Hidden Markov Model-based Pedestrian Navigation System using MEMS Inertial Sensors

Yingjun Zhang, Wen Liu, Xuefeng Yang, Shengwei Xing
2015 Measurement Science Review  
In this paper, a foot-mounted pedestrian navigation system using MEMS inertial sensors is implemented, where the zero-velocity detection is abstracted into a hidden Markov model with 4 states and 15 observations. Moreover, an observations extraction algorithm has been developed to extract observations from sensor outputs; sample sets are used to train and optimize the model parameters by the Baum-Welch algorithm. Finally, a navigation system is developed, and the performance of the pedestrian
more » ... vigation system is evaluated using indoor and outdoor field tests, and the results show that position error is less than 3% of total distance travelled.
doi:10.1515/msr-2015-0006 fatcat:i6wioemcarglbics23tl5ne5lu