A Degradation Prediction Algorithm for Maritime Distress Reporting Based on Deep Learning

Owing to the bad weather, equipment immersion with water, antenna difficult to point at the satellite and so on, the accuracy and the reliability of the positional information of the pilot in distress sent by the distress message signal sending device is low, which will be reduced with the increase of working time. In order to improve the reliability of distress message signal sending device based on BeiDou satellite, a prediction method for signal sending time and a prediction method for signal transmitting delay time are firstly proposed based on the deep neural network. In the process of prediction, a lot of sensor information is used, especial in the prediction of signal transmitting delay time, multiple-sampling information from the sensors is adopted. The experimental results show that the probability of successful message signal sending is increased from 36.3% to 73.3%, moreover, the working time of the equipment was extended from 6.0 hours to 8.6 hours.