To improve the effectiveness of the long-term structural monitoring of ships in irregular waves, an autonomous signal recognition method is studied using machine learning technology. On the basis of fast Fourier filtering and a backpropagation neural network, the intelligent recognition method can accurately distinguish the time-domain signals at different monitoring positions. Its feasibility is proved by the monitoring data from a model test. In addition, factors such as the multidimensional signal input and learning library size are further analyzed in detail. The fluctuation trends of signal recognition rates are also observed at different frequency thresholds. Finally, some parameters of the signal recognition method for general application scenarios are given.
With the development of large-scale and high-speed ships, the characteristics of hull structural conditions in the complex marine environment have attracted more and more attention from ship designers. Monitoring technology has been gradually developed to observe and study the fluctuation characteristics of the hull structure conditions exactly.