Open AccessDesign Issues for MEMS-Based Pedestrian Inertial Navigation Systems
Author(s) -
Pavel S. Marinushkin,
Valentina A. Bakhtina,
Ilya Podshivalov,
O Stukach
Publication year - 2015
Publication title -
nauka i obrazovanie
Language(s) - English
Resource type - Journals
ISSN - 1994-0408
DOI - 10.7463/0615.0786740
Subject(s) - pedestrian , inertial navigation system , microelectromechanical systems , computer science , inertial frame of reference , inertial measurement unit , aerospace engineering , systems engineering , aeronautics , engineering , transport engineering , physics , optoelectronics , quantum mechanics
The paper describes design issues for MEMS-based pedestrian inertial navigation systems. By now the algorithms to estimate navigation parameters for strap-down inertial navigation systems on the basis of plural observations have been already well developed. At the same time mathematical and software processing of information in the case of pedestrian inertial navigation systems has its specificity, due to the peculiarities of their functioning and exploitation. Therefore, there is an urgent task to enhance existing fusion algorithms for use in pedestrian navigation systems. For this purpose the article analyzes the characteristics of the hardware composition and configuration of existing systems of this class. The paper shows advantages of various technical solutions. Relying on their main features it justifies a choice of the navigation system architecture and hardware composition enabling improvement of the estimation accuracy of user position as compared to the systems using only inertial sensors. The next point concerns the development of algorithms for complex processing of heterogeneous information. To increase an accuracy of the free running pedestrian inertial navigation system we propose an adaptive algorithm for joint processing of heterogeneous information based on the fusion of inertial info rmation with magnetometer measurements using EKF approach. Modeling of the algorithm was carried out using a specially developed functional prototype of pedestrian inertial navigation system, implemented as a hardware/software complex in Matlab environment. The functional prototype tests of the developed system demonstrated an improvement of the navigation parameters estimation compared to the systems based on inertial sensors only. It enables to draw a conclusion that the synthesized algorithm provides satisfactory accuracy for calculating the trajectory of motion even when using low-grade inertial MEMS sensors. The developed algorithm can be implemented in the individual navigation equipment designed to solve practical tasks requiring autonomous positioning