A Novel Minimum Distance Constraint Method Enhanced Dual-Foot-Mounted Inertial Navigation System for Pedestrian Positioning

Tao Liu,Jian Kuang,You Li,Xiaoji Niu

DOI: https://doi.org/10.1109/jiot.2023.3271309

IF: 10.6

2023-01-01

IEEE Internet of Things Journal

Abstract:Foot-mounted inertial navigation system (Foot-INS) with the zero velocity update (ZUPT) has become one of the indispensable technical means in professional pedestrian positioning fields due to the advantages of self-constraint and immune to environmental factors. The dual-Foot-INS can provide more excellent autonomous positioning performance than a single-Foot-INS because it utilizes more opportunities for zero velocity correction and additional distance constraint information. However, the classical dual-Foot-INS does not fully exploit the distance constraint potential for positioning improvement. In this article, we proposed a novel minimum distance constraint (MDC) method that achieves higher positioning accuracy than the traditional dual-Foot-INS methods. To obtain an accurate and consistent state estimation under the nonlinear distance constraint problem, we propose an iterative distance constraint (IDC) algorithm. The IDC is transformed into an approximate linear constraint model, and an alternative estimate is obtained by the estimation projection method. To solve the problem that the distance constraint moment in the traditional method is affected by the recursive foot positions, we propose a more reasonable and reliable minimum distance moment detection (MDMD) method. The proposed MDMD method maximizes the positioning performance improvement of the dual-foot pedestrian system. Two rigorous experimental tests with a long walking trajectory without turn around and closed loop were conducted to verify the effectiveness of the proposed method, the positioning error of the proposed method is reduced by 83.5% and 62.9% compared to the classical ZUPT and MDC methods, respectively.