• 한국태양에너지학회 논문집
• /
• 제37권1호
• /
• pp.81-90
• /
• 2017

We have analyzed the performance of 58 kWp photovoltaic (PV) power systems installed in Jeddah, Saudi Arabia. Performance ratio (PR) of 3 PV systems with 3 desert-type PV modules using monitoring data for 1 year showed 85.5% on average. Annual degradation rate of 5 individual modules achieved 0.26%, the regression model using monitoring data for the specified interval of one year showed 0.22%. Root mean square error (RMSE) of 6 big data analysis models for power output prediction in May 2016 was analyzed 2.94% using a support vector regression model.

• 유체기계공업학회:학술대회논문집
• /
• 유체기계공업학회 2000년도 유체기계 연구개발 발표회 논문집
• /
• pp.176-183
• /
• 2000

The performance of gas turbines decreases as their operating hours increase. Compressor fouling is the main reason for this time-dependent performance degradation. Airborne particles adhere to the blade surface and results in the change in the blade shape. It is difficult to exactly analyze the mechanism of the compressor fouling because the growing process of the fouling is very slow and the dimension of the fouled depth is very small compared with blade dimensions. In this study, an analytic method to predict the motion of particles and their deposition inside axial flow compressors is proposed. The analytic model takes into account the blade shape and the flow within the blade passage. Comparison of simulation result with field data shows the feasibility of the model. Influence of the particle distribution on the compressor fouling is also examined.

• KEPCO Journal on Electric Power and Energy
• /
• 제6권4호
• /
• pp.413-418
• /
• 2020

초기 학습 데이터의 과적합으로 인한 전력망 상태예측 모델의 성능 감소를 방지하고 예측모델의 예측 정확도 유지를 통한 계속적인 현장활용을 위해서는 기계학습 모델의 예측 정확도를 지속적으로 관리할 필요가 있다. 이를 위해, 본 논문에서는 다양한 요인에 의해 끊임없이 변화하는 전력망 상태 데이터의 특성을 고려하여 예측모델의 정확성과 신뢰성을 높이고 현장 적용 가능한 수준의 품질을 유지하기 위한 기계학습 기반 전력망 상태예측 모델의 성능 유지관리 자동화 기법을 제안한다. 제안 기법은 워크플로우 관리 기술의 적용을 통해 전력망 상태예측 모델 성능 유지관리를 위한 일련의 태스크들을 워크플로우의 형태로 모델링하고 이를 자동화하여 업무를 효율화 하였다. 또한, 기존 기술에서는 시도되지 않았던 학습데이터의 통계적 특성 변화 정도와 예측의 일반화 수준을 모두 고려한 예측모델의 성능 평가를 통해 성능 결과의 신뢰성을 확보하고 이를 통해 예측 모델의 정확도를 일정 수준으로 유지관리하고 더욱 성능이 우수한 예측모델의 신규 개발이 가능하다. 결과적으로 본 논문에서 제안하는 전력망 상태예측 모델 성능 유지관리 자동화 기법을 통해 예측모델의 성능 저하문제를 해결하여 분산자원 연계 등 외부 환경의 변화에 유연한 예측모델 관리를 통해 정확성과 신뢰성이 보장된 예측 모델의 지속적인 활용이 가능하다.

• International Journal of Contents
• /
• 제13권4호
• /
• pp.23-28
• /
• 2017

Of the total economic loss caused by disasters, 40% are due to floods and floods have a severe impact on human health and life. So, it is important to monitor the water level of a river and to issue a flood warning during unfavorable circumstances. In this paper, we propose a modified error function to improve a hydrological modeling using a multi-layer perceptron (MLP) neural network. When MLP's are trained to minimize the conventional mean-squared error function, the prediction performance is poor because MLP's are highly tunned to training data. Our goal is achieved by preventing overspecialization to training data, which is the main reason for performance degradation for rare or test data. Based on the modified error function, an MLP is trained to predict the water level with rainfall data at upper reaches. Through simulations to predict the water level of Nakdong River near a UNESCO World Heritage Site "Hahoe Village," we verified that the prediction performance of MLP with the modified error function is superior to that with the conventional mean-squared error function, especially maximum error of 40.85cm vs. 55.51cm.

• 지능정보연구
• /
• 제16권4호
• /
• pp.99-112
• /
• 2010

앙상블 학습은 분류 및 예측 알고리즘의 성과개선을 위하여 제안된 기계학습 기법이다. 그러나 앙상블 학습은 기저 분류자의 다양성이 부족한 경우 다중공선성 문제로 인하여 성과개선 효과가 미약하고 심지어는 성과가 악화될 수 있다는 문제점이 제기되었다. 본 연구에서는 기저 분류자의 다양성을 확보하고 앙상블 학습의 성과개선 효과를 제고하기 위하여 유전자 알고리즘 기반의 범위 최적화 기법을 제안하고자 한다. 본 연구에서 제안된 최적화 기법을 기업 부실예측 인공신경망 앙상블에 적용한 결과 기저 분류자의 다양성이 확보되고 인공신경망 앙상블의 성과가 유의적으로 개선되었음을 보여주었다.

• Computers and Concrete
• /
• 제17권4호
• /
• pp.499-521
• /
• 2016

In this paper, a probabilistic- and finite element-based approach to evaluate and predict the lifetime performance of reinforced concrete (RC) bridges undergoing various maintenance actions is proposed with the time-variant system reliability being utilized as a performance indicator. Depending on their structural state during the degradation process, the classical maintenance actions for RC bridges are firstly categorized into four types: Preventive type I, Preventive type II, Strengthening and Replacement. Preventive type I is used to delay the onset of steel corrosion, Preventive type II can suppress the corrosion process of reinforcing steel, Strengthening is the application of various maintenance materials to improve the structural performance and Replacement is performed to restore the individual components or overall structure to their original conditions. The quantitative influence of these maintenance types on structural performance is investigated and the respective analysis modules are written and inputted into the computer program. Accordingly, the time-variant system reliability can be calculated by the use of Monte Carlo simulations and the updated the program. Finally, an existing RC continuous bridge located in Shanghai, China, is used as an illustrative example and the lifetime structural performance with and without each of the maintenance types are discussed. It is felt that the proposed approach can be applied to various RC bridges with different structural configurations, construction methods and environmental conditions.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2004년도 학술대회 논문집 정보 및 제어부문
• /
• pp.82-85
• /
• 2004

This paper proposes the development of delay compensator to minimize performance degradation caused by time delays in network-based real-time control systems. The delay compensator uses the time-stamp method as a direct delay measuring method to measure time delays generated between network nodes. The delay compensator predicts the network time delays of next period in the views point of time delays and minimizes performance degradation from network through considering predicted time delays. Control output considering network time delays is generated by the defuzzification of probable time delays of next period. The time delays considered in the delay compensator are modeled by using a timed Petri net model. The proposed delay prediction mechanism for the delay compensator is evaluated through some simulation tests by measuring deviation of the predicted delays from simulated delays.

• 한국전기전자재료학회:학술대회논문집
• /
• 한국전기전자재료학회 1998년도 춘계학술대회 논문집
• /
• pp.297-300
• /
• 1998

In this paper, we obtained the data, which is required in training the neural network and diagnosing the degradation degree, by introducing the AE detection that is effective method in ordinary degradation diagnosis on activation. Automatic detection system to detect acoustic. As the results of generalization tests by appling neural network to the unknown AE patterns obtained from specimens, firstly as to evaluate an objective performance of neural network, the recognition ratio for no-void specimen is appeared. Also, in the evaluation for the adaptability of neural network with a untrained type of no-void specimen, it is confirmed that the result appears.

• IEIE Transactions on Smart Processing and Computing
• /
• 제6권3호
• /
• pp.183-192
• /
• 2017

k-nearest neighbor (K-NN) is a well-known classification algorithm, being feature space-based on nearest-neighbor training examples in machine learning. However, K-NN, as we know, is a lazy learning method. Therefore, if a K-NN-based system very much depends on a huge amount of history data to achieve an accurate prediction result for a particular task, it gradually faces a processing-time performance-degradation problem. We have noticed that many researchers usually contemplate only classification accuracy. But estimation speed also plays an essential role in real-time prediction systems. To compensate for this weakness, this paper proposes correlation coefficient-based clustering (CCC) aimed at upgrading the performance of K-NN by leveraging processing-time speed and plurality rule-based density (PRD) to improve estimation accuracy. For experiments, we used real datasets (on breast cancer, breast tissue, heart, and the iris) from the University of California, Irvine (UCI) machine learning repository. Moreover, real traffic data collected from Ojana Junction, Route 58, Okinawa, Japan, was also utilized to lay bare the efficiency of this method. By using these datasets, we proved better processing-time performance with the new approach by comparing it with classical K-NN. Besides, via experiments on real-world datasets, we compared the prediction accuracy of our approach with density peaks clustering based on K-NN and principal component analysis (DPC-KNN-PCA).

• 한국정보과학회논문지:컴퓨팅의 실제 및 레터
• /
• 제15권4호
• /
• pp.280-283
• /
• 2009

H.264/AVC의 확장 표준으로 제정된 SVC(Scalable Video Coding)는 공간적 확장성의 압축 효율을 높이기 위해 기존 H.264/AVC에서 제공하는 인트라 예측과 인터 예측뿐만 아니라 계층 간 예측을 추가로 수행한다. 그로인해 부호화 계산량이 더욱 증가되는 문제점이 있다. 본 논문에서는 공간적 향상 계충에서 인트라 예측 모드를 효율적으로 선택하는 방법을 제안한다. 제안한 방법은 실험을 통한 Intra_BL 모드의 RD 값을 이용하여 미리 Intra_BL 모드를 선택한 후, 나머지 모드를 다 수행하지 않고 대표적인 DC 모드만을 비교하여 빠른 인트라 예측 모드를 결정한다. 실험 결과 화질 저하는 적은 데 비해 인트라 예측 모드 부호화 시간은 약 59% 감소되었다.