The recognition and applications of floating cars´ driving behaviors

This paper classified drivers´ driving behaviors into rotation of steering wheel, the stampede of clutch, brake and throttle, the shift of gear and some other primary driving behaviors. Concerning about the needs of practical application and considering the requirements of convenience, real time and accuracy fully, we researched how to measure these primary driving behaviors on the simulated driving platform constructed by ourselves. And then we discussed how to conclude floating cars´ driving behaviors from these primary driving behaviors combining with velocity, location and direction information of floating cars. After that, we applied logical reasoning methods and D-S reasoning theory to describe city road traffic conditions based on recognizing floating cars´ driving behaviors. We developed a cognitive system for floating cars´ driving behaviors on this basis. This cognitive system has a significant impact on forecasting and solving city traffic problems.