起伏噪声中单矢量传感器预警检测方法  

A target datection and early warning method with a vector-sensor under fluctuating background noise

作　　者：郭虎生[1] 颜冰[1] 

机构地区：[1]海军工程大学兵器工程系,武汉430033

出　　处：《振动与冲击》2014年第10期189-192,共4页Journal of Vibration and Shock

摘　　要：针对现有振动传感器能量检测方法在较低信噪比及环境噪声变化较大时检测性能较差问题,提出矢量传感器协方差检测方法。该方法利用矢量传感器多通道采样数据构造协方差矩阵,并以最大特征值与最小特征值之比为检验统计量进行目标判决。用该方法检测目标不仅无需噪声先验知识,可有效克服噪声起伏变化带来的影响。仿真及实验表明,该方法检测性能良好,计算复杂度小,适合于矢量传感器对目标检测预警。The ship motion information can be detected with a vibration sensor deployed at the bottom of sea. Energy detection used in a vibration sensor is the most traditional method.However,its performance has a significant loss under a dramatic noise fluctuation and a low signal to noise ratio.To solve this problem,a sample covariance detection algorithm for a vector sensor was proposed here.According to the characteristics of signals received by the vector sensor, the ratio of the maximum eigenvalue to the minimum one was taken as the detection statistic to make target determinations. The proposed method exhibited a good performance and robustness against noise fluctuation without having to know any prior information of the noise.The simulation and experiment results showed that the new detection method can be used to improve the target detection performance of a vector sensor under noise fluctuation,and its computational complexity is smaller,it is suitable for target detection and early warning with a vector sensor.

关 键 词：信息处理技术 矢量传感器 采样协方差 起伏噪声 

分 类 号：TN911.7[电子电信—通信与信息系统]