A Hyperheuristic for the Dial-a-Ride Problem with Time Windows

Table 3

Results for DARPTW obtained by using the hyperheuristic model with the greedy heuristic selection operator and the random permutation descent operator.

	HH with greedy					HH with random permutation descent
	Route duration		Ride time		CPU time	Route duration		Ride time		CPU time
	Avg.	Best	Avg.	Best	min	Avg.	Best	Avg.	Best	min
pr01	1011.28	848.12	294.06	222.63	1.08	996.14	875.03	293.83	225.29	1.15
pr02	2061.82	2012.68	624.42	454.18	1.48	2044.30	1813.17	668.81	545.79	1.55
pr03	2779.83	2673.89	1112.08	967.50	1.93	2825.89	2669.76	1234.97	1044.24	2.00
pr05	4265.13	4007.28	1592.12	1396.41	3.00	4392.61	4212.04	2004.69	1566.09	3.09
pr11	917.10	794.14	221.50	174.97	1.07	921.48	787.68	226.73	151.52	1.14
pr12	1550.91	1475.81	368.66	294.49	1.49	1619.21	1475.81	399.42	268.65	1.56
pr15	4129.36	3937.46	1131.83	938.83	3.02	4338.53	3937.46	1219.13	1027.10	3.05
pr16	4979.85	4708.17	1618.35	1416.12	3.71	5108.65	4708.17	1892.12	1497.70	3.78
pr17	1237.40	1146.47	389.75	257.48	1.25	1263.79	1146.47	353.18	276.71	1.33
pr19	3626.02	3518.29	1804.54	1406.06	2.80	3677.65	3518.29	2266.98	1809.17	2.88