Gravity Gradiometer Survey and Real Time Techniques for Improving Inertial Navigation System Accuracy

摘要

As inertial navigation instrumentation becomes more advanced, erros associated with imprecise knowledge of the true gravity vector become significant. With the development of the moving base gravity gradiometer, a means for detecting the true gravity vector is available. This research explores both real time and a priori schemes for using gradiometer information to reduce the position and velocity error associated with an inertial navigation system. In the real time application, information derived from the gradiometer is filtered to produce estimates of position and velocity error. The problem is to derive accurate filters in the presence of an inherently transcendental gravity field. Conventional Kalman filters are cumbersome to apply in this instance. However, the sensitivity of inertial navigation systems at Schuler frequency makes it possible to neglect errors outside of a narrow band of frequencies centered at Schuler frequency. This approximation leads to low order filters which may be applied with surprising accuracy. For the survey application, the gradiometer is used to establish a reference gravity model to be used in open loop inertial navigation system operation. By determining the root mean squared values of the resulting position and velocity errors. It is possible to evaluate the effectiveness of the survey scheme and to compare the survey application with the real time gradiometer application. Analytic and numeric covariance methods are used again to do these analyses.