Calibration of an electronic measuring system for ultrasonic analysis of solutions

The problem of calibrating a system that measures the concentration of methanol in water is described. The limitations of polynomial and inverse-polynomial models used for this purpose are demonstrated. The cubic-spline interpolation is chosen and examined as a flexible means of calibration. A method for experimental validation of the results of calibration is proposed. It is concluded that spline-based calibration is much more flexible than the procedures dedicated to the fixed-polynomial or inverse-polynomial models. It allows the utilization of a much greater amount of calibration data, thus improving the accuracy of calibration, and requires much less a priori information on the relationships between quantities involved in the measurement process.<
>