Bike-Sharing Fleet Allocation Optimization Based on Demand Gap and Cycle Rebalancing Strategies

Table 2

The rental and returning amount of station id 72 from March 1 to 7.

Station id	Date	Rental amount	Returning amount
72	3.1	56	62
72	3.2	34	32
72	3.3	29	32
72	3.4	64	41
72	3.5	68	66
72	3.6	94	116
72	3.7	107	109

Note that, under different parameters, convergence of genetic algorithm is different, as shown in Figure 6. Under the conditions of the crossover rate , mutation rate , a relatively best result can be obtained, which may be used in experiments of this paper.