论文部分内容阅读
本系列文章的工作是在舰船噪声谱图的基础上，利用模糊神经网络对舰船进行分类识别。文章Ⅰ叙述了舰船噪声的谱特征由可分离的平稳谱和非平稳谱组成，介绍了既利用有效识别特征（有类别共性和异性的特征）又对特定舰船特征作专门记忆的工作路线及识别框架，对特定舰船的记忆具体体现在特定舰船特征模库一包含有线谱模板库、双重频率谱模板库和平均功率谱模板库。文章Ⅰ又具体讨论了特征提取和建立线谱模板时所碰到的理论、模型、分析参数及线谱提取方法，舰船噪声实际情况和理论之间的差异等问题。文章最后介绍了用机器自动提取线谱的一种方法。系列文章Ⅱ将讨论线谱稳定性、唯一性和线谱模板图；文章Ⅲ将讨论双重谱和平均功率谱的特征提取和模板建立；文章Ⅳ将讨论模糊神经网络和识别。
The work of this series of articles is based on the ship noise spectrum, the use of fuzzy neural network classification of the ship classification. Article I describes that the spectral characteristics of ship noise are composed of separable smoothed and non-smoothed spectra. A working route is described that utilizes both effective identification of features (common and heterosexual features of classes) and specialized memory of specific ship features And identification framework, the memory of a particular ship is embodied in a ship-specific signature database, which includes a library of spectral spectral templates, a database of dual spectral spectral templates and an average spectral template library. Article I discussed in detail the theory, model, analysis parameters and line spectrum extraction methods encountered in the feature extraction and the establishment of line spectrum templates, the actual problems of ship noise and theoretical differences. The article concludes with a method of automatically extracting line spectra using a machine. Series II will discuss line spectrum stability, uniqueness, and line-spectrum template maps; Section III will discuss feature extraction and template creation for dual-spectrum and average power spectra; and Section IV discusses fuzzy neural networks and recognition.