Research on Shared Bikes Scheduling Optimization Considering Excess Capacity Penalty

As a green and low-carbon mode of travel, shared bikes are an important means of transportation for public transportation connections and “last kilometer” travel, providing users with a more convenient and efficient choice for short and medium distance travel. However, in the daily operation and management process, the time-space mismatch between supply and demand is prominent, especially in the commute peak period, there is an obvious tidal phenomenon, and it is urgent to use dispatch vehicles to achieve the spatio-temporal allocation of shared bike resources. Therefore, in view of the problems existing in some existing shared bike scheduling methods, such as single optimization goal, scheduling point can only be visited once is not considered, and scheduling point capacity is not considered, a multi-objective optimization model for shared bike scheduling is established, which aims at minimum total demand loss and excess capacity penalty. The model considers the situation that the demand for scheduling point in peak hours is much larger than the scheduling vehicle capacity. It allows the scheduling vehicle to repeatedly visit the scheduling point and carry out continuous scheduling of multiple scheduling vehicles. A multi-objective ant colony algorithm is designed to solve the problem. A non-dominated sorting method is introduced to divide the solution set into different non-dominated levels, and the highest level solution is selected. In addition, a max-minimum ant system is introduced to improve the state transition probability rule and pheromone update rule, so that it can be applied to solving multi-objective optimization problems. In order to verify the feasibility of the model and algorithm, a numerical example analysis was carried out, and the results showed that the solution method of the multi-objective ant colony algorithm had good convergence, which realized the overall arrangement of the driving path, loading and unloading quantity, so as to improve the utilization efficiency of shared bike resources and scheduling vehicle resources.