A Real-Time Train Timetable Rescheduling Method Based on Deep Learning for Metro Systems Energy Optimization under Random Disturbances

Table 4

The rescheduled timetable with the MGA-GRU method in the two-train system.

Section	Spacing (m)	Train no.	Departure instant	Arrival instant	Dwell time (s)	Coasting speed (m/s)
Xujiahui ⟶ Hengshan Road	1458.5	1	8:00:00	8:01:30	29.6	21.8
		2	8:02:00	8:03:35	21.2	18.04
Hengshan Road ⟶ Changshu Road	1125.7	1	8:01:59	8:02:21	41.9	18
		2	8:03:56	8:05:12	22	18
Changshu Road ⟶ South Shaanxi Road	979.2	1	8:04:03	8:05:11	29.6	18.28
		2	8:05:34	8:06:43	20	18.2
South Shaanxi Road ⟶ South Huangpi Road	1377.4	1	8:05:41	8:07:11	29.8	18
		2	8:07:03	8:08:33	20	18.08
South Huangpi Road ⟶ People’s Square	1526.8	1	8:07:41	8:09:13	20	18
		2	8:08:53	8:10:26	21.8	18.04
People’s Square ⟶ Xinzha Road	961.2	1	8:09:33	8:10:39	—	18.56
		2	8:10:48	8:11:55	—	18