Dual detection technology for abnormal state of vehicle-mounted CANFD network

The invention belongs to the technical field of intelligent driving, and particularly relates to a dual detection method for an abnormal state of a vehicle-mounted CANFD network, which comprises the following steps of: 1, performing CANFD message identification level abnormity monitoring based on relative entropy and active entropy, calculating an entropy value in a certain time period in real time in the stage, comparing the entropy value with a standard entropy value range, identifying whether a message is normal or abnormal, and reporting an error or early warning; step 2, CANFD message content level anomaly detection based on a decision tree, in the stage, the decision tree is constructed for different content attributes of different CANFD messages, the messages are sequentially put into a decision tree model to be checked, and whether the message content is abnormal or not is judged; according to the method, the structure is reasonable, CANFD message identification level abnormity is detected through the