• The Transactions of the Korea Information Processing Society
• /
• v.3 no.7
• /
• pp.1773-1780
• /
• 1996

KBANN (knowledge-based artificial neural network) combining the symbolic approach and the numerical approach has been shown to be more effective than other machine learning models. However KBANN doesn't have the theory refinement ability because the topology of network can't be altered dynamically. Although TopGen was proposed to extend the ability of KABNN in this respect, it also had some defects due to the link-ing of hidden nodes to input nodes and the use of beam search. The algorithm which could solve this TopGen's defects, by adding the hidden nodes linked to next layer nodes and using hill-climbing search with backtracking, is designed.

• Journal of Electrical Engineering and Technology
• /
• v.12 no.6
• /
• pp.2146-2156
• /
• 2017

When a failure occurs in active distribution system, it will be isolated through the action of circuit breakers and sectionalizing switches. As a result, the network might be divided into several connected components, in which distributed generations could supply power for customers. Aimed at decreasing customer interruption cost, this paper proposes a theoretically optimal island partition model for such connected components, and a simplified but more practical model is also derived. The model aims to calculate a dynamic island partition schedule during the failure recovery time period, instead of a static islanding status. Fluctuation and stochastic characteristics of the renewable distributed generations and loads are considered, and the interruption cost functions of the loads are fitted. To solve the optimization model, a heuristic search algorithm based on the hill climbing method is proposed. The effectiveness of the proposed model and algorithm is evaluated by comparing with an existing static island partitioning model and intelligent algorithms, respectively.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.33 no.10
• /
• pp.1029-1037
• /
• 2009

Recently, the technologies of mobile robots have been growing rapidly in the fields such as cleaning robot, explosive ordnance disposal robot, patrol robot, etc. However, the researches about the autonomous underwater robots have not been done so much, and they still remain at the low level of technology. This paper describes a model of 3-joint (4 links) fish robot type. Then we calculate the dynamic motion equation of this fish robot and use Singular Value Decomposition (SVD) method to reduce the divergence of fish robot's motion when it operates in the underwater environment. And also, we analysis response characteristic of fish robot according to the parameters of input torque function and compare characteristic of fish robot with 3 joint and fish robot with 2 joint. Next, fish robot's maximum velocity is optimized by using the combination of Hill Climbing Algorithm (HCA) and Genetic Algorithm (GA). HCA is used to generate the good initial population for GA and then use GA is used to find the optimal parameters set that give maximum propulsion power in order to make fish robot swim at the fastest velocity.

• Journal of IKEEE
• /
• v.22 no.2
• /
• pp.460-475
• /
• 2018

In the variable-speed wind energy system, to achieve maximum power point tracking (MPPT), the wind turbine should run close to its optimal angular speed according to the wind speed. Non-linear control methods that consider the dynamic behavior of wind speed are generally used to provide maximum power and improved efficiency. In this perspective, the mechanical power is estimated using Kalman filter. And then, from the estimated mechanical power, the wind speed is estimated with Newton-Raphson method to achieve maximum power without anemometer. However, the blade shape and air density get changed with time and the generator efficiency is also degraded. This results in incorrect estimation of wind speed and MPPT. It causes not only the power loss but also incorrect wind resource assessment of site. In this paper, the adaptive maximum power point tracking control algorithm for wind turbine system based on the estimation of wind speed is proposed. The proposed method applies correction factor to wind turbine system to have accurate wind speed estimation for exact MPPT. The proposed method is validated with numerical simulations and the results show an improved performance.

• Journal of Institute of Control, Robotics and Systems
• /
• v.17 no.6
• /
• pp.587-595
• /
• 2011

In this paper, we focus on the issue of automatic selection for multi-level threshold, and we greatly improve the efficiency of Otsu's method for image segmentation based on genetic algorithm. We have investigated and evaluated the performance of the Otsu and Valley-emphasis threshold methods. Based on this observation we propose a method for automatic threshold method that segments an image into more than two regions with high performance and processing in real-time. Our paper introduced new peak detection, combines with evolution algorithm using MAGA (Modified Adaptive Genetic Algorithm) and HCA (Hill Climbing Algorithm), to find the best threshold automatically, accurately, and quickly. The experimental results show that the proposed evolutionary algorithm achieves a satisfactory segmentation effect and that the processing time can be greatly reduced when the number of thresholds increases.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2002.11a
• /
• pp.321-327
• /
• 2002

본 논문은 기존의 귀납적 결정 트리 방식에서의 문제점 개선을 통한 사용자 관심 프로파일 구축을 목적으로 한다. 특히 사용자 관심 프로파일의 정확도 향상을 위한 속성 선택에 대한 연구에 초점을 맞추고 있다. 사용자의 관심, 비관심 문서를 대상으로 사용자 관심 키워드를 생성하고 이를 바탕으로 초기 문서들을 재표현한다. 재표현된 문서를 입력 집합으로 하여 기계학습을 진행한다. 본 논문의 의사 결정 트리 생성 알고리즘은 입력 집합을 클래스별로 가장 잘 나누는 속성을 선택하여 노드를 구성하는 면에서는 기존의 알고리즘과 같다. 그러나 기존의 의사 결정 트리 알고리즘에서는 hill-climbing.방식을 사용함으로써 사용자의 관심을 나타내는 중요한 단어가 사용자 관심 프로파일에서 숨겨질 경우가 발생한다. 이를 최소화하기 위해 특징 추출을 통해 선택된 속성을 그대로 학습의 입력 데이터로 사용하는 것이 아니라 입력데이터를 가장 잘 나누는 속성과 그 다음 속성을 대상으로 disjunctive 연산을 통해 새로운 속성을 생성하여 이것을 속성 집합에 포함시키고 이를 학습의 입력 데이터로 이용한다. 이와 같이 disjunctive operator를 이용하여 새로운 속성을 의사 결정 트리 형성 시 이용하면 사용자의 중요한 관심을 포함하는 의미 있는(semantic) 사용자 관심 프로파일 구축이 가능해지고, 사용자 관심 프로파일을 기반으로 사용자가 관심 있는 문서를 제공할 수 있는 개인화 서비스를 제공한다.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.9 no.4
• /
• pp.319-325
• /
• 2004

The major factors that make ZEV affordable are the range and cost. The development of advanced batteries such as Ni-MH battery can solve the problem partly； on the hand the battery management system is an efficient way. Ni-MH battery and battery ECU is a key component influencing ZEV performance, such as range, acceleration and hill-climbing capability. Because most problems related to battery such as short circuit, over-discharge and overcharge occur easily during operation, it is necessary to develop a dedicated battery ECU for HEV. This paper proposes a new SOC algorithm for the HEV based on the terminal voltage and current integration. And battery ECU was designed and analyzed. Also, the validity is confirmed through experiment.

• Journal of Power Electronics
• /
• v.16 no.2
• /
• pp.666-674
• /
• 2016

In this paper, a high efficiency satellite electrical power supply system is proposed. The increased efficiency of the power supply system allows for downscaling of the solar array and battery weight, which are among the most important satellite design considerations. The satellite power supply system comprises two units, namely a generation unit and a storage unit. To increase the efficiency of the solar array, a maximum power point tracker (MPPT) is used in the power generation unit. In order to improve the MPPT performance, a novel algorithm is proposed on the basis of the hill climbing method. This method can track the main peak of the array power curve in satellites with long duration missions under unpredicted circumstances such as a part of the array being damaged or the presence of a shadow. A lithium-ion battery is utilized in the storage unit. An algorithm for calculating the optimal rate of battery charging is proposed where the battery is charged with the maximum possible efficiency considering the situation of the satellite. The proposed system is designed and manufactured. In addition, it is compared to the conventional power supply systems in similar satellites. Results show a 12% increase in the overall efficiency of the power supply system when compared to the conventional method.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.19 no.5
• /
• pp.7-13
• /
• 2011

The climate change due to the increased consumption with fossil fuel and rise of the oil price have been serious global issues. Automobile industry consumes 30% of the oil every year and causes air pollution and global warming by the exhaust emissions and carbon dioxide ($CO_2$). The demand of two-wheeled vehicle increases every year due to the parking and traffic problem caused by the increased automobiles in the urban area. Approximately 50,000,000 two-wheeled vehicles were produced in 2008. The development and sales of the hybrid two-wheeled vehicle industry become active due to its increased market demands. In this paper, the change of the motor and battery efficiency, driving distance, hill climbing ability with the change of the motor capacity was analyzed. Simulation of the peculiarities in urban driving schedule(World-wide Motorcycle Test Cycle(WMTC), Manhattan driving schedule), constant speed(10 km/h, 35 km/h) of small electronic two-wheeled vehicle was also carried out. Through the simulation result, appropriate capacities of the motor and battery for urban driving was acquired.

• KIISE Transactions on Computing Practices
• /
• v.21 no.1
• /
• pp.40-45
• /
• 2015

Online social networks(OSN) are very popular nowadays. As OSNs grows, the commercial markets are expanding their social commerce by applying Influence Maximization. However, in reality, there exist more than two players(e.g., commercial companies or service providers) in this same market sector. To address the Influence Maximization problem between adversaries, we first introduced Influence Maximization against the social adversaries' problem. Then, we proposed an algorithm that could efficiently solve the problem efficiently by utilizing social network properties such as Betweenness Centrality, Clustering Coefficient, Local Bridge and Ties and Triadic Closure. Moreover, our algorithm performed orders of magnitudes better than the existing Greedy hill climbing algorithm.