Abstract:

Positioning signals in the underground environment without navigation satellite system leads to common ground location methods which can not be used in underground space.Traditional dead reckoning by cell phone usually requires fixed its attitude or combined with other wireless location methods to correct gait and positioning error.In order to facilitate people to use,reduce the extra hardware layout and avoid interference between wireless signals,this paper improves the gait detection algorithm to eliminate the influence of cell attitude on gait information and takes advantage of the characteristics of people walking path in underground space.The cumulative error is corrected based on the direction of the path,and good results are achieved.This paper selects the ground floor of the teaching building as the experimentation area.When the sampling frequency is 50 Hz,the average positioning error is less than 2 m.

Key words: cell phone, inertial sensor, underground positioning, cumulative error