Modeling of Mountain Highway Event Warning Based on Multi-agent System

Because of the alignment, slope, road conditions and other factors, the incident rate in Mountain highway was significantly higher than other roads. Traditional Events Early Warning Systems, usually used a "multi-bank" means in the structure of the system, was difficult to apply to the complex mountain highway system. Firstly, this paper established mountain road event early warning systems\´ framework based on Multi Agent System(MAS) theory; Secondly, this paper used Visual C++ to program structure of the main agent in the system, and thus the structure of MAS organizations was finished. At last, the paper created new models which could well apply in the mountain highway incident early warning based on MAS. Model validation results showed that both accuracy and calculation speed of the new models was much better than the traditional models. So the study of this paper enhanced the intelligent level of mountain traffic management and reduced traffic accidents.