- Agent Theory And Applied Research In The Scheduling Of Train Operation

The construction of passenger-dedicated line and the high-speed of train traffic bring new light for train transportation. on the other hand, the high-speed of train cause train traffic course become more complicated, so the hidden trouble security and uncontrolled factor is higher than ever. Compared with the existing line, the passenger-dedicated line brings forward higher demands for the capability of train traffic control system, it demands train traffic control system have more real-time, intelligent and responsible specialize. Train traffic control system is the most important system of the intelligent railway transportation systems and guarantees the train's safety & punctuality. Then it demands new control strategy for high-speed of the train's multi-goal of safety, punctuality, comfortable. To the background of train traffic control, this paper study on the agent structure, multi-agent theory, and train traffic control. The main studying content as follows:(1) This paper studied on multi-agent theory and theory correlated with train traffic control, it indicated train traffic control is a typical distributed problem, while multi-agent is fit for solving distributed problem, and then bring forward the studying content of this paper;(2) This paper studied on the structure of multi-agent system, bring forward train traffic control agent with learning machinery, for converging and rebuilding information which learned from agent from train traffic control system ,it is used for new dispatching, and for re-dispatching. and then improve the intelligence of train traffic control.(3) Emulating on train traffic control system with learning mechanism, this paper validated the high efficiency of train traffic control system with learning mechanism.Through studying on the agent structure and the train traffic control, this paper brings forward agent structure which is suit for train traffic with learning mechanism. by emulating experiment to validate the efficiency. This theory has important theoretical and practical value for improving the intelligent of train traffic control system.