Finite size and dimensional dependence in the Euclidean traveling salesman problem

被引:68
作者
Percus, AG
Martin, OC
机构
[1] Division de Physique Théorique, Institut de Physique Nucléaire, Université Paris-Sud, Orsay
关键词
D O I
10.1103/PhysRevLett.76.1188
中图分类号
O4 [物理学];
学科分类号
0702 ;
摘要
We consider the Euclidean traveling salesman problem for N cities randomly distributed in the unit d-dimensional hypercube, and investigate the finite size scaling of the mean optimal tour length L(E). With toroidal boundary conditions we find, movtivated by a remarkable universality in the kth nearest neighbor distribution, that L(E)(d = 2) = (0.7120 +/- 0.0002) N-1/2 [1 O +(1/N)] and L(E) (d = 3) = 0.6979 +/- 0.0002) N-2/3 [1 + O(1/N)]. We then consider a mean-field approach in the limit N --> infinity which we find to be a good approximation (the error being less than 2.1 % at d = 1, 2, and 3), and which suggests that L(E)(d) = N-1-1/d root d/2 pi e (pi d)(1/2d)[1 + O(1/d)] at large d.
引用
收藏
页码:1188 / 1191
页数:4
相关论文
共 13 条
[1]  
[Anonymous], P CAMB PHILO SOC, DOI DOI 10.1017/S0305004100034095
[2]  
BERSIMAS DJ, 1990, OPER RES LETT, V9, P223
[3]  
CERF NJ, IN PRESS
[4]  
JOHNSON DS, IN PRESS 7TH P ANN A
[5]  
JOHNSON DS, 1990, 17TH P C AUT LANG PR, P446
[6]   THE CAVITY METHOD AND THE TRAVELING-SALESMAN PROBLEM [J].
KRAUTH, W ;
MEZARD, M .
EUROPHYSICS LETTERS, 1989, 8 (03) :213-218
[7]   EFFECTIVE HEURISTIC ALGORITHM FOR TRAVELING-SALESMAN PROBLEM [J].
LIN, S ;
KERNIGHAN, BW .
OPERATIONS RESEARCH, 1973, 21 (02) :498-516
[8]   LARGE-STEP MARKOV-CHAINS FOR THE TSP INCORPORATING LOCAL SEARCH HEURISTICS [J].
MARTIN, O ;
OTTO, SW ;
FELTEN, EW .
OPERATIONS RESEARCH LETTERS, 1992, 11 (04) :219-224
[9]   MEAN-FIELD EQUATIONS FOR THE MATCHING AND THE TRAVELING SALESMAN PROBLEMS [J].
MEZARD, M ;
PARISI, G .
EUROPHYSICS LETTERS, 1986, 2 (12) :913-918
[10]  
Mezard M., 1987, Spin glass theory and beyond: An Introduction to the Replica Method and its Applications, Vvol 9