作者: Francesco Cappello , Subramanian Ramasamy , Roberto Sabatini
DOI: 10.4271/2015-01-2459
关键词:
摘要: Multi-Sensor Data Fusion (MSDF) techniques involving satellite and inertial-based sensors are widely adopted to improve the navigation solution of a number mission- safety-critical tasks. Such integrated Navigation Guidance Systems (NGS) currently do not meet required level performance in all flight phases small Remotely Piloted Aircraft (RPAS). In this paper an innovative Square Root-Unscented Kalman Filter (SR-UKF) based NGS is presented compared with conventional UKF governed design. The system architectures adopt state-of-the-art information fusion approach on low-cost including; Global Satellite (GNSS), Micro-Electro-Mechanical System (MEMS) Inertial Measurement Unit (IMU) Vision Based (VBN) sensors. Additionally, Dynamics Model (ADM), which essentially knowledge module, employed compensate for MEMS-IMU sensor shortcomings high-dynamics attitude determination ADM acts as virtual its measurements processed non-linear estimation order increase operational validity time. An improvement state vector (i.e., position, velocity attitude) obtained, thanks accurate modeling aircraft dynamics advanced processing techniques. SR-UKF VBN-IMU-GNSS-ADM (SR-U-VIGA) architecture design was implemented typical (U-VIGA) RPAS (AEROSONDE) integration arrangement exploring representative cross-section envelope. comparison position data shows that SR-U-VIGA U-VIGA fulfill relevant RNP criteria, including precision