Customer-oriented multi-objective optimization on a novel collaborative multi-heterogeneous-depot electric vehicle routing problem with mixed time windows

Through combination of four types of customer size, three types of depot quantity and three types of battery swapping station quantity, 36 benchmark instances were generated. Similar to the traditional benchmark instances, one depot was randomly generated near position (50, 50) in a 100 × 100 grid, and the other depots were randomly generated near positions (25, 50), (75, 50), (50, 75) or (50, 25). Five types of products were generated that can be stored in multiple heterogeneous depots. Three of the five types of products were selected randomly and stored in each depot. The sizes of customers were set as 40, 80, 120 or 160 in these instances and the locations of customers were randomly scattered. The numbers of depots were set as 3, 4, or 5. The demand type for each customer was randomly selected from one of the five product types, and the demand quantity was randomly selected from 5, 10 or 15. The time window types were randomly selected from the hard or soft time window. The numbers of battery swapping stations were set as 2, 4 or 6. The locations of the battery swapping stations were randomly generated near the positions (25, 25), (75, 75), (25, 75) or (75, 25). Each instance is named in the form of “number of customers_number of depots_number of battery swapping stations”. For example, the instance “40_4_6” implies that it involves 40 customers, 4 depots and 6 battery swapping stations.