基于负熵的人员脚步震动信号检测识别算法  

Algorithm for Person Footstep Detection and Identification of Seismic Signal Based on Egentropy

作　　者：梁志强[1] 魏建明[1] 赵俊钰[1] 刘海涛[1] 

机构地区：[1]中国科学院上海微系统与信息技术研究所,上海100050

出　　处：《小型微型计算机系统》2010年第1期190-192,共3页Journal of Chinese Computer Systems

基　　金：上海市科委重点项目(07dz15011)资助

摘　　要：提出一种基于震动信号的人员脚步检测识别算法,该算法根据信息论中的负熵概念,采用高阶累积量的负熵近似计算方法.仿真与实测结果证明,与一般的模式识别算法相比,该算法具有三个重要的优点,包括环境适应性强、识别准确率高和运算量小.这些优点使得该算法更适用于能量受限、随机自组的无线传感器网络,能够在野外环境下准确、简单的检测识别人员脚步震动信号.This paper provides an algorithm for person footstep detection and identification of seismic signal. This algorithm makes use of the computational method of negentropy which is based on high-order cumulants according to information theory. Simulation and implementation results proves that the algorithm described in this paper shows high performance that includes better adaptability to en- vironments, more accurateness of identification and less computational resource compared with general pattern recognition algorithm. These advantages make the algorithm more adaptable to energy-limited and random-organized Wireless Sensors Network （WSN） and ensure the accurateness and simpleness in person footstep identification under open country.

关 键 词：负熵 无线传感器网络 模式识别 人员脚步 高阶累积量 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]