多波信号到达时间自动识别的核函数方法研究  

Research on Automatic Recognition of the Arrival Time of Multiwave Signal Model Based on Kernel Function

作　　者：任波[1] 宋瑞雪[1] 付志伟[1] 

机构地区：[1]沈阳理工大学自动化与电气工程学院,沈阳110159

出　　处：《沈阳理工大学学报》2016年第4期27-32,共6页Journal of Shenyang Ligong University

摘　　要：信号所到达时间的具体信息,对于信号发射事件的定位,识别和信号发射源机理的分析都十分重要。提出了一种基于核函数的自动识别算法,该算法利用噪声信号和发射信号的核函数模型,利用Akaike的信息准则,对信号到达的时间进行了自动识别。该方法较其他方法具有计算简便的优点。仿真试验结果表明,基于核函数模型的信号到达时间识别的算法运行速度快,可以实时对信号的到达时间进行识别,能够给出较为准确的结果。Signal arrival time information is very important for the localization, identifica- tion of signal emission event and the analysis of mechanism of signal emission source. A new algorithm for automatic recognition based on kernel function is presented. The algo- rithm uses the kernel function model of the noise and the transmitted signal. By Akaike cri- terion application, the signal arrival time is automatically recognized. This method has the advantages of simple calculation in comparison with other methods. Simulation results show that the signal arrival time algorithm based on kernel function model runs fast. It could iden- tify the arrival time of the real-time signal and zive more accurate results.

关 键 词：模型 核函数 到达时间 自动识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]