摘要
间接法测量是在时间长度为待测信号周期整倍数的测量时间内,通过计算高频基准时钟个数来得到待测频率。基于间接法的回归法是把整个测量时间分成很多相等的小时间段,对每一小时间段内记录到的高频时钟个数进行回归处理,可以得到比间接法更高的分辨率。本研究通过严格的理论计算和实验,证明了应用回归法测量频率时,测量时间必须分成36段以上的小时间段,且测量时间内的小时间段数越多,测量分辨率越高。
The period measuring method counts the total number of high frequency pulses during certain period of time to obtain the frequency of input signal. The regression method takes into consideration the numbers of pulses counted during each individual short interval. Regression method can improve resolution, but the analysis and conclusions presented by Staffan Johansson are not correct in the world at present. Correct analysis and conclusions are presented in this paper and verified by experiments. Analysis and experiments show that the regression method is superior to the period measuring method when the number of counted input cycles is greater than 36 and the regression method will have better performance as the number of counted input cycles grows larger.
出处
《天津农学院学报》
CAS
2016年第1期54-56,共3页
Journal of Tianjin Agricultural University
基金
天津市大学生创新创业训练计划项目"微波介电测量系统的最优化设计"(201410061030)
国家级大学生创新创业训练计划项目"斯特林发电机整流变频电路的设计"(201410061125)
关键词
频率测量
线性回归
分辨率
frequency measurement
linear regression
resolution
