Considering the close relationship between rearend accident and the vehicle’s travelling state in highway tunnel group,this paper takes braking distance,travelling speed and the type of vehicle as characteristic parameters for forewarning the rearend accidents occurring in highway tunnel group.Then,on the basis of artificial immune mechanism,an earlywarning model is proposed to identify and determine the tendency caused by vehicle’s abnormal state,and some important earlywarning information is given as the levels of accidents.Finally,Xihan highway tunnel group is studied as a simulation example to verify the accuracy and utility of the earlywarning model.It is expected that the findings from the study can provide a significant theoretical basis for the safety operation management in highway tunnel group.