Online error compensation of moving-base rotating accelerometer gravity gradiometer

Imperfections in the factors of a rotating accelerometer gravity gradiometer (RAGG), such as accelerometer mounting errors, circuit gain mismatch, accelerometer linear scale factor imbalances, and accelerometer second-order error coefficients, make the RAGG susceptible to its own motion. These motion errors easily cause saturation of the RAGG so that it is unable to work normally. In this study, we propose a scheme for continually adjusting the linear scale factors or mounting angles of the accelerometers to reduce motion sensitivity and for generating a compensation signal based on an analytical model of the RAGG, to compensate motion errors. A numerical model of the RAGG is used to simulate a real imperfect RAGG to allow an online error compensation experiment to be performed. In the experiment, the mean and standard deviation of the air turbulence are 100 mg and 20 mg (1 g = 9.81 m/s2), respectively, and those of the angular velocity are 100 deg/h and 50 deg/h. It takes about 15 min for the RAGG online error compensation system to achieve convergence. In the converged state, the motion noise density of the RAGG is about 1 E/√Hz and scale factor balances of the order of 10−9 g/g are maintained. These experimental results suggest that the proposed online error compensation method is valid.