A Framing Link Based Tabu Search Algorithm for Large-Scale Multidepot Vehicle Routing Problems
Table 4
Sensitivity data of parameter and its update step .
Solution value
CPU time
% gap 1
% gap 2
% gap 3
% gap 4
p08
249
2
0.85
0.10
11.54
1.51
11.81
29.2
p09
249
3
0.75
0.08
9.82
1.79
19.13
34.48
p10
249
4
0.65
0.04
4.35
1.57
10.83
30.79
p11
249
5
0.65
0.04
17.68
1.2
16.61
51.14
p15
160
4
0.65
0.04
3.24
1.69
7.99
26.85
p16
160
4
0.65
0.06
16.32
1.95
8.68
41.14
p17
160
4
0.65
0.04
9.49
1.43
14.09
45.51
p18
240
6
0.7
0.06
9.59
2.59
12.67
48.92
p19
240
6
0.65
0.04
18.23
1.01
19.57
37.92
p20
240
6
0.65
0.06
13.66
1.99
17.61
44.74
p21
360
9
0.65
0.04
6.95
2.84
4.9
22.24
p22
360
9
0.65
0.04
8.49
1.9
19.68
57.11
p23
360
9
0.65
0.04
10.8
2.55
15.73
21.33
pr04
192
4
0.65
0.04
18.3
2.52
8.75
50.53
pr05
240
4
0.7
0.06
11.59
1.59
14.52
25.42
pr06
288
4
0.75
0.08
12.44
1.13
13.44
35.39
pr09
216
6
0.6
0.04
6.96
1.73
19.96
46.54
pr10
288
6
0.65
0.04
17.1
2.07
12.92
57.18
% Gap 1: % Gap between the worst and the best solution value with different and . % Gap 2: % Gap between the worst and the best solution value with different and . % Gap 3: % Gap between the longest and the shortest CPU time with different and . % Gap 4: % Gap between the longest and the shortest CPU time with different and .
