Deep data fusion method for missile-borne inertial/celestial system
Abstract
Strap-down inertial-celestial integrated navigation system has the advantages of autonomy and high precision and is very useful for ballistic missiles. The star sensor installation error and inertial measurement error have a great influence for the system performance. Based on deep data fusion, this paper establishes measurement equations including star sensor installation error and proposes the deep fusion filter method. Simulations including misalignment error, star sensor installation error, IMU error are analyzed. Simulation results indicate that the deep fusion method can estimate the star sensor installation error and IMU error. Meanwhile, the method can restrain the misalignment errors caused by instrument errors.
- Publication:
-
6th International Conference on Computer-Aided Design, Manufacturing, Modeling and Simulation (CDMMS 2018)
- Pub Date:
- May 2018
- DOI:
- 10.1063/1.5038986
- Bibcode:
- 2018AIPC.1967b0014Z