• Environmental Engineering Research
• /
• 제24권3호
• /
• pp.404-411
• /
• 2019

In this study, we designed a data-driven model to predict chlorophyll-a using M5P model tree and extreme learning machine (ELM). The Juksan weir in the Youngsan River has high chlorophyll-a, which is the primary indicator of algal bloom every year. Short-term algal bloom prediction is important for environmental management and ecological assessment. Two models were developed and evaluated for short-term algal bloom prediction. M5P is a classification and regression-analysis-based method, and ELM is a feed-forward neural network with fast learning using the least square estimate for regression. The dataset used in this study includes water temperature, rainfall, solar radiation, total nitrogen, total phosphorus, N/P ratio, and chlorophyll-a, which were collected on a daily basis from January 2013 to December 2016. The M5P model showed that the prediction model after one day had the highest performance power and dropped off rapidly starting with predictions after three days. Comparing the performance power of the ELM model with the M5P model, it was found that the performance power of the 1-7 d chlorophyll-a prediction model was higher. Moreover, in a period of rapidly increasing algal blooms, the ELM model showed higher accuracy than the M5P model.

• 한국정보통신학회논문지
• /
• 제20권12호
• /
• pp.2348-2354
• /
• 2016

최근 농업분야에서 IoT(Internet of Things)기술을 통해 다양한 생체 및 환경 정보를 DB(data base)로 구축할 수 있게 되면서 빅 데이터를 이용한 기계학습 분석이 증가하고 있다. 기계학습 분석을 통해 농업의 생산량과 가축의 질병 등을 예측할 수 있게 되어 농업경영에서 효율적인 의사결정을 돕는다. 본 논문에서는 스마트 돈사의 다양한 환경데이터와 몸무게데이터를 이용하여 환경정보와 일당증체의 연관성 모델을 도출하고 그 정확도를 분석하였다. 이를 위해 기계학습의 M5P tree기법을 적용하였다. 분석을 통해 일당증체량이 풍속에 큰 영향을 받는 것을 확인하였다.

• Advances in concrete construction
• /
• 제14권1호
• /
• pp.15-34
• /
• 2022

The use of waste rubber in concrete can reduce natural aggregate consumption and improve some technical properties of concrete. Although there are several equations for estimating the mechanical properties of concrete containing waste rubber, limited numbers of machine learning-based models have been proposed to predict the mechanical properties of rubbercrete. In this study, an extensive database of the mechanical properties of rubbercrete was gathered from a comprehensive survey of the literature. To model the mechanical properties of rubbercrete, M5P tree and linear gene expression programming (LGEP) methods as two machine learning techniques were employed to achieve reliable mathematical equations. Two procedures of input variable selection were considered in this study. The crucial component ratios of rubbercrete and concrete age were assumed as the input variables in the first procedure. In contrast, the volumes of the coarse and fine waste rubber and the compressive strength of concrete without waste rubber were considered the second procedure of the input variables. The results show that the models obtained by LGEP are more accurate than those achieved by the M5P model tree and existing traditional equations. Besides, the volumes of the coarse and fine waste rubber and the compressive strength of concrete without waste rubber are better predictors of the mechanical properties of rubbercrete compared to the first procedure of input variable selection.

• Computers and Concrete
• /
• 제24권2호
• /
• pp.137-150
• /
• 2019

In this paper, surrogate models such as multivariate adaptive regression splines (MARS) and M5P model tree (M5P MT) methods have been investigated in order to propose a new formulation for the 28-days compressive strength of self-compacting concrete (SCC) incorporating metakaolin as a supplementary cementitious materials. A database comprising experimental data has been assembled from several published papers in the literature and the data have been used for training and testing. In particular, the data are arranged in a format of seven input parameters covering contents of cement, coarse aggregate to fine aggregate ratio, water, metakaolin, super plasticizer, largest maximum size and binder as well as one output parameter, which is the 28-days compressive strength. The efficiency of the proposed techniques has been demonstrated by means of certain statistical criteria. The findings have been compared to experimental results and their comparisons shows that the MARS and M5P MT approaches predict the compressive strength of SCC incorporating metakaolin with great precision. The performed sensitivity analysis to assign effective parameters on 28-days compressive strength indicates that cementitious binder content is the most effective variable in the mixture.

• Journal of Forest and Environmental Science
• /
• 제33권3호
• /
• pp.187-196
• /
• 2017

The use of unmanned aerial vehicles (UAV) for the estimation of tree height is gaining recognition. This study aims to assess the effectiveness of tree height estimation of Pinus densiflora Sieb. et Zucc. and Pinus koraiensis Sieb. et Zucc. using digital surface model (DSM) generated from UAV-acquired imageries. Images were taken with the $Trimble^{(R)}$ UX5 equipped with Sony ${\alpha}5100$. The generated DSM, together with the digital elevation model (DEM) generated from a digital map of the study areas, were used in the estimation of tree height. Field measurements were conducted in order to generate a regression model and carry out accuracy assessment. The obtained coefficients of determination (R2) and root mean square error (RMSE) for P. densiflora (R2=0.71; RMSE=1.00 m) and P. koraiensis (R2=0.64; RMSE=0.85 m) are comparable to the results of similar studies. The results of the paired two-tailed t-test show that the two tree height estimation methods are not significantly different (p-value=0.04 and 0.10, alpha level=0.01), which means that tree height estimation using UAV imagery could be used as an alternative to field measurement.

• Environmental Engineering Research
• /
• 제21권2호
• /
• pp.152-163
• /
• 2016

Present study investigates the potential of cassava peel and rubber tree bark for the removal of Cr (VI) from aqueous solution. Removal efficiency of more than 99% was obtained during the kinetic adsorption experiments with dosage of 3.5 g/L for cassava peel and 8 g/L for rubber tree bark. By comparing popular isotherm models and kinetic models for evaluating the kinetics of mass transfer, it was observed that Redlich-Peterson model and Langmuir model fitted well ($R^2$ > 0.99) resulting in maximum adsorption capacity as 79.37 mg/g and 43.86 mg/g for cassava peel and rubber tree bark respectively. Validation of pseudo-second order model and Elovich model indicated the possibility of chemisorption being the rate limiting step. The multi-linearity in the diffusion model was further addressed using multi-sites models (two-site series interface (TSSI) and two-site parallel interface (TSPI) models). Considering the influence of interface properties on the kinetic nature of sorption, TSSI model resulted in low mass transfer rate (5% for cassava peel and 10% for rubber tree bark) compared to TSPI model. The study highlights the employability of two-site sorption model for simultaneous representation of different stages of kinetic sorption for finding the rate-limiting process, compared to the separate equilibrium and kinetic modeling attempts.

• 한국환경과학회지
• /
• 제20권7호
• /
• pp.881-890
• /
• 2011

In this study, thermal environment changes for a marathon course of IAAF World Championship, Daegu 2011 were modeled to provide improvements of thermal environment, so that runners could have the maximum condition and citizens pleasant streets. The three biggest size of intersections were selected for the study. Envi-met, 3G microclimate model, were used for a thermal environment analysis and three different cases - present status, planting roadside tree scenario, and roof-garden scenario - were compared. The followings are the results of the study. 1. The highest thermal distribution were shown at 1 p.m., but there was no significant difference between a thermal distribution at 1 p.m. and that at 5 p.m. since a heat flux from buildings affects thermal distributions rather than insolation does. 2. Tree planting or adding environmental friendly factors might lead a temperature drop effect, but the effect was not significant for areas covered with impermeability packing materials such as concrete or asphalt (especally, for Site case 2) 3. The combination of tree planting and adding environmental friendly factors also brought a temperature drop effect (Site 1 and 2) and this case showed even better result if green spaces (especially, parks) were closed.

• 한국지능시스템학회논문지
• /
• 제16권5호
• /
• pp.608-615
• /
• 2006

본 논문에서는 퍼지 클러스터 기법을 이용하여 구간 분할된 퍼지 모델트리의 제안과 이를 이용한 데이터 모델링 기법을 다룬다. 제안된 방법은 먼저 입력과 출력변수의 속성을 고려한 퍼지 클러스터링에 의해 중심벡터를 계산한 후, 중심벡터들과 입력속성간의 소속도를 이용하여 구간 분할된 영역별로 각각의 선형모델을 구축한다. 노드의 확장은 부모노드(parent node)에서 만들어진 모델에서 계산된 오차값과 자식노드(child node)에서 계산된 오차값을 비교하여 이루어진다. 출력값 예측 단계에서는 입력된 데이터와 잎노드에서 계산된 클러스터 중심값과 비교하여 소속도가 높은 선형모델을 선택하여 데이터에 대한 출력값을 예측하게 된다. 제안된 방법의 우수성을 보이기 위해 다양한 데이터를 대상으로 실험한 결과, 기존의 모델트리방식 및 뉴럴 네트워크 기반의 신경회로망 보다 향상된 성능을 보임을 알 수 있었다.

• 한국기후변화학회지
• /
• 제8권4호
• /
• pp.415-424
• /
• 2017

This study was conducted to estimate carbon stocks of Pinus densiflora with drawing volume of trees in each tree height and DBH applying the suitable stem taper equation and tree specific carbon emission factors, using collected growth data from all over the country. Information on distribution area, tree age, tree number per hectare, tree volume and volume stocks were obtained from the $5^{th}$ National Forest Inventory (2006~2010) and Statistical yearbook of forest (2016), and method provided in IPCC GPG was applied to estimate carbon stock and uptake. Performance in predicting stem diameter at a specific point along a stem in Pinus densiflora by applying Kozak's model, $d=a_{1}DBH^{a_2}a_3^{DBH}X^{b_{1}Z^2+b_2ln(Z+0.001)+b_3\sqrt{Z}+b_4e^z+b_5(\frac{DBH}{H})}$, which is well known equation in stem taper estimation, was evaluated with validations statistics, Fitness Index, Bias and Standard Error of Bias. Consequently, Kozak's model turned out to be suitable in all validations statistics. Stem volume table of P. densiflora was derived by applying Kozak's model and carbon stock tables in each tree height and DBH were developed with country-specific carbon emission factors ($WD=0.445t/m^3$, BEF = 1.445, R = 0.255) of P. densiflora. As the results of analysis in carbon uptake for each province, the values were high with Gangwon-do $9.4tCO_2/ha/yr$, Gyeongsandnam-do and Gyeonggi-do $8.7tCO_2/ha/yr$, Chungcheongnam-do $7.9tCO_2/ha/yr$ and Gyeongsangbuk-do $7.8tCO_2/ha/yr$ in order, and Jeju-do was the lowest with $6.8tC/ha/yr$. Total carbon stocks of P. densiflora were 127,677 thousands tC which is 25.5% compared with total percentage of forest and carbon stock per hectare (ha) was $84.5tC/ha/yr$ and $7.8tCO_2/ha/yr$, respectively.

• 한국농림기상학회지
• /
• 제17권4호
• /
• pp.348-357
• /
• 2015

본 연구는 그간 우리나라에서 경제적인 가치를 인정 받아온 수종인 잣나무를 대상으로 잣나무의 현존 분포를 파악하고, RCP (Representative Concentration Pathway) 8.5 기후변화 시나리오와 생태적 지위 모형에 기반하여 향후 잣나무의 분포 변화를 예측하기 위해 수행되었다. 이를 위해 5년간의 NFI 자료에서 조사지점별 잣나무의 풍부도 자료를 추출하여 사용하였으며, 수종에 영향을 미치는 환경요인변수를 선정하기 위해 생태적 지위 모형 가운데 하나인 GARP (Genetic Algorithm for Rule-set Production)를 이용하였다. 총 27개의 환경요인변수에 대해 각각 모형을 구동하고 컨퓨전 매트릭스(Confusion Matrix) 기반 산출 통계량인 AUC (Area Under Curve)가 0.6 이상인 변수들을 선발하여 최종 잠재분포모형을 작성하였다. 그 결과 작성된 모형은 비교적 높은 적합도를 나타냈는데 잣나무는 현재 표고의 범위가 300m에서 1,200m 사이인 지역 및 남부에서 북부에 이르기까지 넓게 자리 잡고 있는 것으로 나타났다. 작성된 모형에 RCP 8.5 기후변화 시나리오를 적용한 결과, 잣나무는 2020년대부터 잠재분포역이 큰 폭으로 축소되며, 2090년대에는 우리나라 대부분의 지역이 잣나무의 생육에 불리할 것으로 예측되었다. 본 연구를 통해 기후변화가 잣나무 분포에 미치는 영향을 파악하고, 잣나무와 기후변화와의 상관성에 대한 이해를 높임으로써 향후 지역별 조림수종 선정 및 경제수종 교체 등의 조림적 관점에서 도움이 될 수 있을 것으로 판단된다.