When the temperature of the submarine cable rises abnormally, if it is not repaired in time, it will develop into a thermal breakdown accident, which will affect the safety of production and even cause significant economic losses. It is of practical engineering significance and economic value to study the real-time temperature data of the submarine cable and identify and diagnose the abnormal temperature rise. In this paper, an abnormal temperature rises detection method based on eigenvalue amplification is designed, which can amplify the abnormal values and make the characteristics of the abnormal values more obvious. This scheme realizes the early detection of abnormal temperature rise of submarine cable, and has high accuracy of temperature difference identification. The abnormal temperature monitoring and simulation of submarine cable based on exponential characteristics can effectively protect the safety of submarine cable and meet the accuracy and real-time requirements of submarine cable monitoring.