Estratégias de estabilização do canal vertical em sistemas de navegação inercial via integração barométrica

Abstract

Inertial Navigation Systems (INS) are navigation systems composed of accelerometers and rategyros. From these sensors’ measurements, the knowledge of initial conditions, one is able to obtain values of position, velocity, and attitude. After its initialization, an INS has the advantage of being independent on external signals for its continuous operation, as well as having its navigation solution uniquely determined from inertial sensors data, and from a gravity model. However, in the long term, the INS vertical channel, that is, its altitude and vertical velocity, is unstable, accumulating large errors in a relatively short time period. To solve this problem, the integration of the INS with a barometer has been widely used, since the latter presents a non-drifty altitude solution in the long term. This work investigates the INS vertical channel instability issue, and presents different methods of integrating the barometer with the INS aiming at stabilizing this channel. As the main contribution of the work, comparisons are established between integration methods (referred to as mechanizations) traditionally used in the literature (mostly based on empirical tuning of control loops), and new integration strategies based on optimal control and minimization of performance indices, namely, the Linear Quadratic Regulator (LQR), and the Integral Absolute Error (IAE), Integral Squared Error (ISE), Integral of Time Multiplied Absolute Error (ITAE), and Integral of Time Multiplied Squared Error (ITSE) criteria. Simulated and experimental results are presented to highlight the performance of the investigated integration methods. As main conclusions, one verifies that the optimal control-tuned mechanizations presented more stable results, with smaller oscillations. In particular, the tuning obtained via minimization of ITAE criterion presented the best result, with the most stable solution and with the smallest values of Root Mean Square Errors (RMSE) in altitude and vertical velocity.