• 한국융합학회논문지
• /
• 제8권4호
• /
• pp.9-18
• /
• 2017

무인항공기는 2000년대 중반부터 탐색, 조사, 매핑, 수색 및 구조 등의 3D 작업에 적극적으로 사용되기 시작했다. 이러한 세계적인 추세에 따라, 무인항공기의 정밀한 제어는 엄청난 응용 산업들의 측면에 있어서 혁명을 가져올 것이다. 논문의 첫 번째 파트에서는 시스템식별 기법을 사용하여 간략화 된 무인항공기의 모델을 이전의 이산시간 선형모델과 비교분석 한다. 두 번째 파트에서는 동적 모델의 세 가지 변수가 배치추정기법과 RLS추정기법을 사용하여 추정된다. 쿼드로터 무인항공기 호버링 기동 시의 각가속도 데이터가 항상 수렴한다고 분석되었다. 또한 실험 및 MATLAB 시뮬레이션의 쿼드로터 무인항공기 비행 데이터에 의하면, 배치추정기법이 RLS추정기법보다 더 정확하다고 판명되었다.

• Journal of Astronomy and Space Sciences
• /
• 제26권1호
• /
• pp.75-88
• /
• 2009

이 연구의 목적은 선형화 과정이 필요 없는 Unscented 변환(Unscented Transformation)을 사용한 후처리 배치 알고리즘을 소개하고, 기존 최소자승법을 이용한 후처리 배치 필터와 반복 UKF 스무더(Iterative Unscented Kalman Filter Smoother)들과 비교하여 추정 방법 간의 성능비교와 장단점을 분석하는 것이다. 연구에 사용된 위성 궤도 결정시스템의 동역학 방정식은 지구의 비대칭 중력장의 영향, 대기항력, 태양복사압 및 달과 태양의 중력으로 구성되었다. 관측 데이터로는 지상국으로부터 측정한 위성의 거리, 방위각과 고도각이 사용되었다. 특히, 비선형성의 영향에 대한 추정 방법 간의 성능과 장단점의 비교를 위해 위성의 포기 궤도오차별, 관측데이터의 관측 잡음의 크기별 테스트를 수행하였다. 이 연구를 통해 소개된, 선형화 과정이 필요 없는 Unscented 변환 기반의 후처리 배치 필터는, 비선형성의 특징이 증대된 상황에서 기존의 후처리 배치 알고리즘들에 비해 초기 궤도오차별, 관측데이터 잡음의 크기별 테스트 시 평균적으로 각각 약 5%와 12%정도의 정밀도 향상결과를 보였다. 또한, 기존 최소자승법을 이용한 후처리 배치필터가 발산한 상황에서도, 수렴성을 확보하는 안정적인 결과를 얻을 수 있었다. 그러므로 Unscented 변환 기반의 후처리 배치필터가 인공위성 궤도 결정 시스템에 효율적으로 사용할 수 있음을 제시하였다.

• Biotechnology and Bioprocess Engineering:BBE
• /
• 제7권5호
• /
• pp.275-280
• /
• 2002

In order to control glucose concentration during fed-batch culture for antibiotic production, we applied so called “software sensor” which estimates unmeasured variable of interest from measured process variables using software. All data for analysis were collected from industrial scale cultures in a pharmaceutical company. First, we constructed an estimation model for glucose feed rate to keep glucose concentration at target value. In actual fed-batch culture, glucose concentration was kept at relatively high and measured once a day, and the glucose feed rate until the next measurement time was determined by an expert worker based on the actual consumption rate. Fuzzy neural network (FNN) was applied to construct the estimation model. From the simulation results using this model, the average error for glucose concentration was 0.88 g/L. The FNN model was also applied for a special culture to keep glucose concentration at low level. Selecting the optimal input variables, it was possible to simulate the culture with a low glucose concentration from the data sets of relatively high glucose concentration. Next, a simulation model to estimate time course of glucose concentration during one day was constructed using the on-line measurable process variables, since glucose concentration was only measured off-line once a day. Here, the recursive fuzzy neural network (RFNN) was applied for the simulation model. As the result of the simulation, average error of RFNN model was 0.91 g/L and this model was found to be useful to supervise the fed-batch culture.

• Journal of Astronomy and Space Sciences
• /
• 제36권4호
• /
• pp.249-264
• /
• 2019

In this study, the precise orbit determination (POD) software is developed for optical observation. To improve the performance of the estimation algorithm, a nonlinear batch filter, based on the unscented transform (UT) that overcomes the disadvantages of the least-squares (LS) batch filter, is utilized. The LS and UT batch filter algorithms are verified through numerical simulation analysis using artificial optical measurements. We use the real optical observation data of a low Earth orbit (LEO) satellite, Cryosat-2, observed from optical wide-field patrol network (OWL-Net), to verify the performance of the POD software developed. The effects of light travel time, annual aberration, and diurnal aberration are considered as error models to correct OWL-Net data. As a result of POD, measurement residual and estimated state vector of the LS batch filter converge to the local minimum when the initial orbit error is large or the initial covariance matrix is smaller than the initial error level. However, UT batch filter converges to the global minimum, irrespective of the initial orbit error and the initial covariance matrix.

• 상하수도학회지
• /
• 제18권2호
• /
• pp.113-120
• /
• 2004

The biodegradability of organics has become essential for the design and modeling of a biological nutrient removal process. Respirometry for the batch test just with wastewater has been conducted to estimate active biomass and readily biodegradable organics in municipal wastewater simultaneously. Municipal wastewater contains significant active biomass, which is estimated about 17% of COD. Compared to the batch test seeded with sludge, the batch test just with wastewater represents a little higher readily biodegradable organics. This might be due to the different environment of the logarithmic growth of active biomass. The nitrate uptake rate test has been also performed for the estimation of the readily biodegradable organics. The nitrate uptake rate test results in a little higher readily biodegradable organics compared to the batch test seeded with sludge and similar organics compared to the batch test just with wastewater. This might be caused by the different sludge of a sequencing batch reactor process. Taking the result of the previous research into account, the readily biodegradable, slowly biodegradable, active biomass, soluble inert, and particulate inert organics are estimated about 11%, 49%, 17%, 11%, and 12% of COD, respectively.

• Journal of Microbiology and Biotechnology
• /
• 제13권5호
• /
• pp.641-646
• /
• 2003

An accurate and efficient method for measuring the mass of hairy roots using conductometry is established. A conductivity equation expressed in terms of the concentration of the ion species in the medium is suggested. By using this equation, the effect of the individual ions on the total conductivity can be quantitatively analyzed. An equation for the in situ estimation of the cell growth coefficient for determining the mass of hairy roots is established based on measurements of the nitrogen concentration and conductivity during cultivation. The proposed equation does not require preliminary experiments to determine the cell growth coefficient. Instead, the physiological characteristics of the plant species are reflected by introducing the cellular nitrogen content. Since the cell growth coefficient is determined by measuring the major ionic nutrient concentrations, it is more effective to express the dynamics of an actual culture system. This improved method for determining the mass of hairy roots was successfully utilized in a fed-batch culture system.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1993년도 한국자동제어학술회의논문집(국내학술편); Seoul National University, Seoul; 20-22 Oct. 1993
• /
• pp.595-600
• /
• 1993

The heat of reaction has been estimated from heat balance relationships around the reactor. The heat balance equations were formulated with the assumptions that the reactor temperature is uniformly distributed and the jacket temperatures are axially distributed. We have obtained the temperature distribution of jacket contents by FDM. And then, we have rearranged the heat balance equations so that the heat of reaction can be estimated from the finite number of temperature measurements, i.e., temperatures of the reactor contents, at the jacket inlet and outlet, respectively. The proposed method for reaction heat estimation on were applied to industrial batch reactors ; one is ABS polymerization reactor and the other is SAN polymerization reactor. We have also examined the variation of overall heat transfer coefficients for the reactors during reaction.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제18권3호
• /
• pp.670-684
• /
• 2024

3D Cross-Modal Retrieval (3DCMR) is a task that retrieves 3D objects regardless of modalities, such as images, meshes, and point clouds. One of the most prominent methods used for 3DCMR is the Cross-Modal Center Loss Function (CLF) which applies the conventional center loss strategy for 3D cross-modal search and retrieval. Since CLF is based on center loss, the center features in CLF are also susceptible to subtle changes in hyperparameters and external inferences. For instance, performance degradation is observed when the batch size is too small. Furthermore, the Mean Squared Error (MSE) used in CLF is unable to adapt to changes in batch size and is vulnerable to data variations that occur during actual inference due to the use of simple Euclidean distance between multi-modal features. To address the problems that arise from small batch training, we propose a Noisy Center Loss (NCL) method to estimate the optimal center features. In addition, we apply the simple Siamese representation learning method (SimSiam) during optimal center feature estimation to compare projected features, making the proposed method robust to changes in batch size and variations in data. As a result, the proposed approach demonstrates improved performance in ModelNet40 dataset compared to the conventional methods.

• Biotechnology and Bioprocess Engineering:BBE
• /
• 제10권6호
• /
• pp.522-527
• /
• 2005

Various processes which produce L-lactic acid using ammonia-tolerant mutant strain, Rhizopus sp. MK-96-1196, in a 3L airlift bioreactor were evaluated. When the fed-batch culture was carried out by keeping the glucose concentration at 30g/L, more than 140 g/L of L-lactic acid was produced with a product yield of 83%. In the case of the batch culture with 200g/L of initial glucose concentration, 121g/L of L-lactic acid was obtained but the low product yield based on the amount of glucose consumed. In the case of a continuous culture, 1.5g/L/h of the volumetric productivity with a product yield of 71% was achieved at dilution rate of $0.024\;h^{-1}$. Basis on these results three processes were evaluated by simple variable cost estimation including carbon source, steam, and waste treatment costs. The total variable costs of the fed-batch and continuous cultures were 88% and 140%, respectively, compared to that of batch culture. The fed-batch culture with high L-lactic acid concentration and high product yield decreased variable costs, and was the best-suited for the industrial production of L-lactic acid.

• Natural Product Sciences
• /
• 제23권2호
• /
• pp.75-83
• /
• 2017

The optimization and microwave assisted extraction of stem bark of Terminalia arjuna, quantitative estimation of the marker compounds arjunic acid and arjunolic acid using HPTLC and the evaluation of free radical scavenging activity has been performed in this study. The central composite design was used for optimization and the values of parameters for optimized batch of microwave assisted extraction were 1000 W (Power), 3 minutes (Time) and 1/120 (Solid/solvent ratio). The solvent system to carry out the HPTLC was toluene: acetic acid: ethyl acetate (5: 5: 0.5) and quantitative estimation was done using standard equations obtained from the marker compounds. The in-vitro free radical scavenging activity was performed spectrophotometrically using ascorbic acid as standard. The value of estimated percentage yield of arjunic acid and arjunolic acid was 1.42% and 1.52% which upon experimentation was obtained as 1.38% and 1.51% respectively. The DPPH assay of the different batches of microwave assisted extraction and marker compounds taken suggested that the marker compounds arjunic acid and the arjunolic acid were responsible for the free radical scavenging activity as the batch having the maximum percentage yield of the marker compounds showed best free radical scavenging effect as compared to standard ascorbic acid. The $IC_{50}$ value of the optimized batch was found to be 24.72 while that of the standard ascorbic acid was 29.83. Hence, the yield of arjunic acid and arjunolic acid has direct correlation with the free radical scavenging activity of stem bark extract of Terminalia arjuna and have potential to serve as active lead compounds for free radical scavenging activity.