面向伪周期信号在线处理的模板参数估计简化方法  

Simplified template parameter estimation for pseudo-periodic signal on-line processing

作　　者：刘少强[1] 汪立林[1] 王爱民[2] 李勇周[1] 庞新苗[1] 

机构地区：[1]中南大学信息科学与工程学院,长沙410075 [2]东南大学仪器科学与工程学院,南京210096

出　　处：《东南大学学报（自然科学版）》2009年第5期904-908,共5页Journal of Southeast University：Natural Science Edition

基　　金：国家自然科学基金资助项目(60776834);国家高技术研究发展计划(863计划)资助项目(2009AA01Z311)

摘　　要：为适应资源有限系统对伪周期信号在线时域处理及数据压缩的要求,提出了一种模板参数估计简化方法.首先,将伪周期信号的每个一循环视为单周期信号,并以模板与单周期信号的均方误差作为目标函数,选择傅立叶级数作为模板.然后,根据两相邻周期信号的极大点或极小点位置差估计模板函数基频,利用三角函数正交性直接计算傅氏系数及误差指标,并通过预测与在线调整级数阶次,快速确定出给定精度下的最佳模板阶次.通过进一步在线微调信号周期,改善基频估计的准确性,得到符合精度要求的各伪周期信号模板参数.最后,基于超低功耗微控制器实现了对伪周期信号脉搏波的时域处理实验.实验结果表明,该方法简单有效,有利于减小计算功耗,适用于无线传感器节点等资源有限系统.In order to process pseudo-periodic signals on-line in time domain and data compression in hardware resource-constrained systems, a simplified method of template parameters estimation is presented. First, considering each cycle of pseudo-periodic signals as one single periodic and taking the mean square error between the template and measurements as a cost function, Fourier series is used as the template of the pseudo-periodic signal. Then, the fundamental frequency of the series is estimated by measuring the distance of the maximums or the minimums between each successive cyclic signal. With the orthogonality of trigonometric function, the Fourier coefficients and the error index are computed, and the best degree of series for a given tolerance can be quickly determined by predicting and on-line adjusting the degree of the series. Furthermore, through nicely adjusting the period of current cycle on-line, the matched errors are below the given tolerance, and therefore the coefficients corresponding to each pseudo-period can be refined. Finally, based on an ultra-low power microcontroller, the pseudo-periodic pulse wave on-line processing is implemented. The experimental results show that this method is simple and valid. It is helpful to reduce the power dissipation of computation and adaptive to resource-constrained system such as wireless sensors.

关 键 词：伪周期信号 在线处理 模板参数估计 傅立叶级数 资源有限系统 

分 类 号：TP274.2[自动化与计算机技术—检测技术与自动化装置]