DOA estimation of unknown frequency for vector hydrophone array based on baffle

When the rectangular air cavity sonar baffle is applied to the direction estimation of vector hydrophone uniform linear array, the performance of the direction estimation algorithm based on the reflection coefficient of the baffle is degraded due to the unknown reflection coefficient, In order to reduce the influence of the reflection coefficient of the baffle on the azimuth estimation results, the training set and the test set are obtained under the condition of considering the vibration and sound structure coupling of the steel plate layer of the rectangular air cavity baffle, Through the training and testing of the network, it is proved that the machine learning algorithm can solve the azimuth estimation problem of the vector hydrophone array based on the baffle when the frequency of the incoming wave is unknown, and obtain better performance than the algorithm based on the reflection coefficient of the baffle.