Aerospace Instrument-Making Annotation << Back TRAINING AND CALIBRATION TECHNOLOGY VARIANT INTELLIGENT INERTIAL SENSORS V.S. Kulabukhov, V.V. Bulgakov, V.A. Novikov A variant of smart sensor calibration technology is proposed based on the synergistic effect of two innovations – system identification based on the General principle of isomorphism and model approximation using artificial neural networks. The goal is to simplify the development and production of sensors through individual training using the developed "smart" algorithm. Using the example of an angular velocity sensor, we demonstrate the procedure for identifying its unknown real two-dimensional measuring characteristic based on calibration experiments. The technology allows using the known code of the analog-to-digital angular velocity converter and the current temperature to obtain at the output of the sensor, which has an individually special conversion function and is supplemented by a "smart" algorithm, the true values of the measured physical quantity with the required accuracy in the expected operating conditions. The article presents the results of modeling and training an artificial neural network that implements the above "smart" algorithm. Keywords: identifi cation, angular velocity sensor, calibration, neural network, training. DOI: 10.25791/aviakosmos.10.2020.1182 Pp. 11-25.