舰船辐射噪声信号弱调制线谱检测与估计研究

The ship noise contains information of the ship's working conditions and propeller structure, which is an important basis for target detection and classification of passive sonar. Recently, the power and the modulation characteristics of the ship noise decrease with the improvement of vibration reduction and noise reduction technology. These factors lead to the increase of the difficulty of the conventional DEMON processing to extract the characteristic of the wideband modulation of the ship noise. In this study, the mechanism and experimental data of the ship noise will be studied, and a fine signal model of the ship noise will be designed, and the reasonable distribution range of model parameters will be given. Based on the fine model, the quantitative relationship between the broadband modulation line spectrum extraction performance and signal parameters, demodulation processing method, processing parameters will be studied, and the processing method and processing parameters selection criteria will be given. Through the study of coupling mechanism between modulation line spectrum detection and frequency estimation, a improved demon process method will be given to improve the modulation line spectrum extraction performance on weak signal modulation and nonstationary modulated signal. The proposed approach will be verified by practical measured data. Through the research, we aims to provide a method for the wideband modulation feature extraction of the ship noise, which can provide the support for the accuracy and reliability of the wideband demodulation feature extraction for the actual weak underwater acoustic target.

舰船辐射噪声的调制特征蕴含了舰船工况及螺旋桨结构等信息，是被动声纳目标检测和分类识别的重要依据。减震降噪技术的不断完善使得舰船辐射噪声功率和调制特征明显减弱，常规的噪声调制包络检测（DEMON）处理难以有效提取目标噪声弱调制特征。本课题拟结合水中螺旋桨辐射噪声的产生机理及水洞实验数据，设计精细的舰船辐射噪声信号模型，给出合理的模型参数分布范围。基于该精细模型开展推导和仿真分析，建立信号质量、解调方法和处理参数等因素与宽带调制线谱提取性能的量化关系，给出解调方法和处理参数的选取准则。通过研究DEMON分析时线谱检测和频率估计的耦合机制，设计数据自适应的DEMON处理流程，提高对弱调制信号和目标机动状态下的短时非平稳调制信号的调制线谱提取性能，并开展试验验证。通过本课题的研究，力求提供一种精细的螺旋桨辐射噪声宽带调制线谱提取方法，为实际水声目标弱调制特征提取的准确性和可靠性提供支持。

舰船辐射噪声的调制特征蕴含了舰船工况及螺旋桨结构等信息，是被动声纳目标检测和分类识别的重要依据。减震降噪技术的不断完善使得舰船辐射噪声功率和调制特征明显减弱，常规的噪声调制包络检测处理难以有效提取目标噪声弱调制特征。在此背景下，本课题针对性开展了研究。. 首先，开展了螺旋桨噪声数据收集和采集工作，根据噪声的发声机理和信号统计特性，结合辐射噪声数据，设计了一种较为精细的舰船辐射噪声模型。. 其次，开展了水中螺旋桨辐射噪声信号参数、解调制处理方法和处理参数等因素与调制线谱提取性能的相关性研究，分析推导了信号参数、特征提取方法与特征提取性能之间的量化关系。. 然后，开展了解调制处理方法和处理参数的选取准则研究，基于调制特征提取的影响设计了频段权重设置准则，提出了一种调制频段补偿的宽带解调方法，提高了对调制特征的提取性能。. 最后，针对时频图中的非平稳时变线谱和多线谱检测问题，针对时频图中弱线谱检测问题和多线谱检测问题，提出了一种基于时频图的多线谱检测方法，获得较好的线谱轨迹提取效果，并开发了一套调制特征提取软件。