A MEMS based, low cost GPS-aided INS for UAV motion sensing

This paper presents a new design and development of a low cost GPS (Global Positioning System)-aided INS (Inertial Navigation System) using Micro-ElectroMechanical-System (MEMS) inertial sensor. A typical MEMS type inertial sensor consists of three orthogonally aligned accelerometers and three orthogonally aligned gyroscopes confined in a very small chip. In this paper, the intensive preprocessing and modeling techniques of MEMS inertial sensor's primitive, noisy motion data are outlined. These techniques transform the erroneous motion data into usable motion indicators illustrated in three-dimensional position, three-dimensional velocity, and three-dimensional orientation. GPS serves as an aiding device to tune the MEMS inertial sensor's measurements through Kalman filter, while extra sensory feedbacks from magnetometers are employed to improve the system performance. Experiment was conducted to evaluate the performance of the GPS-aided INS by installing it into an Unmanned Aerial Vehicle (UAV) for motion sensing. Lastly, the experiment results are evaluated and verified using the motion data generated from a commercial navigation system.