Early Warning System Based on Dempster-Shafer Reasoning for Drunk Driving

The general detection of drink driving is only the alcohol content testing,its means is single,and it relies on human self-consciousness.In order to detect and prevent drunk driving effectively,an early warning model of drunk driving built by the Dempster-Shafer evidence reasoning method based on the multi-source information fusion technology was proposed.Many kinds of sensors were used to extract the features of drivers in the model,and the information obtained was analyzed and forecasted,which approached the status of the driver more,and detection results were more accurate.