论文部分内容阅读
研究了基于小波域多尺度树结构的水下噪声信号统计概率模型的建模方法。实验分析表明,船舶辐射噪声的模型特征不仪与海洋环境噪声的有很大差异,而且在不同工况下自身也表现出了不同特点。根据海洋环境噪声和船舶辐射噪声在模型参数上的特征差异,提出了一种从海洋环境噪声中检测船舶辐射噪声的方法,实验证明了该方法的检测性能大大优于以前提出的几种检测方法。另外,为了更好地解决船舶辐射噪声的分类问题,在研究了基于隐马尔可夫统计模型的分类方法的基础上,还提出结合支撑向量机和隐马尔可夫模型的综合分类方法,实验分析也取得了较好的结果。
The modeling method of statistical probability model of underwater noise signal based on multi-scale tree structure in wavelet domain is studied. Experimental analysis shows that the model noise characteristics of ship radiated noise are greatly different from that of marine environment noise and show different characteristics under different working conditions. According to the characteristic difference of marine environment noise and ship radiation noise on the model parameters, a method of detecting ship radiation noise from marine environmental noise is proposed. The experimental results show that the detection performance of this method is much better than the previous detection methods . In addition, in order to solve the classification problem of ship radiation noise better, based on the classification method based on hidden Markov statistical model, a comprehensive classification method based on support vector machine and hidden Markov model is also proposed. The experimental analysis Also achieved good results.