Processing of Redundant Information in Airborne Electronic Systems by means of Neural Networks

摘要

Processing of redundant information of the noncollinear measuring instrument developed by the technology of microelectromechanical systems is considered. The algorithm of information processing based on neural networks is proposed. The structural schemes of neural networks for single axis and triaxial inertial sensors are given. The neural network with one hided layer for information processing of noncollinear redundant measuring instrument based on inertial sensors is represented. The features of information processing algorithm are described. A process of a training process is represented. Results of information processing by means of neural networks are given. The obtained results can be useful for improving information used for navigation of moving vehicles of the wide class.