Abstract: To meet the personalized needs of passenger travel in urban industrial agglomeration areas and to alleviate the problem of oversaturation of the road network, a customized bus dispatching model considering parking penalty on congested roads is proposed. First, the operation conditions and scheduling rules are analyzed. Then, the penalty functions for the cost of stopping and picking up passengers as well as violating passenger′s time window are demarcated. A customized bus dispatching decision-making model in response to the real-time demand is established with the goal of optimizing the total cost of the system, including the cost of carrying passenger time, the cost of bus operation time, the cost of violating passenger′s time window penalty and the cost of stopping on congested roads. Improved genetic algorithm and insertion algorithm are designed to solve the problem, and Zhongguancun Software Park is used as an example to verify the effectiveness of the model and algorithm. The results show that the customized bus dispatching model can achieve the expected results with respect to the targets for shift length, full-load ratio, and cost control, thus confirming the feasibility of the model and algorithm.

Key words: customized bus, parking penalty on congested roads, time window constraint, demand response, improved genetic algorithm