基于改进D-S证据理论的水下航行器声隐身性能评估

Acoustic stealth situation assessment for an underwater vehicle based on improved D-S theory

针对传统水下航行器声隐身性能评估方法计算时间长、实时性不强的缺点,将评估由数值计算问题变为基于多传感器信息融合的模式识别问题来解决。通过将改进D-S证据理论模型和图像欧氏距离模型相结合,提出了一种新的水下航行器声隐身性能快速评估方法。首先利用加速度传感器测得壳体表面振动信息,抽取分析频段内每个频带的功率作为特征向量,然后通过图像欧氏距离方法进行证据生成,最后用一种改进的D-S证据合成方法进行识别分类,快速评估出航行器当前的声隐身状态。并利用计算机仿真和水下振动实验相结合的方式,通过类柱体结构模拟航行器舱段,详细讨论了方法的先进性,并研究指出了需要改进的问题。探讨了一种水下航行器声隐身性能快速评估的实现方法。

Traditional algorithm on underwater vehicle radiated noise calculation can not satisfy real-time request in practical use due to longer computing time.Aiming at this disadvantage,the problem was converted into a pattern recognition problem based on multi-sensor data fusion,and a new method of fast evaluation acoustic stealth condition of underwater vehicle was studied.The method combined an improved D-S theory and an image Euclidean distance model.Firstly,the power of each frequency bands in analysis bands was extracted as the feature vectors.And then the evidence was built with the image Euclidean distance method.At last,the rapid assessment was made by the improved D-S theory.The availability of the method was proved with computer simulation and an underwater test.The advantages were analyzed in detail,and disadvantages also were pointed out.A new approach to evaluate acoustic stealth condition of an underwater vehicle was explored.