TY - JOUR
T1 - Nodal ordering for bandwidth reduction using ant system algorithm
AU - Kaveh, A.
AU - Sharafi, P.
PY - 2009
Y1 - 2009
N2 - Purpose - For the solution of equations with sparse matrices, the problem of bandwidth reduction is an important issue. Though graph theoretical algorithms are available, the purpose of this paper is to examine the feasibility of ant systems (AS). Design/methodology/approach - For band optimization an ant colony algorithm based on AS is utilized. In this algorithm a local search procedure is also included to improve the solution. Findings - AS algorithms are found to be suitable for bandwidth optimization. Originality/value - Application of AS to the bandwidth reduction is the main purpose of this paper, which is successfully performed. The results are compared to those of a graph theoretical bandwidth optimization algorithm.
AB - Purpose - For the solution of equations with sparse matrices, the problem of bandwidth reduction is an important issue. Though graph theoretical algorithms are available, the purpose of this paper is to examine the feasibility of ant systems (AS). Design/methodology/approach - For band optimization an ant colony algorithm based on AS is utilized. In this algorithm a local search procedure is also included to improve the solution. Findings - AS algorithms are found to be suitable for bandwidth optimization. Originality/value - Application of AS to the bandwidth reduction is the main purpose of this paper, which is successfully performed. The results are compared to those of a graph theoretical bandwidth optimization algorithm.
UR - http://handle.uws.edu.au:8081/1959.7/549030
U2 - 10.1108/02644400910943644
DO - 10.1108/02644400910943644
M3 - Article
SN - 0264-4401
VL - 26
SP - 313
EP - 323
JO - Engineering Computations
JF - Engineering Computations
IS - 3
ER -