Application of Physics-Informed Neural Networks in Weak Signal Detection

This paper proposes a weak signal detection method based on Physics-Informed Neural Networks(PINN). The method effectively enhances network detection performance in low signal-to-noise ratioenvironments by incorporating physical constraints such as signal amplitude characteristics and spectraldistribution during neural network training. The experiments utilize high-precision sensors for datacollection and combine Wiener filtering for preprocessing. Under -10dB signal-to-noise ratio conditions,the proposed method achieves a detection accuracy significantly outperforming traditional detectionmethods. Experimental results demonstrate that this method exhibits strong robustness and accuracy incomplex noise environments, providing a new solution for weak signal detection.