Cooperative Positioning Method of Dual Foot-Mounted Inertial Pedestrian Dead Reckoning Systems

Abstract

The inertial-based pedestrian dead reckoning (PDR) system assisted by zero-velocity update (ZUPT) algorithm has been widely researched, as it can independently provide effective pedestrian's position in indoor environment. However, in the realistic test, the system still often suffers from drift, due to the ZUPT algorithm's poor observation of system's heading. In this article, based on full analysis of the existing maximum range constraint algorithm between both-side feet, first, a constraint method based on the maximum distance of one-side foot has been proposed. It mainly uses the principle that the distance between the position of one-side foot at current time and that of the previous stationary instant should be within a maximum range. Second, a one-side foot and both-side feet cooperative maximum range constraint algorithm has been proposed. It mainly uses the one-side foot maximum range and the both-side feet maximum range to intersect the most likely position of pedestrian's current moving foot and then uses this position to constrain the heading divergence. Finally, to further improve the heading constraint accuracy, a maximum range decoupling algorithm has been proposed, which decouples the maximum ranges of one-side foot and both-side feet into the distance in the forward direction. Then, using this distance, the heading divergence can be constrained more effectively.