• 전자공학회논문지B
• /
• 제33B권12호
• /
• pp.95-105
• /
• 1996

This paper presents structure optimization of a feedforward neural netowrk controller using the genetic algorithm. It is important to design the neural network with minimum structure for fast response and learning. To minimize the structure of the feedforward neural network, a genralization of multilayer neural netowrks, the genetic algorithm uses binary coding for the structure and floating-point coding for weights. Local search with an on-line learnign algorithm enhances the search performance and reduce the time for global search of the genetic algorithm. The relative fitness defined as the multiplication of the error and node functions prevents from premature convergence. The feedforward neural controller of smaller size outperformed conventional multilayer perceptron network controller.

• 산업경영시스템학회지
• /
• 제21권47호
• /
• pp.35-45
• /
• 1998

Scheduling problem in Flexible Manufacturing Systems(FMS) is complex because of various situation of Manufacturing Systems. Especially, in case of short-term scheduling demanding high efficiency, low cost at short-period, efficient scheduling is a serious problem. To solve this problem, many dispatching rules are developed. But, it leave much to be desired, because real situation in shop floor is complex and real-time scheduling is needed in real manufacturing shop floor. In this paper, search algorithm that allocate different dispatching rules to each machine is presented to complement lack of dispatching rule and develop practical real-time scheduling system. The search algorithm is described in detail. First, algorithm detect machine breakdown, evaluate each dispatching rule. dispatching rules for each machine meeting performance criteria are ranked. The algorithm selects new dispatching nile for bottleneck machine. The effectivenes and efficiency of the mixed dispatching rule and search algorithm is demonstrated.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2000년도 ITC-CSCC -2
• /
• pp.803-806
• /
• 2000

The correspondence problem is known to be difficult to solve because false positives and false negatives almost always exist in real image sequences. In this paper, we propose a robust feature tracking algorithm considering incomplete trajectories such as entering and/or vanishing trajectories. We solve the correspondence problem as the optimal graph search problem, by considering false feature points and by properly reflecting motion characteristics. The proposed algorithm finds a local optimal correspondence so that the effect of false feature points can be minimized in the decision process. The time complexity of the proposed graph search algorithm is given by O(mn) in the best case and O(m$^2$n) in the worst case, where m and n are the number of feature points in two consecutive frames. The proposed algorithm can find trajectories correctly and robustly, which has been shown by experimental results.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제8권3호
• /
• pp.220-224
• /
• 2008

This paper presents the polygon-based Q-leaning and Parallel SVM algorithm for object search with multiple robots. We organized an experimental environment with one hundred mobile robots, two hundred obstacles, and ten objects. Then we sent the robots to a hallway, where some obstacles were lying about, to search for a hidden object. In experiment, we used four different control methods: a random search, a fusion model with Distance-based action making (DBAM) and Area-based action making (ABAM) process to determine the next action of the robots, and hexagon-based Q-learning, and dodecagon-based Q-learning and parallel SVM algorithm to enhance the fusion model with Distance-based action making (DBAM) and Area-based action making (ABAM) process. In this paper, the result show that dodecagon-based Q-learning and parallel SVM algorithm is better than the other algorithm to tracking for object.

• Journal of Information Processing Systems
• /
• 제15권1호
• /
• pp.116-126
• /
• 2019

In this paper, an improved one-against-one support vector machine algorithm is used to classify multiple power quality disturbances. To solve the problem of parameter selection, an improved particle swarm optimization algorithm is proposed to optimize the parameters of the support vector machine. By proposing a new inertia weight expression, the particle swarm optimization algorithm can effectively conduct a global search at the outset and effectively search locally later in a study, which improves the overall classification accuracy. The experimental results show that the improved particle swarm optimization method is more accurate than a grid search algorithm optimization and other improved particle swarm optimizations with regard to its classification of multiple power quality disturbances. Furthermore, the number of support vectors is reduced.

• Structural Engineering and Mechanics
• /
• 제77권6호
• /
• pp.779-795
• /
• 2021

In the recent decades, various optimization algorithms have been considered for the optimization of structures. In this research, a new enhanced algorithm is used for the size and topology optimization of truss structures. This algorithm, which is obtained from the combination of Crow Search Algorithm (CSA) and the Cellular Automata (CA) method, is called CA-CSA method. In the first iteration of the CA-CSA method, some of the best designs of the crow's memory are first selected and then located in the cells of CA. Then, a random cell is selected from CA, and the best design is chosen from the selected cell and its neighborhood; it is considered as a "local superior design" (LSD). In the optimization process, the LSD design is used to modify the CSA method. Numerical examples show that the CA-CSA method is more effective than CSA in the size and topology optimization of the truss structures.

• 대한교통학회지
• /
• 제24권3호
• /
• pp.177-187
• /
• 2006

IT 산업이 발달하고, 정보의 양이 넘쳐날수록, 사람들은 획일화되어 제공되는 정보보다는, 스스로 다양한 경로를 통해 정보를 찾아내며, 이를 가공하여 판단하고 반응한다. 그러므로 정보 제공자들은 이러한 개인들의 성향을 만족시키기 위해 서는 획일화된 정보보다는 소비자들이 스스로 판단할 수 있도록 다양한 정보를 제공해 주어야만 할 것이다 이를 위하여 비용의 비교를 통해 경로를 선택하는 기존 알고리즘과 달리 최저비용과의 차이를 통한 알고리즘을 제안한다. 이를 위해 본 연구는 기존의 노드 기반 탐색법에 비해 네트워크 구조의 변화 없이 효율적으로 환승이나 회전제약을 표현할 수 있는 링크 기반 탐색법을 기반으로 운전자들의 다양한 needs를 최대한 반영할 수 있는 즉 유연한 탐색 알고리즘의 개발을 목표로 한다. 이러한 목표를 위해, 기존의 최적 경로와 다경로 탐색 알고리즘을 대상으로 이론적 배경을 고찰하고, 다목적 정보제공을 위한 다경로 탐색기법을 위한 통행원리를 개념화한 후, 이를 알고리즘에 적용하는 방안을 제안하며, 가상의 네트워크에 적용하여 알고리즘 수행과정을 보여주고자 한다

• 정보처리학회논문지:컴퓨터 및 통신 시스템
• /
• 제1권2호
• /
• pp.79-86
• /
• 2012

선박경제운항이란 기상예측정보를 활용하여 연료소모량을 최소화하도록 선박을 운항하는 것으로서 최근 다양한 선박경제운항 시스템이 연구되고 있다. 성공적인 선박경제운항을 위해서는 효율적인 최적의 지리적 경로탐색 방법이 필요한데 기존의 시스템에서는 주로 Dijkstra 알고리즘 기반의 최소정적비용 경로탐색 알고리즘으로 접근하고 있다. 그러한 접근법을 적용하기 위해서는 특히 연료소모량으로 정의되는 간선의 비용을 고정해야 하는데 선박이 그 간선을 실제 지날 때의 기상 상황에 따라 연료소모량이 변할 수 있다는 점에서 적절하지 않은 가정이다. 이에 본 논문에서는 그러한 단점을 극복하기 위해 Dijkstra 알고리즘을 변형한 최소동적비용 경로탐색 알고리즘을 제안한다. 또한, 실행시간을 단축하기 위해 $A^*$ 알고리즘을 활용하여 탐색공간을 효과적으로 줄이기 위한 방법도 제시한다. 총 10개의 테스트 노선에 대해서 본 논문에서 제안된 시스템을 기존의 단순한 최단거리 운항방법과 비교한 결과, 운항소요시간은 거의 차이가 없으면서도 연료소모량을 평균 2.36%, 최대 4.82% 개선시킬 수 있었다.

• International Journal of Internet, Broadcasting and Communication
• /
• 제12권4호
• /
• pp.102-108
• /
• 2020

One of the key features of a mobile crowd-sensing (MCS) system is task allocation, which aims to recruit workers efficiently to carry out the tasks. Due to various constraints of the tasks (such as specific sensor requirement and a probabilistic guarantee of task completion) and workers heterogeneity, the task allocation become challenging. This assignment problem becomes more intractable because of the deadline of the tasks and a lot of possible task completion order or moving path of workers since a worker may perform multiple tasks and need to physically visit the tasks venues to complete the tasks. Therefore, in this paper, a hybrid search algorithm for task allocation called HST is proposed to address the problem, which employ a traveling salesman problem heuristic to find the task completion order. HST is developed based on the tabu search algorithm and exploits the premature convergence avoiding concepts from the genetic algorithm and simulated annealing. The experimental results verify that our proposed scheme outperforms the existing methods while satisfying given constraints.

• 전기학회논문지
• /
• 제56권9호
• /
• pp.1521-1526
• /
• 2007

This paper presents the application of Adaptive Evolutionary Algorithm (AEA) to search an optimal setting of overcurrent relay coordination to protect ring distribution systems. The AEA takes the merits of both a genetic algorithm (GA) and an evolution strategy (ES) in an adaptive manner to use the global search capability of GA and the local search capability of ES. The overcurrent relay settings and coordination requirements are formulated into a set of constraint equations and an objective function is developed to manage the overcurrent relay settings by the Time Coordination Method. The domain of overcurrent relays coordination for the ring-fed distribution systems is a non-linear system with a lot of local optimum points and a highly constrained optimization problem. Thus conventional methods fail in searching for the global optimum. AEA is employed to search for the optimum relay settings with maximum satisfaction of coordination constraints. The simulation results show that the proposed method can optimize the overcurrent relay settings, reduce relay mis-coordinated operations, and find better optimal overcurrent relay settings than the present available methods.