Application of an ANFIS Algorithm to Sensor Data Processing

Some sensors require frequent recalibrations; therefore, calibration cost must be limited. In this paper, a new calibration technique is presented. It is a two-phase method which is based on adaptive neuro-fuzzy networks, and it shows superior performances with respect to traditional algorithms, requiring fewer calibration points and less computational power in the recalibration phase. Feasibility has been demonstrated with a pyroelectric biaxial positioning system, reaching performance to the limit of the adopted test bench, on the order of 20 um with respect to a whole sensible area of 7 mm x 7 mm.