• Journal of Korea Game Society
• /
• v.19 no.6
• /
• pp.83-90
• /
• 2019

Heuristic based Genetic Algorithm Pathfinding(H-GAP), a method without the need for node and edge information, can compensate the disadvantages of existing pathfinding algorithm, and perform the path search at high speed. However, because the pathfinding by H-GAP is non-node-based, it may not be an optimal path when it includes unnecessary path information. In this paper, we propose an algorithm to optimize the search path using H-GAP. The proposed algorithm optimizes the path by removing unnecessary path information through ray-tracing algorithm after the H-GAP path search is completed.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.13 no.8
• /
• pp.1615-1622
• /
• 2009

Generally, the Terminal Based Navigation System(TBNS) used embedded road data searches a path that has less qualitative than The Center Based Navigation System(CBNS). TBNS has not used real time road data but it is recently able to use it with technique such as TPEG. However, it causes to increase a cost of exploring by using real time road data for improvement quality of a path, because of limited performance. In this paper, we propose a Dynamic Heuristic to improve quality of path in the TBNS. Dynamic Heuristic(DH) is not fixed data and is dynamically modified using transferred real time road data from server. In this paper, we propose path-lading algorithm with Maximum Speed Dynamic Heuristic (DH-MAX) and do an experiment. The DH-MAX is to be used the highest speed as DH, in real map divided by same size. And proposed algorithm searches path using the priority searching only of the fixed data, but also the highest speed with real time information. In the performance test, the quality of path is enhanced but the cost of searching is increased than A* algorithm.

• The KIPS Transactions:PartA
• /
• v.10A no.1
• /
• pp.43-48
• /
• 2003

Two-dimensional packing algorithm can be used for allocating submeshes in mesh multiprocessor systems. Previously, we developed an efficient packing algorithm called TP heuristic, and showed how the results of the packing could be used for allocating submeshes. In this paper, we present theoretical performance bounds for TP heuristic. We also present a parallel version of the algorithm that consumes reduced time when it is executed by multiple processors in mesh multiprocessors.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.10a
• /
• pp.457-459
• /
• 2003

클러스터 시스템은 상대적으로 가격이 싼 컴퓨터를 고성능의 네트워크(Network)로 묶어서 슈퍼컴퓨터와 같은 고성능을 가지도록 만들어진 시스템이다. 이런 클러스터 컴퓨팅 환경에서 효율적인 스케줄링은 그 성능에 직접적인 영향을 주는 요소이다. 이런 시스템에서 완전한 동시 스케줄링(Coscheduling)은 서로 교환해야하는 정보가 많아지기 때문에 그 구현이 어렵다. 이 상황에서 메시지를 기다리는 정보와 메시지의 도착 정보를 이용해서 즉 단지 그 노드(Node) 자체의 정보만을 이용해 동시 스케줄링의 효과를 구현할 수 있다. 그리고 이것을 이용한 알고리즘 중에 주기적 추진(Periodic Boost(PB))이 있다. 이 논문에서는 주기적 추진에 휴리스틱을 이용하du 더 효과적인 스케줄링을 할 수 있는 알고리즘을 소개한다. 그리고 이 휴리스틱의 효과를 검증하기 위해서 클러스터 노드 2개를 이용해서 실험을 했다. 실험은 계산대 통신 비율(Communication-to-Computation ratio)을 변화시켜가면서 총 수행시간을 측정하고, 서로 통신하는 양이 다른 프로세스를 섞어서 그 성능을 실험한 결과 휴리스틱이 주기적 추진(PB)에서 불필요하게 낭비되는 자원을 효율적으로 사용할 수 있음을 알 수 있었다.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.4
• /
• pp.890-900
• /
• 1999

In homogeneous multiprocessor systems, the task allocation algorithm which equally assigns tasks to processors if possible is generally used. But this algorithm is not suitable to accomplish to accomplish effective task allocation in heterogeneous multiprocessor systems. JSQ (Join the Shortest Queue) algorithm is often used in heterogeneous multiprocessor systems. Unfortunately, JSQ algorithm is not efficient when the differences of capabilities of processors are far large. To solve this problem, we suggest a heuristic task allocation algorithm that makes use of dynamic information such as task arrival time, task service time, and number of finished tasks. The results of simulation show that the proposed heuristic allocation algorithm improves the system performance.

• Journal of KIISE:Computer Systems and Theory
• /
• v.29 no.7
• /
• pp.377-383
• /
• 2002

The TSP(Traveling Salesman Problem) has been known as NP-complete, there have been various studies to find the near optimal solution. The nearest neighbor heuristic is more simple than the other algorithms which are to find the optimal solution. This paper designs and implements a new distributed nearest neighbor heuristic with bounding function for the TSP using the master/slave model of PVM(Parallel Virtual Machine). Distributed genetic algorithm obtains a near optimal solution and distributed nearest neighbor heuristic finds an optimal solution for the TSP using the near optimal value obtained by distributed genetic algorithm as the initial bounding value. Especially, we get more speedup using a new genetic operator in the genetic algorithm.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.22 no.5
• /
• pp.540-548
• /
• 2012

The Weapon Target Allocation(WTA) problem is the NP-Complete problem. The WTA problem is that the threatful air targets are assigned by weapon of allies for killing the targets. A good solution of NP-complete problem is heuristic algorithms. Genetic algorithms are commonly used heuristic for global optimization, and it is good solution on the diverse problem domain. But there has been very little research done on the generation of their initial population. The initialization of population is one of the GA step, and it decide to initial value of individuals. In this paper, we propose to the population initialization method to improve a Genetic Algorithm. When it initializes population, the proposed algorithm reflects the characteristics of the WTA problem domain, and inherits the dominant gene. In addition, the search space widely spread in the problem space to find efficiently the good quality solution. In this paper, the proposed algorithm to verify performance examine that an analysis of various properties and the experimental results by analyzing the performance compare to other algorithms. The proposed algorithm compared to the other initialization methods and a general genetic algorithm. As a result, the proposed algorithm showed better performance in WTA problem than the other algorithms. In particular, the proposed algorithm is a good way to apply to the variety of situation WTA problem domain, because the proposed algorithm can be applied flexibly to WTA problem by the adjustment of RMI.

• Journal of Korea Game Society
• /
• v.13 no.1
• /
• pp.31-40
• /
• 2013

A* algorithm is a well known pathfinding algorithm. However, there may be a limit to use A* algorithm in real-time in a map where many interactions occur between objects or many obstacles exist. Therefore, it may be necessary to find a naturally looking path quickly instead of finding a shortest path in games. In this paper, we propose a new heuristic function to exploit path information in a map. We also show that the pathfinding algorithm based on the proposed heuristic function can find a good path much faster than A* algorithm on several grid maps.

• The KIPS Transactions:PartB
• /
• v.15B no.2
• /
• pp.159-170
• /
• 2008

Nowadays, one of the important functionalities required from robot task planners is to generate plans to compose existing component services into a new service. In this paper, we introduce the design and implementation of a heuristic search planner, JPLAN, as a kernel module for component service composition. JPLAN uses a local search algorithm and planning graph heuristics. The local search algorithm, EHC+, is an extended version of the Enforced Hill-Climbing(EHC) which have shown high efficiency applied in state-space planners including FF. It requires some amount of additional local search, but it is expected to reduce overall amount of search to arrive at a goal state and get shorter plans. We also present some effective heuristic extraction methods which are necessarily needed for search on a large state-space. The heuristic extraction methods utilize planning graphs that have been first used for plan generation in Graphplan. We introduce some planning graph heuristics and then analyze their effects on plan generation through experiments.

• Journal of Digital Convergence
• /
• v.12 no.5
• /
• pp.125-130
• /
• 2014

The competitiveness in today's global market forces many companies to develop families of products to provide enough variety for the marketplace. The challenge when designing a product family is in resolving the tradeoff between product commonality and distinctiveness. Simultaneously it is necessary to consider environmental performance to design a product family as well as to shorten lead-times, improve quality and reduce costs. This paper proposes a metaheuristic algorithm based redesign methodology for green product family considering environmental performance. The proposed method uses a genetic algorithm as metaheuristic algorithm and green product family index (GPFI) to support green product family design. In addition, it provides the redesign methodology such as product family level and component level. A case study used table lamps as an product family's example shows the verification and effectiveness of the proposed method.