During the operation of transmission lines, there are sudden failures and a large number of slowdeveloping, preventable “gradual” failures, which have seriously threatened the safe and stable operation of the transmission system. Based on analyzing the multidimensional environmental factors affecting line safety, the study proposes a method for identifying the operating state of transmission lines based on the AdaBoost integrated learning algorithm, and develops a set of transmission line hidden danger monitoring system. A decision pile based on Ginin indicators is used as a weak classifier, and the hidden danger monitoring results and their confidence levels are output by training and weighted summation of multiple weak classifiers. Using historical data for validation experiments, the proposed method achieves an accuracy of 95.92% in recognizing the operating state of transmission lines, which is a more superior performance compared with traditional machine learning methods. The system can basically realize the hidden danger monitoring of transmission lines, so as to assist the field operation and maintenance personnel of transmission lines to carry out fault investigation, and reduce the transmission line tripping due to the development of hidden danger into fault.