An Experimental Study of UAV Motion Sensing using GPS Aided MEMS Inertial Sensor

Due to the nature of Micro-Electro- Mechanical System (MEMS) inertial sensor, its outputs are normally being jeopardized by measurement errors. A common practice to regulate the MEMS inertial measurements into usable motion data is by fusing the Global Positioning System's (GPS) measurement with the MEMS inertial measurement through Kalman filter. Such integrated system is known as GPS aided MEMS Inertial Navigation System. Note that the robustness of such integrated system is heavily relied on the accuracy of the modelling of stochastic error of MEMS inertial sensor. In this paper, the on-board motion sensing experimental of the GPS aided MEMS Inertial Navigation System implemented on an Unmanned Aerial Vehicle (UAV) airplane is carried out. The results of two different stochastic noise models implementation, namely the Gauss-Markov (GM) Model and Autoregressive (AR) Model, are studied and compared. Results of short intervals (20 seconds) of no GPS condition are studied as well. The outcomes indicate that the stochastic noise model using AR modelling produces better estimation results than GM modelling.