Abstract:

The traditional method does not conduct the training and learning of the characteristics of the transmission information when conducting the transmission information security defense, which leads to the unsatisfactory effect of the information security defense. To this end, non-scanning vulnerability detection technology is introduced to optimize transmission information security defense. In a specific time, the execution program obtains the program execution event sequence, completes the dynamic feature extraction of the transmission information based on information aggregation; uses the TF-IDF algorithm to realize the feature vectorization of the dynamic sample feature set, and the feature quantity is input into the machine learning model through Text-CNN, Logistic, and random forest algorithms train and learn the extracted dynamic features to obtain the vulnerability detection results; for the security problems caused by the detected security vulnerabilities, the security defense, secure storage and security isolation of the line detection terminal are implemented transmission information security defense. Experiments verify that this method accurately detects power grid loopholes and has a good effect on transmission information security defense.

Key words: Non-scanning detection, power transmission information, security holes, TF-IDF algorithm