Real-time Tour Planning For Flexible Route Bus Considering Passengers’ Waiting Behavior

Flexible route bus serves passengers who are located in low-density district, and is one part of demand response transit mode. As one of the typical passenger service systems, the principle of flexible route bus is people-oriented. If the tour planning is conducted pursuing operational efficiency, ignoring passenger’s waiting psychology and behavior; the scheduling would violate the people-oriented principle. Passengers and their behaviors are vital; passengers’bounded rationality would direct the tour planning. Thus, real-time tour planning for flexible route bus considering passengers’waiting behavior is of both practical and theoretical significance. Real-time Tour Planning for flexible route bus itself is NP-hard problem; it would make it harder if considering passengers’waiting psychology and behavior. We apply operation research, waiting psychology, behavior science, and intelligent algorithm to investigate this problem. The research plan includes passengers’waiting behavior, scheduling strategies, response strategies, model generation and intelligent algorithm. Overall, the following four aspects are conducted in details.(1) The analysis of real-time tour planning for flexible route bus and passengers’waiting behavior. First, the characteristics of real-time tour planning for flexible route bus were analyzed. Second, real-time scheduling strategy and response strategy were proposed. Then were the related theories about waiting behavior. All of those are provided as guideline for the model generation.(2) The model generation of real-time flexible tour scheduling considering passengers’ waiting behavior. Waiting behavior concerns multi aspects, including reference point, loss aversion, group waiting, fairness, social injustice, etc. Those elements were quantized and embedded into the real-time flexible tour scheduling model with the objective of maximizing passenger satisfaction.(3) The solving method for the new model. The generated model is of multi-objectives and high dynamism degree. Accordingly, the problem is divided into sequential decision problems, then NSGA-Ⅱ was designed for each periodical decision making.(4) Numerical experiment analysis. The experiment is designed to test the scheduling strategy and response strategy. And the parameters concerning waiting behavior is tested by sensitivity analysis. According to the analysis results, the managerial suggestions were provided in order to improve passenger satisfaction. In order to attain high passenger satisfaction, the development and evaluation of the real-time flexible tour planning considering passengers’ waiting behavior is the main focus of this paper. Specifically, the coordination and fusion of behavior theory, psychology theory, operation research and intelligent algorithm were exploited. The proposed models, approaches and methods for tour planning can be characterized by the following contributions. Decision support was provided for managers facing real-time flexible tour planning, and is of great practical significance for improving passenger satisfaction. Moreover, this research is the extension of multi-objective and dynamic optimization problem.