Data Acquisition And Processing Software For Liquid Drop Fingerprint

In Fiber-Capacitive Drop Analysis technology, "liquid drop fingerprint" is a unique figure that can be obtained as the fiber signal changes with the drop volume during the drop growth. It can be used as a new method to distinguish two different liquid precisely, for its good reflections of the liquid's density, viscosity, refractive index, surface tension, etc. In order to obtain the "liquid drop fingerprint" accurately, this paper give its emphasis to a series of data processes, including data filtering, data normalization, and feature extraction.The paper is supported by the State Natural Sciences Fund (Project Number:60702004). In this paper, sampling precision and sampling frequency of the data acquisition system required by the Fiber-Capacitive Drop Analyzer is researched. Meanwhile, the development of the data acquisition system and the FORTRAN subroutines is introduced.In this paper, after comparing the features of normal filter and subsection filter, a "first-order differential" method for optimizing subsection filter is explored and analyzed. Experiment improves that it is an effective and concise way to divide the signals into segments for subsection filter.In this paper, the importance and objection of normalization is introduced. And then, the two different styles of normalizations, respectively based on time dividing and capacitive dividing, are compared and reached. Additionally, the development of the capacitive based dividing method FORTRAN subroutines is introduced.In this paper, after comparing and researching the two different wave analysis method, neighborhood comparing and poly detecting method, "differential calculus" is introduced. It has the advantage in immunity from data undulation and robustness for various situations.