基于负熵的人员脚步震动信号检测识别算法

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.

Algorithm for Person Footstep Detection and Identification of Seismic Signal Based on Egent-ropy

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.