Measurement device with learned sensor

The main subject of the paper is calibration of the sensor by learning method of the fuzzy system. Fuzzy system is used for simultaneous compensation of the nonlinearity and temperature sensitivity. The compensation of pressure sensors is intended for application in the Air Data Computer. The results obtained experimentally show that fuzzy system is superior to the polynomial regressions because it is not necessary to compute high degree polynomials. After the compensation it is supposed measurement device can be periodically verified and if necessary corrected. General property of two-point correction is presented. In the case of tested sensor two-point re-calibration is assumed as being effective because of one-side nonlinearity.