Nonlinear Dynamical Detection Of The Acoustic Object

As oceanic reverberation is the main interference to acoustic object detection, suppressing the chaos by signal processing is an important task. This dissertation mainly discusses the problem, especially on how to detect static object in a reverberation background. The paper also analyses the methods and results used or got from the study.Firstly, the paper introduces the history of the nonlinear dynamics. Then describe the basis of the nonlinear dynamics theory, among which, the Takens reconstruction theory and Lyapunov exponents are mainly discussed. These analysis shows that the signals being used have the characters of chaos, so the nonlinear methods are proper way to do the procession.Secondly, the locally linear prediction and AR model is introduced for the chaos sequences have the character of local predictability. Then the local projection method is introduced for the smooth property of reconstruction phase space.At last, based on the compare of different results, we can see the effect of various algorisms. After that some thoughts about how to improve the methods are provided as an end.The result shows that some methods do have a help to the static signal detection, but there is more job to do to get a better effect. To improve the object detection skill is still an important task.