基于时域线性插值的信号周期计算方法及误差分析  

Time-domain linear interpolation algorithm and its error analysis for estimating signal period

作　　者：周峰[1] 赵春宇[1] 黄震宇[1] 陈大跃[1] 

机构地区：[1]上海交通大学电子信息与电气工程学院,上海200240

出　　处：《仪器仪表学报》2011年第8期1724-1730,共7页Chinese Journal of Scientific Instrument

摘　　要：信号周期或信号基波周期的准确测量是周期信号分析的重要环节。基于时域线性插值的周期检测方法是非同步采样情况下的常用方法,该算法通过计算信号与设定阈值的交叉点间的时间间隔来获得信号周期。插值算法的测量不确定度主要受到阈值、采样频率和信噪比3个因素的影响,推导出了3个影响因素与测量误差之间的关系式,并进行了理论分析。仿真实验证明,通过选取信号平均值作为阈值,提高采样频率,利用带通滤波器降低噪声干扰等方法,可有效地减小插值算法测量误差,提高测量准确度。时域插值算法在谐波分析仪中的应用验证了该算法的实用性。Accurate measurement of the fundamental period of a periodic signal is an important task in periodic signal analysis. Time-domain linear interpolation （TLI） algorithm is a common measuring method, which can be applied in the course of asynchronous sampling. Using the TLI algorithm, the signal period can be obtained by calculating the time interval between adjacent crossing positions where the signal crosses a given threshold; therefore, the measure- ment uncertainty is affected by the threshold, sampling frequency and noise. This paper investigates the relationships between the measurement error and these three critical factors. Simulation results show that the measurement errors can be reduced effectively and the measurement accuracy can be improved by setting the signal mean value to be the threshold, increasing the sampling frequency and diminishing the noise interference using a band-pass filter. The application of the proposed TLI algorithm in a harmonic analyzer validates the effectiveness of the proposed method.

关 键 词：周期测量 时域插值算法 误差分析 

分 类 号：TN911.6[电子电信—通信与信息系统]