• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2008년도 지능정보 및 응용 학술대회
• /
• pp.87-90
• /
• 2008

동적 휴리스틱이란 고정된 정보가 아닌 서버로부터 실시간 도로정보를 전송받아 동적으로 변경되는 휴리스틱이다. 이 논문에서는 최고속력 동적휴리스틱을 이용하는 경로탐색 기법을 제안하고, 성능실험을 하였다. 최고속력 동적 휴리스틱이란 지도를 일정크기로 나눈 그리드내의 도로 중 최고속력을 동적 휴리스틱으로 활용하며, 최고속력이 높은 곳을 우선 탐색함으로 인해 거리에 따른 정보뿐만 아니라 실시간 정보를 활용하여 경로를 탐색한다. 성능 실험 결과 $A^*$ 알고리즘에 비해 경로의 질은 향상되었으나 탐색비용이 조금 늘었다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제7권8호
• /
• pp.1825-1842
• /
• 2013

Rapid growth of the IT industry has led to significant energy consumption in the last decade. Data centers swallow an enormous amount of electrical energy and have high operating costs and carbon dioxide excretions. In response to this, the dynamic consolidation of virtual machines (VMs) allows for efficient resource management and reduces power consumption through the live migration of VMs in the hosts. Moreover, each client typically has a service level agreement (SLA), this leads to stipulations in dealing with energy-performance trade-offs, as aggressive consolidation may lead to performance degradation beyond the negotiation. In this paper we propose a heuristic based resource allocation of VM selection and a VM allocation approach that aims to minimize the total energy consumption and operating costs while meeting the client-level SLA. Our experiment results demonstrate significant enhancements in cloud providers' profit and energy savings while improving the SLA at a certain level.

• 산업경영시스템학회지
• /
• 제35권2호
• /
• pp.57-63
• /
• 2012

소프트웨어 개발에 있어서 소프트웨어를 시장에 출시하는 계획을 수립하는 것은 소프트웨어를 이루고 있는 기능들을 구현하는 데 제약이 되는 조건들(기술, 자원, 위험, 예산 등)을 만족하면서 계획된 출시기간에 이들 기능들을 할당하는 일이다. 이와 같이 소프트웨어 출시를 계획하는 것은 소프트웨어 제품라인에 대해서 고려할 때 더욱 복잡해진다. 본 연구에서는 소프트웨어 제품라인에 있어서 소프트웨어 출시 계획을 수립하기 위한 문제를 우선순위 제약하의 다수 0-1 배낭문제로 수리 모형화하고, 이를 풀기 위한 최적해법이 개발된다. 최적해법은 동적 계획법이 주가 되고, 문제의 크기를 줄이기 위하여 휴리스틱과 축소방법이 이용된다.

• 전자공학회논문지B
• /
• 제30B권11호
• /
• pp.84-92
• /
• 1993

Calibration of Engine models is a painstaking process but very important for successful application to automotive industry problems. A combined heuristic and machine learning approach has therefore been adopted to improve the efficiency of model calibration. We developed an intelligent calibration program called ICALIB. It has been used on a daily basis for engine model applications, and has reduced the time required for model calibrations from many hours to a few minutes on average. In this paper, we describe the heuristic control strategies employed in ICALIB such as a hill-climbing search based on a state distance estimation function, incremental problem solution refinement by using a dynamic tolerance window, and calibration target parameter ordering for guiding the search. In addition, we present the application of amachine learning program called GID3*for automatic acquisition of heuristic rules for ordering target parameters.

• 제어로봇시스템학회논문지
• /
• 제9권8호
• /
• pp.623-631
• /
• 2003

We have fabricated the two-dimensional inverted pendulum system and designed its controller. The two-dimensional inverted pendulum system, which is composed of X-Y table, is actuated through timing belt by each of two geared DC motors. And the control goal is that the rod is always kept to a vertical position to any distrubance and is quickly moved to the desired position. Because this system has generally nonlinear dynamic characteristics and X-axis and Y-axis move together, it is very difficult to find its exact mathematical model and to design its controller. Therefore, we have designed the PID controller with simple structure and excellent performance. Genetic algorithm(GA), which is blown as one of probabilistic searching methods, and human's heuristic control strategy are introduced to design an optimal PID controller. The usefulness of the proposed GA based PID coefficient searching technique is verified through the experiments and computer simulations.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제52권5호
• /
• pp.283-289
• /
• 2003

This paper proposes an expert system which can determine automatically the shunting routes corresponding to the given shunting works by considering totally the train operating environments in the station. The expert system proposes the multiple shunting routes with priority of selection based on heuristic search strategy. Accordingly, system operator can select a shunting route with the safety and efficiency among the those shunting routes. The expert system consists of a main inference engine and a sub inference engine. The main inference engine determines the shunting routes with selection priority using the segment routes obtained from the sub inference engine. The heuristic rules are extracted from operating knowledges of the veteran route operator and station topology. It is implemented in C computer language for the purpose of the implementation of the inference engine using the dynamic memory allocation technique. And, the validity of the builted expert system is proved by a test case for the model station.

• 산업공학
• /
• 제22권1호
• /
• pp.63-72
• /
• 2009

This paper considers a transporter scheduling problem under dynamic block transportation environment in shipbuilding. In dynamic situations, there exist the addition, cancellation or change of block transportation requirements, sudden breakdowns and maintenance of transporters. The transportation of the blocks in the shipyard has some distinct characteristics. Some blocks are available to be picked up at a specific time during the planning horizon while some other blocks need to be delivered before a specific time. These requirements cause two penalty times: 1) delay times incurred when a block is picked up after a required start time, and 2) tardy times incurred when a block shipment is completed after the required delivery time. The blocks are located at different areas in the shipyard and transported by transporters. The objective of this paper is to propose a heuristic algorithm based on a network flow model which minimize the weighted sum of empty transporter travel times, delay times, and tardy times. Also, a rolling-horizon scheduling method is proposed for dynamic block transportation environment. The performance of the proposed heuristic algorithms are evaluated through a simulation experiment.

• Journal of Electrical Engineering and Technology
• /
• 제11권2호
• /
• pp.329-337
• /
• 2016

With the increasing deployment of distributed generators in the distribution system, a very large search space is required when dynamic programming (DP) is applied for the optimized dispatch schedules of voltage and reactive power controllers such as on-load tap changers, distributed generators, and shunt capacitors. This study proposes a new optimal voltage and reactive power scheduling method based on dynamic programming with a heuristic searching space reduction approach to reduce the computational burden. This algorithm is designed to determine optimum dispatch schedules based on power system day-ahead scheduling, with new control objectives that consider the reduction of active power losses and maintain the receiving power factor. In this work, to reduce the computational burden, an advanced voltage sensitivity index (AVSI) is adopted to reduce the number of load-flow calculations by estimating bus voltages. Moreover, the accumulated switching operation number up to the current stage is applied prior to the load-flow calculation module. The computational burden can be greatly reduced by using dynamic programming. Case studies were conducted using the IEEE 30-bus test systems and the simulation results indicate that the proposed method is more effective in terms of saving electric charges and improving the voltage profile than loss minimization.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2008년도 춘계종합학술대회 A
• /
• pp.179-182
• /
• 2008

실제 도로 교통망에서 경로를 탐색할 때 방향전환에 대해 고려해야 한다. 왜냐하면 차량이 이동할 때 교차로나 진입도로 등과 같이 방향전환이 있을 경우 교통신호 또는 속력감속으로 인한 운행시간이 지연될 수 있기 때문이다. 만약 경로의 직진성을 높일 경우 실제 이동시간이 줄어들 가능성이 있다. 턴 휴리스틱을 적용한 알고리즘은 이와 같은 경우를 고려한 경로탐색 기법이다. 방향전환에 따라 서로 다른 가중치를 부여한 이 기법은 경로의 직진성을 높였으나, 탐색 비용이 증가하였다. 이 논문에서는 개선된 턴 휴리스틱을 적용한 알고리즘을 제안한다. 제안하는 기법은 가변적인 턴 휴리스틱을 사용한다. 즉, 경로탐색 중 일부 구간에 대해서만 턴 휴리스틱을 적용함으로써, 기존연구에 비해 탐색 비용을 줄일 가능성을 높였다.

• 한국정보처리학회논문지
• /
• 제4권10호
• /
• pp.2470-2476
• /
• 1997

본 연구에서는 실시간 휴리스틱 탐색 알고리즘을 개발하고 이것을 기존의 mini-min lookahead 알고리즘과 비교하였다. 많은 실시간 휴리스틱 탐색의 접근 방법에서 종종 전체 문제를 몇 개의 부 문제로 문제를 분할한다. 본 연구에서는 분할된 부 문제에서 마감시간을 적용할 뿐만 아니라 전체 해를 구하는데 있어서도 마감시간을 적용하는 알고리즘을 제안한다. 실시간 휴리스틱 탐색 알고리즘으로 제안된 $RTA^{\ast}$, SARTS, DYNORA 등의 알고리즘들은 탐색에 필요한 시간의 예측을 휴리스틱 평가 함수로부터 얻기 때문에 휴리스틱 평가의 정확도가 그 알고리즘의 성능을 보장하게 된다. 그러나 실세계의 문제에서 정확한 휴리스틱 평가 함수를 구하는 것은 매우 어려운 일이므로 부 문제 공간에서의 탐색 상황을 반영한 마감시간을 적용할 필요가 있다. 본 연구에서는 동적 마감시간 전략인 cut-off 방법을 사용하는 새로운 알고리즘을 제안한다.