Multiple Targets Feature Extraction Based On Linear Array

The task of passive underwater acoustic signal processing includes target detection, location and identification. Feature extraction is the key of underwater target recognition and target power spectrum analysis the most commonly used feature extraction method. However, the insufficiency of the sensor array directionality results in the identified target power spectrum containing interference ingredients. As a result, it affects the reliability of the results of target recognition. In this paper, we proposed a multi-objective interference suppression algorithm based on iterative interference signal reconstruction and linear array. It used passive sonar to receive signal which was passed to a bandpass filter and purified to obtain the interference signal with high signal to noise ratio. The interference component was estimated by utilizing array structure and interference signal orientation. We removed the interference component from the time-domain signal and get it purified again to obtain a more pure target signal. Finally, the desired signal line spectrum feature was correctly extracted by making use of LOFAR spectrum and DEMON spectrum feature extraction method. The main research work of this paper is as follows.1. We discussed the different multi-target interference impact on the beamforming output signal and the subsequent feature extraction and analyzed the extracted feature optimization direction under multi-target interference by combination of linear sensor array directivity function.2. The array element signal and the radiated noise of the ship were simulated according to their statistical characteristics respectively under the consideration of the actual engineering situation. The beamforming output signal was analyzed by the LOFAR spectrum and DEMON spectrum to extract the line spectrum of the mechanical noise and the shaft and leaves frequency of the propeller. The performance of the extracted feature of the single signal was compared with that of the presence of multiple target interference by utilizing the above method.3. We discussed the anti-interference performance of the array output signal under the different circumstances of zero guide beamforming and MVDR guide beamforming. A multi-target interference suppression algorithm was proposed based on iterative interference signal reconstruction. After the multi-target interference suppression, the target feature extraction performance was simulated and analyzed.