio and the dry unit weight and the cohesion is mainly influenced by the void ratio, the dry unit weight and the plastic index. Meanwhile, the shear strength estimated by the developed model is similar with that of the direct shear test. Therefore, the developed model may be used to estimate the shear strength of soils in the same condition of study area.

• 한국윤활학회:학술대회논문집
• /
• 한국윤활학회 1990년도 제12회 학술강연회초록집
• /
• pp.58-62
• /
• 1990

The size and concentration of wear debris in lubricating oil often reveal the operating condition of the rotating machinery. To evaluate the possible application of light attenuation measurement for the estimation of wear debris concentration in the lubricating oil, the light transmittance through the lubricating oil cell contaminated with various concentrations of diatomite particles was measured, the attenuation coefficient was estimated from the transmittance measurement and the coefficients were compared with those obtained from the scattering theory. The comparision showed good agreements between them. It is also noted that the experimentally determined attenuation coefficient showed almost linear relation with particle weight concentrations for the concentrations within the range of 2000 ppm. For the case of 0 ppm weight concentration of diatomite particles in the lubricating oil cell, the thickness of the cell required to give $100 \mu W$ light attenuation is 7.75 mm. This result indicates that the light attenuation method will be one of the possible candidates of machine failure diagnostic sensors for the estimation of wear debris concentration in the lubricating oil.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2003년도 학술회의 논문집 정보 및 제어부문 A
• /
• pp.187-190
• /
• 2003

In this paper, we address the robust adaptive backstepping controller using fuzzy neural network (FHIN) for a class of uncertain output feedback nonlinear systems with disturbance. A new algorithm is proposed for estimation of unknown bounds and adaptive control of the uncertain nonlinear systems. The state estimation is solved using K-fillers. All unknown nonlinear functions are approximated by FNN. The FNN weight adaptation rule is derived from Lyapunov stability analysis and guarantees that the adapted weight error and tracking error are bounded. The compensated controller is designed to compensate the FNN approximation error and external disturbance. Finally, simulation results show that the proposed controller can achieve favorable tracking performance and robustness with regard to unknown function and external disturbance.

• Tribology and Lubricants
• /
• 제6권2호
• /
• pp.94-98
• /
• 1990

The size and concentration of wear debris in lubricating oil often reveal the operating condition of the rotating machinery. To evaluate the possible application of light attenuation measurement for the estimation of wear debris concentration in the lubricating oil, the light transmittance through the lubricating oil cell contaminated with various concentrations of diatomire particles was measured, the attenuation coefficient was estimated from the transmittance measurement and the coefficients were compared with those obtained from the scattering theory. The comparision showed good agreements between them. It is also noted that the experimentally determined attenuation coefficient showed almost linear relation with particle weight coucentrations for the concentrations within the range of 2000 ppm. For the case of 0 ppm weight concentration of diatomire particles in the lubricating oil cell, the thickness of the cell required to give $100 \muW$ light attenuation is 7.75 mm. This result indicates that the light attenuation method will be one of the possible candidates of machine failure diagnostic sensors for the estimation of wear debris concentration in the lubricating oil.

• ETRI Journal
• /
• 제38권3호
• /
• pp.487-493
• /
• 2016

Two new methods are proposed for an unsupervised adaptation of a language model (LM) with a single sentence for automatic transcription tasks. At the training phase, training documents are clustered by a method known as Latent Dirichlet allocation (LDA), and then a domain-specific LM is trained for each cluster. At the test phase, an adapted LM is presented as a linear mixture of the now trained domain-specific LMs. Unlike previous adaptation methods, the proposed methods fully utilize a trained LDA model for the estimation of weight values, which are then to be assigned to the now trained domain-specific LMs; therefore, the clustering and weight-estimation algorithms of the trained LDA model are reliable. For the continuous speech recognition benchmark tests, the proposed methods outperform other unsupervised LM adaptation methods based on latent semantic analysis, non-negative matrix factorization, and LDA with n-gram counting.

• Journal of the Korean Data and Information Science Society
• /
• 제17권3호
• /
• pp.775-784
• /
• 2006

This study is to investigate the effect of the number of co-rated users to the MAE. User based collaborative algorithm generally uses similarity weight to compute the relation of active user and other users. The original estimation algorithm of the GroupLens used the Pearson's correlation coefficient, soon after other researchers used various weighting. The Pearson’s correlation coefficient and Vector similarity, which is used in the field of information retrieval, are commonly used to the estimation algorithm. In prediction, we analyze the effect of the number of co-rated users on the user based recommender system.