Abdul Samad Shibghatullah
96
30052008
4.5.1 Experiment Environment
A PC with Pentium IV 1.2 GHz and Windows XP operating system was used for the experiments.
4.6 Overview of Experimental Results
This section presents the sample of the results of the experiments for all the events as shown in Tables 4.23 to 4.28 LFSO Table 4.23, LFR Table 4.24, LFSW Table
4.25, DFSO Table 4.26, DFSW Table 4.27 and UNV Table 4.28. The full results are shown in 54 tables in Appendix F.
As shown in Tables 4.23 to 4.27, lateness illustrates how many minutes the crew is late, late-crew ready time shows the time when the late-crew is ready for work, time shows
how fast in seconds the system performs the rescheduling process, round shows how
many rescheduling rounds take place to find a match, and rescheduling shows which
crew is assigned to which duty. For example, in the case of rescheduling result S, V, Z, AB, crew S which the first from left is the late-crew and it takes the duty of AB
which is the first from right. Crew V takes crew S’s duty, crew Z takes crew V’s duty, and crew AB takes crew Z’s duty. Minutes late shows the late minutes due to
adjustments to find a match, and crew involved shows how many crews are affected by
the rescheduling. For example, Table 4.23 shows the result for the event of LFSO for large schedule type
and maximum distribution. As an example, in the first row, the crew is late by 15 minutes; the late-crew is only ready for work at 6:20; the rescheduling process takes
1.873 seconds and only one round of matching and the result is that crew S which is the late-crew takes crew V’s duty, and crew V takes crew S’s duty. In this process, two
crews are affected. For the 60 minutes late example as shown in the last row, the crew is late for 60 minutes, the late-crew is only ready for work at 7:05, and the rescheduling
process takes 5.544 seconds and three rounds of matching are needed. The rescheduling results is that crew S which is the late-crew takes crew AF’s duty, crew V takes crew
S’s duty, crew Z takes crew V’s duty, crew AB takes crew Z’s duty, crew AC takes crew AB’s duty, crew AD take crew AC’s duty, crew AE takes crew AD’s duty, and
Abdul Samad Shibghatullah
97
30052008
crew AF takes crew AE’s duty. In this process, there is 1 minute lateness and 8 crews are affected.
Table 4.23: LFSO LargeMaximum
Lateness Minutes
Late-Crew Ready Time
Time s Round Rescheduling Crew ID
Minutes Late
Crew Involved
15 06:20:00 1.873
1 S,
V 2
20 06:25:00 1.912
1 S,
V 2
25 06:30:00
3.756 2
S, V, Z 3
30 06:35:00
1.943 1
S, V, Z 3
35 06:40:00 3.866
2 S,
V,Z,AB 4
40 06:45:00 3.764
2 S,
V,Z,AB,AC 5
45 06:50:00 1.982
1 S,
V,Z,AB,AC 5
50 06:55:00 3.655
2 S,
V,Z,AB,AC,AD 6
55 07:00:00 3.806
2 S,
V,Z,AB,AC,AD, AE
7 60 07:05:00
5.544 3
S, V,Z,AB,AC,AD,
AE, AF
1 8
Table 4.24: LFR LargeMaximum
Lateness Minutes
Late-Crew Ready Time
Time s Round Rescheduling Crew ID
Minutes Late
Crew Involved
15 12:12:00 1.873
1 X,
T 2
20 12:17:00 2.123
1 X,
T 2
25 12:22:00 1.963
1 X,
T 2
30 12:27:00 1.963
1 X,
T 2
35 12:32:00 1.933
1 X,
T 2
40 12:37:00 1.963
1 X,
T 2
45 12:42:00 1.903
1 X,
T 2
50 12:47:00 1.862
1 X,
T 2
55 12:52:00 1.983
1 X,
T 2
60 12:57:00 1.843
1 X,
T 2
Table 4.25: LDR LargeMaximum
Lateness Minutes
Late-Crew Ready Time
Time s Round Rescheduling Crew ID
Minutes Late
Crew Involved
15 13:23:00 1.903
1 AD,
BW 2
20 13:28:00 1.872
1 AD,
BW 2
25 13:33:00 1.923
1 AD,
BW 2
30 13:38:00 1.973
1 AD,
BW 2
35 13:43:00 1.993
1 AD,
AF 2
40 13:48:00 1.923
1 AD,
AF 2
45 13:53:00 1.882
1 AD,
AI 2
50 13:58:00 1.882
1 AD,
AI 2
55 14:03:00 1.923
1 AD,
AI 2
60 14:08:00 1.993
1 AD,
AI 2
Abdul Samad Shibghatullah
98
30052008
Table 4.26: DFSOLargeMaximum
Lateness Minutes
Late-Crew Ready Time
Time s Round Rescheduling Crew ID
Minutes Late
Crew Involved
80 07:25:00
1.873 1
S, V,Z,AB,AC,AD, AE, AF, AG 8
9 100
07:45:00 -
- No Match
- -
120 08:05:00
- -
No Match -
- 140
08:25:00 -
- No Match
- -
160 08:45:00
- -
No Match -
- 180
09:05:00 -
- No Match
- -
Table 4.27: DFSW LargeMaximum
Lateness Minutes
Late-Crew Ready Time
Time s Round Rescheduling Crew ID
Minutes Late
Crew Involved
80 14:28:00 1.803
1 AD,
AI 2
100 14:48:00 1.813 1
AD, AI
2 120 15:08:00
1.773 1 AD,
AI 2
140 15:28:00 3.555 2
AD, AI,
AO 3
160 15:48:00
- -
No Match -
- 180
16:08:00 -
- No Match
- -
Table 4.28: UNV LargeMaximum
Unavailable Minutes Start Time
End Time Rescheduling
Time s Minutes Late
30 13:15:00 13:45
I 1.752
30 13:30:00 14:00
W 1.802
30 13:45:00 14:15
W 1.723
60 13:15:00 14:15
I 1.843
60 13:30:00 14:30
W 1.943
60 13:45:00 14:45
W 1.833
90 13:15:00 14:45
I 1.873
90 13:30:00 15:00
W 1.853
90 13:45:00 15:15
W 1.813
120 13:15:00 15:15
I 1.813 0 120 13:30:00
15:30 W
1.792 0 120 13:45:00
15:45 W
1.812 0
In the event of UNV see Table 4.28, unavailable shows in minutes how long the
uncovered duty is because of the unavailability of a crew; start time shows the start time
of the uncovered duty; end time shows the end time of the uncovered duty; rescheduling
shows which crew is assigned to the unattended-duty; time shows how fast in seconds
the system performs the rescheduling process and minutes late shows the late minutes
due to adjustments to find a match.
Abdul Samad Shibghatullah
99
30052008
As an example, Table 4.28 shows the results of UNV in large schedule type and maximum distribution. It starts from 30 minutes then followed by 60 minutes, 90
minutes and 120 minutes, and for every minute of UNV three different start and end times 15, 30, and 45 past the hour. For example, in the case of 30 minutes
unavailability the start time is 13:15; the end time is 13:45; crew I is reassigned to this uncovered duty; it takes 1.752 seconds for the rescheduling process and is 0 minute late.
In the case of 120 minutes UNV, the start time is 13:14; the end time is 15:45; crew W is reassigned to this uncovered duty; it takes 1.812 seconds for the reassignment process
and is 0 minute late. For the full results refer to Appendix F. The next section presents the analysis of the
results in detail according to every event.
4.7 Results Analysis