Train Scheduling Optimization for an Urban Rail Transit Line: A Simulated-Annealing Algorithm Using a Large Neighborhood Search Metaheuristic

Table 4

Optimization results with 2 : 1 train operation proportion of full-length and short-turn routings.

(s)	Number of trains	Number of trains required		Max. loading rate (%)		Average loading rate (%)		(s)
		Full-length routing	Short-turn routing
170	40	30	10	119.7	86.6	76.9	57.4	6.45
165	41	31	10	116.2	84.1	74.6	55.6	6.14
160	42	32	10	112.5	81.2	72.2	54.2	5.86
155	44	33	11	109.1	78.9	70.1	52.3	5.61
150	45	34	11	105.2	76.4	67.9	50.4	5.35
145	47	35	12	102.1	73.6	65.2	48.9	5.12
140	48	36	12	98.6	71.3	63.2	47.3	4.83
135	50	38	12	95.3	68.8	60.1	45.6	4.64
130	52	39	13	91.5	66.1	57.8	43.8	4.33
125	54	41	13	88.1	63.7	55.2	42.1	4.04
120	56	42	14	84.4	61.1	53.7	40.5	3.82