基于DSP技术及互相关算法的光纤液滴传感器

光纤液滴分析技术是通过反映被测液体的液滴生长过程的光纤液滴指纹图,测量被测液体的物理、化学特性.互相关原理被用来对不同液体的光纤液滴指纹图进行对比分析,分析液体性质或判决2种液体的异同,DSP技术的使用提高了互相关运算的速度.基于DSP技术和互相关原理制作了实验室样机,并进行了样品测试.分别利用互相关函数和互相关系数测试了不同浓度的酒精溶液,测试结果显示,互相关函数峰值可以作为一种测量液体浓度的方法,酒精溶液的测量精度达到3.4%,而互相关系数法可以甄别不同浓度的酒精溶液,最小系数在60%以下.

Fiber liquid drop sensors based on DSP technique and cross-correlation algorithm

The fiber liquid drop sensor can be used to monitor the dripping process of a liquid drop and obtain the fiber fingerprint drop trace （FFDT）, which reflects the physical and chemical properties of liquid samples to be tested. The cross-correlation algorithm could be adopted to dig characters of liquids, in addition, the algorithm could be used to judge if two liquid samples are same. The application of DSP technique speeds up the operation of the cross-correlation algorithm. An experimental device was fabricated and applied to test some samples. Both cross-correlation function and cross-correlation coefficient were used to test alcohol solution with different concentrations. The testing results show that the peaks of the cross-correlation functions could be used to measure the concentrations of alcohol solutions with standard deviation as 3.4%, while the cross-correlation coefficient is a good candidate to judge alcohol solutions with different concentrations. The minimum cross-correlation coefficient is less than 60%.