Error Analysis of ZUPT-Aided Pedestrian Inertial Navigation

We present, for the first time, an analytical relation between the navigation errors during Zero-Velocity-Update (ZUPT) aided pedestrian inertial navigation and Inertial Measurement Unit (IMU) performances. Angle Random Walk of gyroscopes and Velocity Random Walk of accelerometers in IMU, velocity measurement uncertainty during the stance phase, and the sampling frequency of IMU are studied and their effects on the overall navigation accuracy are analyzed. Numerical simulation of the ZUPT-aided navigation algorithm is conducted based on a generated pedestrian trajectory to verify the analytical results with the discrepancy less than 30%, showing fidelity of the results. In this study, a foot motion model was developed according to the ambulatory gait analysis as a basis of the numerical simulation. This study may serve as a benchmark for analysis of errors in pedestrian inertial navigation.