Analysis of bridge-traffic system using agent-based cellular automaton traffic model

In the field of bridges' health monitoring, traffic loading's record is a significant part since they have considerable impact on bridges' safety and durability. To rationally estimate the traffic loading's impact, the track of each vehicle passing on the bridge is desired. The health monitoring system generally uses weigh-in- motion (WIM) system to record the vehicle at entrance and exit of the bridge. There is information gap of traffic flow moving on the bridge except the position installed WIM. One option to solve this problem is to use a reasonable traffic flow model to supplement the vacancy information. In this paper, a new six- lane agent-based cellular automaton (Agent-CA) traffic model for bridge is proposed. This model is developed based on the cellular automaton (CA) traffic model, a commonly-used micro-simulation traffic model for long-span bridges in recent years. Compared with CA traffic model, Agent-CA traffic model can consider the profile and driving behavior's differences among various vehicle types. Thus, the Agent-CA traffic model is more close to the real traffic situation. Through parametric analysis of a typical bridge traffic system, it is found that the traffic flow's simulation result between CA and Agent-CA model cannot be neglected. Subsequently, the dynamic response's characteristics of the bridge under various traffic situations are summarized.