A method for modeling drivers´ behavior rules in agent-based traffic simulation

Drivers´ judgment plays an important role in the choice of route. In traditional agent-based models, route planning is based on optimal path algorithm for length, time or combination of them. But drivers choose route not only depends on assistant information but also on his cognition. Based on this consideration, we propose our driver behavior rules to make the route plan process similar to driver´s decision process. The behavior models are composed of two parts. The first part is the memory model. The familiarity of route is reflected by driver´s memory. The driver can remember the traffic density of the route that he had run on. After run on the same route several times, the driver have known the probable traffic density of each route at different times. Due to limited memory, the driver mainly remembers routes that he often runs on and those main routes. The second part is the information utility model. The effectiveness of information is reflected by the attitude of driver to follow the instructions of the information. This research aims at construction behavior rules for agent-based driver model. We developed a prototype system to do simulation experiment. The result shows the interesting difference with traditional models.