摘要
It is becoming increasingly difficult for Chinese citizens to access traditional public transport because of overcrowded community structures. Therefore, novel ideas are required to improve the transport system. In this respect, this study considers the design of a public transport scheduling model for a micro system. The model aims to minimize passenger waiting time and maximize number of passengers one bus carries, by simultaneously optimizing departure intervals and use of traditional and rapid buses. The model is superior to traditional models, as it analyzes the phenomena of vehicle overtaking, vehicle capacity limit, and passenger determination uncertainty. In addition, the model is a sophisticated nonlinear multi-objective optimization problem and contains more than one type of decision variable, therefore two composite algorithms, HPSO and GAPSO, are proposed, which are improvements of the Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). These two algorithms are compared to the classical GA with respect to stability and effect, and the results show them to be strong in both respects. In addition, the simultaneous optimization method has evident advantages compared to single-method optimizations.
It is becoming increasingly difficult for Chinese citizens to access traditional public transport because of overcrowded community structures. Therefore, novel ideas are required to improve the transport system. In this respect, this study considers the design of a public transport scheduling model for a micro system. The model aims to minimize passenger waiting time and maximize number of passengers one bus carries, by simultaneously optimizing departure intervals and use of traditional and rapid buses. The model is superior to traditional models, as it analyzes the phenomena of vehicle overtaking, vehicle capacity limit, and passenger determination uncertainty. In addition, the model is a sophisticated nonlinear multi-objective optimization problem and contains more than one type of decision variable, therefore two composite algorithms, HPSO and GAPSO, are proposed, which are improvements of the Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). These two algorithms are compared to the classical GA with respect to stability and effect, and the results show them to be strong in both respects. In addition, the simultaneous optimization method has evident advantages compared to single-method optimizations.
基金
supported by the National Key Basic Research and Development Program in China(No.2016YFB0100906)
the National Key Technology Research and Development Program(No.2014BAG03B01)
the National Natural Science Foundation of China(Nos.61273238 and 61673232)
Beijing Manicipal Science and Technology Program(No.D15110900280000)