Adaptive In-Flight Alignment of INS/GPS Systems for Aerial Mapping

In this paper, a new in-flight alignment (IFA) method of the integrated inertial navigation system (INS) and global positioning system (GPS) is proposed for the aerial mapping applications. The integrated INS/GPS measurement system is used to provide attitude information and, based on this, the exterior orientation parameters can be derived for the purpose of direct georeference of the airborne imagery. The IFA plays an important role in achieving high accuracy of attitude estimation. However, the statistics of INS noise is usually time-varying and seriously degrades the estimation accuracy in practice. In order to solve the problem, two strategies are taken into account in the paper. First, an adaptive estimation algorithm is developed by adjusting the window size of data processing in IFA, so that the covariances of INS noise can be estimated and updated online to improve the state estimation performance. Second, a strong tracking filter is applied to guarantee the convergence of the IFA algorithm as well as its robustness against parameter perturbations and trajectory maneuvers. Finally, an aerial mapping experiment is implemented, and the results demonstrate the effectiveness of the proposed method.