论文部分内容阅读
The methods of recognizing different types of ship objects according to the shape of average powerspectrum of the radiated noise are discussed in this paper.The pre-processing of the data is con-sidered first.After applying general correlation method,the distance method and weighted distancemethod,a new method i.e.M-m_2 plane partition method,is proposed.The basis of this methodis the shape of averaged power spectrum of ship radiated noise may be approximately described bythe Ece noise model.Such a noise model,which may be characterized by a set of three parametersf_0,f_m and K or w_0,a and K is described in detail.The approaches of extraction these parameters arealso discussed.By using of the recognition features M.which is the location of the maximum value ofaverage power spectrum curve,and m_2,which is the second central moment of normalized averagepower spectrum level and applying a piece-linear classifier,a 92% of correct recognition rateis obtained for two different types of ship objects.Compared w
The methods of recognizing different types of ship objects according to the shape of average powerspectrum of the radiated noise are discussed in this paper. The pre-processing of the data is con-sidered first. After applying general correlation method, the distance method and weighted distancemethod, a new method ieM-m_2 plane partition method, is proposed. The basis of this methodis the shape of averaged power spectrum of ship radiated noise may be described the same by the Ece noise model.Such a noise model, which may be characterized by a set of three parameters f_0, f_m and K or w_0, a and K is described in detail. These approaches of extraction these parameters are also discussed. By using the recognition features M.which is the location of the maximum value ofaverage power spectrum curve, and m_2, which is the second central moment of normalized averagepower spectrum level and applying a piece-linear classifier, a 92% of correct recognition rateis obtained for two different types of ship objects.Compared w