Indoor positioning method based on adaptive non-trace Kalman filtering

Abstract

The invention discloses an indoor positioning method based on adaptive non-trace Kalman filtering. The indoor positioning method comprises the steps that router signals received at the sampling pointare collected according to the indoor environment and combined into a fingerprint storage database, a router signal sequence is collected in real time, the sequence and a fingerprint in the database are matched, and the approximate position is estimated to be obtained; acceleration, angular acceleration and magnetic field strength data are obtained through a sensor, and the step length and direction during travelling are obtained through a step length detection algorithm and a course calculation algorithm in a track calculation algorithm correspondingly; and system equation and measurement equation models suitable for a non-trace Kalman filtering algorithm are constructed through estimated position information and step length and direction information, adaptive noise scaling factors are designed according to a deviation of an observed value and a posterior estimation value, the variance of process noise at the current moment is corrected, and finally the position is estimated through iteration updating. According to the indoor positioning method, excellent performance of filtering under the conditions of unknown models and pedestrian state variability can be achieved, and long-termand stable positioning accuracy is guaranteed.

Description

technical field [0001] The invention belongs to the field of indoor positioning algorithms, in particular to an indoor positioning method based on an adaptive unscented Kalman filter algorithm. Background technique [0002] Indoor positioning technology is indispensable in today's life. The commonly used global positioning system for outdoor positioning and navigation has weak signals and cannot guarantee the validity of information in complex indoor environments. In order to meet the needs of various indoor positioning, researchers have researched and developed Many positioning systems suitable for indoor environments range from single data source positioning to multiple ways of mutually assisting positioning, and some technologies can achieve satisfactory positioning results. [0003] The rapid development of micro-electromechanical systems has brought about the civilianization of inertial navigation systems. The dead reckoning system developed on this basis can estimate r...

AI Technical Summary

Problems solved by technology

Kalman filtering and its extended form are all based on the known model, which can achieve a good estimate for a stable system, but for an uncertain model, the filtering effect will become worse, and the deviation of the estimation result will become larger. This leads to the problem of poor positioning results in the case of a combination of multiple motions

Images