水声信号的混沌特征参数提取与分类研究

Better Feature Extraction and Classification of Ship-Radiated Noise with Three Chaos Feature Parameters

研究了水声信号的混沌特征参数提取以及利用混沌特征参数的水声信号分类。讨论了水声信号的关联维数、最大Lyapunov指数以及时间序列h2熵等混沌特征参数的计算以及它们在水声信号特征提取、分类中的应用。通过对不同类别、一定样本数量的实测水声数据计算它们的混沌特征参数,验证了水声信号不仅具有混沌特性,而且它们的某些混沌特征参数具有可分性。

The three chaos feature parameters are correlation dimension, largest Lyapunov exponent and h2 entropy. Although each of the above-mentioned feature parameters has been previously utilized in feature extraction and classification, we have not, to our best knowledge, seen in the open literature the utilization of any one of the three parameters in feature extraction and classification of ship-radiated noise.We now present our method and research results of using all three of them in feature extraction and classification of ship-radiated noise. In the full paper, we explain how to use our method in mueh detail; here we just give the results of using our method. We apply our method to the feature extraetion and classification of three sets of ship-radiated noise collected at sea by Chinese organizations other than Northwestern Polytechnical University, corresponding respeetively to noises radiated by three different types of ships. We give three figures. Fig. 1 has three subfigures, giving respectively the variations with time of correlation dimension, largest Lyapunov exponent and hz entropy for the first set of ship-radiated noise that corresponds to the first type of ships. Similarly, Fig. 2 and Fig. 3 each has three subfigures respectively for the second and third sets of ship radiated noise eollected at sea. With these noise subfigures we ean achieve better classification of target ships than achievable with feature parameters other than the three chaos feature parameters.