Abstract: In order to improve the level-of-service of common bus line operation under real-time information, this paper proposed an optimization model of common line departure schedule based on real-time information. Firstly, it considered the change of passenger flow distribution caused by the dynamic change of passenger travel behavior under real-time information based on the time-varying passenger flow demand and real-time road condition information, and established the optimization model of departure schedule of common bus lines with the optimization objectives of passenger travel cost and bus enterprise operation cost. Then, it selected the improved particle swarm optimization algorithm (PSO-DT) based on dynamic topology to solve the model. Based on the analysis of two common lines in Dalian city, the optimized departure schedule and vehicle trajectory were given and the difference of pa-ssenger flow distribution with and without real-time information was analyzed. The results show that, as compared with the current scheme, the proposed scheme can reduce the travel cost of passengers by 12.6%, the operating cost of bus companies by 8.3%, the total cost by 12.3%, the number of bus departure by 3, the instantaneous maximum carrying capacity by 4 people, and the standard deviation of carrying capacity of each train by 2.192 people, and increase the average carrying capacity of passengers by 3 people. Therefore, the level-of-service of the optimized line is improved, and the passenger flow distribution among vehicles is more balanced, which verifies the effectiveness of the proposed model. 

Key words: public traffic, real-time information, common bus lines, passenger travel behavior, departure interval, improved particle swarm optimization