IPv6 target address intelligent detection method based on split clustering and entropy structure

The invention discloses an IPv6 target address intelligent detection method based on split clustering and an entropy structure, and relates to the technical field of network address detection. The method comprises the following steps: S1, constructing a tree structure to store an obtained active ipv6 address by utilizing hierarchical split clustering; s2, constructing a model to analyze a clustering result and outputting a prediction address; s3, detecting a predicted address set generated by all leaf nodes by using a zmap scanning tool to obtain an active address; s4, comparing and analyzing the predicted address activity percentages of different output scales by using a broken line graph; and S5, finally constructing the above work into a complete system, and continuously updating and maintaining a data set. According to the method, the active seed addresses are preprocessed and classified by using hierarchical split clustering and tree structure construction, the subdivision degree of the seed set is impro