Abnormal driving behavior detection based on an improved ant colony algorithm

As one of the most serious hazards in world, more than 80% traffic accidents are caused by driver misconduct. The detection abnormal behavior drivers is important to improve safety public transportation. anomaly measurement not only determined objective rules such as laws, but also distinguished due biological characteristics. same driving may present completely opposite judgment results for different categories drivers. In this paper, we propose a novel method that measures preference path length various operations via pheromones, and identifies calculating cumulative conversion probability operation switching. An improved ant colony algorithm based on fixed point simplicial theory proposed convergence efficiency optimizing initial population state. Experimental show can effectively detect significantly reduce false alarms.