Abnormal event early warning method based on rail car operation event association model

The invention discloses an abnormal event early warning method based on a rail car operation event association model, and the method comprises the following steps: finding out a road section where anabnormal event frequently happens through the historical event data of the operation of a rail car; establishing an abnormal event early warning model based on event related rules for different road sections; and combining the early warning model with the real-time event data, predicting the occurrence probability of the abnormal event, and sending early warning information to the driver online. Compared with the prior art, the method has the advantages that the early warning model is inputted on the basis of the rail car operation event data, related factors including driver operation, equipment states and locomotive signals are comprehensive, and comprehensive early warning can be conducted; different early warning models are used for different abnormal event frequent occurrence road sections, and the problem th