Fuzzy regression. III. Calibration of the NH3-and CO2-sensitive potentiometric sensors

A new fuzzy regression algorithm was discussed and successfully applied for calibration of a potentiometric gas sensor for NH3 and CO2 respectively. D ifferent regression techniques were also considered, and for the evaluation of the results more objective quality coefficients were used Depending upo n the structure of data, the fuzzy procedure could improve the results dram atically. Considering the results obtained in this paper and others unpubli shed yet it appears clear that the calibration procedure based upon the fuz zy regression is the most suitable regression method in the potentiometric measurements.