• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2005년도 학술발표회 논문집
• /
• pp.679-683
• /
• 2005

The technique which connects Generalized Regression Neural Networks Model(GRNNM) with Genetic Algorithm (CA) is used to derive rating curve in the river basin. GRNNM architecture consists of 4 layers ; input, hidden, summation and output layer. GA method is applied to estimate the optimal smoothing factor when GRNNM is trained. The derivation of rating curve using GRNNM is considered different kinds of hydraulic characteristics such as water stage, area and mean velocity and is applied two stage stations; Sunsan and Jungam. Furthermore, it is compared with conventional curve-fitting method. Through the training and validation performance, the results show that GRNNM is much superior as compared to the conventional curve-fitting method.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2007년도 학술발표회 논문집
• /
• pp.1427-1431
• /
• 2007

In this study, the new methodology which combines Kohonen self-organizing map(KSOM) neural networks model and the conventional neural networks models such as feedforward neural networks model and generalized neural networks model is introduced to forecast flood stage in Nakdong river, Republic of Korea. It is possible to train without output data in KSOM neural networks model. KSOM neural networks model is used to classify the input data before it combines with the conventional neural networks model. Four types of models such as SOM-FFNNM-BP, SOM-GRNNM-GA, FFNNM-BP, and GRNNM-GA are used to train and test performances respectively. From the statistical analysis for training and testing performances, SOM-GRNNM-GA shows the best results compared with the other models such as SOM-FFNNM-BP, FFNNM-BP, and GRNNM-GA and FFNNM-BP shows vice-versa. From this study, we can suggest the new methodology to forecast flood stage and construct flood warning system in river basin.

• 한국수자원학회논문집
• /
• 제40권1호
• /
• pp.89-99
• /
• 2007

본 논문에서는 본 연구논제(2007)에서 개발된 COMBINE-GRNNM-GA(Type-1)으로부터 최적형태의 구조를 가진 모형을 구성하고, 입력층노드의 기상인자를 제거하기 위하여 불확실성 분석을 실시하였다. 훈련과정중에 가장 최소의 평활인자를 가진 입력층변수는 COMBINE-GRNNM-GA(Type-1)에서 제거되었으며, 변형된 COMBINE-GRNNM-GA(Type-1)은 기상학적 변수의 새로운 최소 평활인자를 구하기 위하여 재훈련된다. 최소 평활인자를 가지는 입력층 노드는 모형결과치에 대하여 가장 유용하지 않는 기상인자인 것을 암시하고 있다. 게다가, 민감하거나 민감하지 않은 기상인자들이 불확실성 분석을 통하여 선택되어진다. 최적 COMBINE-GRNNM-GA(Type-1)은 최소 비용과 노력으로 결측 혹은 미계측 증발접시 증발량과 계측되고 있지 않은 알팔파 기준증발산량을 산정하기 위하여 개발되었다 마지막으로 치적 COMBINE-GRNNM-GA(TyPe-1)을 이용하여 우리나라에서 전반적인 가뭄해석 및 관개배수 시스템 구축을 위한 참고자료를 제공할 수 있는 증발접시 증발량 지도 및 알팔파 기준증발산량 지도도 구축되어질 수 있다.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2006년도 학술발표회 논문집
• /
• pp.115-119
• /
• 2006

The goal of this research is to develop and apply the generalized regression neural networks model (GRNNM) embedding genetic algorithm (GA) for pan evaporation, which is missed or ungaged and for the alfalfa reference evapotranspiration, which is not measured in South Korea. The GRNNM-GA is evaluated using the training, the testing, and reproduction performance respectively for the estimation of the PE and the alfalfa reference evapotranspiration. Since the observed data of the alfalfa reference evapotranspiration using lysimeter have not been measured for a long time in South Korea, the PM method is used to assume and estimate the observed alfalfa reference evapotranspiration. From this research, we evaluate the impact of the limited climatical variables on the accuracy of the GRNNM-GA. We should, furthermore, construct the credible data of the PE and the alfalfa reference evapotranspiration and suggest the reference data for irrigation and drainage networks system in South Korea.

• 한국방재학회:학술대회논문집
• /
• 한국방재학회 2008년도 정기총회 및 학술발표대회
• /
• pp.669-673
• /
• 2008

In this study, the new methodology which combines Kohonen self-organizing map(KSOM) neural networks model and the conventional neural networks models such as feedforward neural networks model and generalized neural networks model is introduced to forecast flood stage in Nakdong river, Republic of Korea. It is possible to train without output data in KSOM neural networks model. KSOM neural networks model is used to classify the input data before it combines with the conventional neural networks model. Four types of models such as SOM-FFNNM-BP, SOM-GRNNM-GA, FFNNM-BP, and GRNNM-GA are used to train and test performances respectively. From the statistical analysis for training and testing performances, SOM-GRNNM-GA shows the best results compared with the other models such as SOM-FFNNM-BP, FFNNM-BP, and GRNNM-GA and FFNNM-BP shows vice-versa. From this study, we can suggest the new methodology to forecast flood stage and construct flood warning system in river basin.

• 한국수자원학회논문집
• /
• 제40권1호
• /
• pp.73-88
• /
• 2007

본 연구의 목적은 결측 혹은 미계측 증발접시 증발량과 우리나라에서 계측되고 있지 않은 알팔파 기준증발산량의 산정을 위하여 유전자 알고리즘이 내재된 일반화된 회귀신경망모형을 개발하고 적용하는데 있다. 우리나라에서는 장기간동안 증발산계를 이용하여 알팔파 기준증발산량의 관측이 시행되지 않고 있으므로, 본 연구에서는 Penman-Monteith(PM) 공식을 이용하여 산정된 값을 계측된 알팔파 기준증발산량으로 가정하였다. 본 연구를 통하여 최적 증발접시 증발량과 알팔파 기준증발산량의 산정을 위한 COMBINE-GRNNM-GA(Type-1) 모형을 개발하였으며, 훈련, 테스트 및 재현과정을 통하여 COMBINE-GRNNM-GA(Type-1) 모형을 평가하였다. COMBINE-GRNNM-GA (Type-1) 모형은 제시된 기상인자를 평가할 수 있으며, 증발접시 증발량과 알팔파 기준증발산량에 대한 신뢰성 있는 자료를 구축할 수 있다. 더 나아가서 우리나라에서 관개배수 시스템 구축을 위한 참고자료를 제공할 수 있을 것으로 판단된다.

• 대한토목학회논문집
• /
• 제30권4B호
• /
• pp.399-412
• /
• 2010

본 연구의 목적은 연 증발접시 증발량의 시간적인 분해를 위하여 신경망모형을 적용하는데 있다. 신경망모형은 각각 다층 퍼셉트론 신경망모형(MLP-NNM)과 일반화된 회귀신경망모형(GRNNM)으로 구성되어 있다. 그리고 신경망모형의 수행평가를 위하여 훈련 및 테스트과정으로 구성되었다. 신경망모형의 훈련과정을 위하여 실측, 모의 및 혼합자료와 같은 세 가지 형태의 자료가 사용되었으며, 테스트과정을 위해서는 실측자료만 이용되었다. 본 연구를 통하여 비선형 시계열자료의 시간적 분해를 위해서 MLP-NNM과 GRNNM의 적용성을 평가하였다. 게다가 연 증발접시 증발량 자료의 시간적 분해로부터 신뢰성있는 월 증발접시 증발량자료를 구축할 수 있을 것이며, 관개배수 네트워크 시스템의 평가를 위한 이용가능한 자료를 제공할 수 있을 것이다.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2009년도 학술발표회 초록집
• /
• pp.1207-1210
• /
• 2009

The goal of this research is to apply the neural networks models for the disaggregation of the pan evaporation (PE) data, Republic of Korea. The neural networks models consist of generalized regression neural networks model (GRNNM) and multilayer perceptron neural networks model (MLP-NNM), respectively. The disaggregation means that the yearly PE data divides into the monthly PE data. And, for the performances of the neural networks models, they are composed of training and test performances, respectively. The training and test performances consist of the only historic data, respectively. From this research, we evaluate the impact of GRNNM and MLP-NNM for the disaggregation of the nonlinear time series data. We should, furthermore, construct the credible data of the monthly PE data from the disaggregation of the yearly PE data, and can suggest the methodology for the irrigation and drainage networks system.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2009년도 학술발표회 초록집
• /
• pp.1215-1218
• /
• 2009

The goal of this research is to apply the neural networks models for the disaggregation of the pan evaporation (PE) data, Republic of Korea. The neural networks models consist of generalized regression neural networks model (GRNNM) and multilayer perceptron neural networks model (MLP-NNM), respectively. The disaggregation means that the yearly PE data divides into the monthly PE data. And, for the performances of the neural networks models, they are composed of training and test performances, respectively. The training data consist of the mixed data The mixed data involves the historic data and the generated data using PARMA (1,1). And, the testing data consist of the only historic data, respectively. From this research, we evaluate the impact of GRNNM and MLP-NNM for the disaggregation of the nonlinear time series data. We should, furthermore, construct the credible data of the monthly PE data from the disaggregation of the yearly PE data, and can suggest the methodology for the irrigation and drainage networks system.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2009년도 학술발표회 초록집
• /
• pp.1211-1214
• /
• 2009

The goal of this research is to apply the neural networks models for the disaggregation of the pan evaporation (PE) data, Republic of Korea. The neural networks models consist of generalized regression neural networks model (GRNNM) and multilayer perceptron neural networks model (MLP-NNM), respectively. The disaggregation means that the yearly PE data divides into the monthly PE data. And, for the performances of the neural networks models, they are composed of training and test performances, respectively. The training data consist of the generated data using PARMA (1,1). And, the testing data consist of the historic data, respectively. From this research, we evaluate the impact of GRNNM and MLP-NNM for the disaggregation of the nonlinear time series data. We should, furthermore, construct the credible data of the monthly PE data from the disaggregation of the yearly PE data, and can suggest the methodology for the irrigation and drainage networks system.