A new method to detect underwater intruders using the disturbed sound field they cause has been developed. Terrorist incursions have always occurred in shallow water. Shallow water acoustic fields were used to model underwater sound transmission. Based on ray theory, targets, such as divers and robots, were modeled as elastic cylinders of finite length. The target's scattering field was then derived. These two sound fields were then united to model a disturbed sound field. An intruder detection algorithm was proposed which counts maximums of the windo wing power spectrum. The proposed algorithm estimates the power spectrum by overlapping the average periodogram which is multiplied with a rectangular window in the frequency domain to extract feature maximums. Experimental results in a lake showed that the method provides a detection probability about 75 % (if threshold changes, the probability could reach 85%) for intruder detection in a known underwater environment.%提出了一种利用水下入侵目标所引起的扰动声场来检测水下入侵目标的方法.针对浅海发起恐怖活动的主体一般为水下蛙人、机器人等小目标,建立了基于有限长弹性圆柱体的目标散射声场模型,并结合基于射线理论的浅海虚源声场模型,建立了入侵目标引起的扰动声场模型.利用滑动平均周期图法估计信号的功率谱,提出了基于频域加窗功率谱极大值数的目标检测算法.湖上试验结果表明,在当时的水声环境中,提出的方法能以75%左右(若改变检测门限,检测概率可达85%左右)的检测概率探测到水下入侵目标.
展开▼
