Research on route planning of connecting demand response bus based on genetic algorithm

In order to meet the travel needs of passengers from residential areas far away from subway stations to subway stations, a time-windowed first-mile feeder DRT route optimization method has been studied. Firstly, based on k-means clustering algorithm, the passenger travel coordinates are clustered to obtain DRT stations. Secondly, a demand-response bus model with the minimum total cost of the system as the objective function, passenger time window, rated passenger capacity of the vehicle, and constraints on boarding and unloading stations is constructed. Then, an improved genetic algorithm for customized cross operation is designed. Finally, an example is given to verify the effectiveness of the model and the algorithm. The results show that the optimized model can output a reasonable multi-DRT route vehicle scheduling scheme, which not only improves the attractiveness and operation efficiency of the bus system, but also satisfies the diversified travel needs of passengers.