精确的互相关算法在超声波流速测量中的应用

互相关算法能够比较准确地计算出声波在流速正反2个方向上的传播时间,进而得出时间差,并根据时间差求出流体的流速。但离散的互相关函数峰值的精度取决于采样间隔的大小,间隔减小精度会相应提高,但在实际的应用中计算量也会大幅度提高。在以单片机为核心的测量仪表中,这种规模的计算通常是不会被采用的,找到一种更简便的计算方法是解决问题的关键。在离散的互相关函数峰值附近,利用抛物线算法可以得出更精确的互相关函数峰值。这种方法允许采样间隔较大,计算时间可大幅度缩短,使同一测量精度的计算时间从3.7 s缩短到0.9 s左右。

Application of accurate cross-correlation method in ultrasonic flow measurement

The propagation time of ultrasonic in positive and negative directions of gas flow can be exactly calculated by using cross-correlation, and the time difference between the two directions can be further computed, and the velocity of gas flow can be also computed. However,because the calculation error of peak of discrete cross-correlation function depends on sample interval,decrease in interval is to increase in precision and certainly it is to increase by a long way in the quantity of calculation. In measuring instruments that are controlled by single chip computer, the method that needs huge calculation is usually not used. The key of the question is to find a relatively simple method. In the vicinity of peak of discrete cross-correlation function,a more accurate peak can be gained by utilizing parabola algorithm, The algorithm permits to have longer sample period and curtails calculating time from 3.7s to 0.9s in the same measurement precision.