UTILIZATION OF A NEURAL NETWORK ALGORITHM FOR BALANCING OF HEMISPHERICAL RESONATOR GYRO

摘要

For the first time, a problem of balancing of the imperfect hemispherical resonator gyro with inhomogeneous angular mass distribution density is considered as a discrete optimization problem, and a neural network algorithm is proposed for finding its solution. The algorithm is based on minimization of the two-dimensional Hopfield network and is analogous to original algorithms for solving some problems of discrete optimization, such as the traveling salesman problem (TSP).