A Multi-Constraint Saturated Acceleration Compensation Method for Pedestrian Inertial Navigation Under High-Dynamic Gaits

The pedestrian inertial navigation system (PINS) based on zero-velocity updates (ZUPT) and foot-mounted MIMU could face the problem that the actual acceleration input could exceed the full scale range (FSR) of some commercial MIMU, causing saturation in its readings and loss of accurate measurements when pedestrians proceed with high dynamic gaits such as fast walking and running. Considering the cost and performance of the MIMU, this paper proposes a Multi-Constraint Saturated Acceleration Compensation (MCSAC) method to compensate for the immeasurable values of saturations of the accelerometer. MCSAC constructs the saturated immeasurable values as unknown vectors, establishes an optimization model based on the constraints of step length difference, velocity deviation, and terminal displacement, and uses the interior-point method for optimization. With the optimal values compensated to the raw inertial data, the PINS solutions are corrected. Experiments validating the effectiveness of MCSAC in suppressing errors caused by insufficient FSR under high-dynamic gaits.