120 / 2023-06-04 17:29:48

Heading estimation study based on indoor geomagnetic and AHDE corrected mobile phone gyroscope data

indoor localization; attitude fusion; quasi-static magnetic field detection; pedestrian heading estimation;

In recent years, indoor location-based services (LBS) have brought various conveniences in all aspects of people's daily life, especially in emergency situations, LBS can play a key role for the safety of people's lives and property. For example, in case of fire, earthquake or other disasters, indoor positioning systems can help rescuers locate and rescue trapped people quickly and accurately, improving safety and chances of survival; in addition, by monitoring the location of people and equipment in real time, managers can better manage and dispatch resources, improve efficiency and reduce energy consumption.

Mobile phone-based pedestrian dead estimation is an indispensable positioning technology in indoor positioning, but the current sensors embedded in smartphones are less expensive and generate higher signal noise, which makes the accuracy of positioning with these sensors low and unstable. In the field of inertial navigation for pedestrians, the mainstream "zero-speed" heading correction algorithms such as ZARU and ZUPT are difficult to implement due to the specificity of the position posture of the cell phone with respect to the human body. The heuristic correction algorithms such as HDE, IHDE, and AHDE are better for pedestrians walking along a straight line and in a specific dominant direction, but they are more limited to the travel route. The indoor geomagnetic heading estimation method is favored by researchers as an effective heading correction method for PDR, but a stable magnetic field environment is required to obtain an accurate heading. Based on this, this paper uses an improved quasi-static magnetic field detection (IQSF) algorithm to detect a more stable magnetic field, and then assist in correcting the heading to improve the reliability of heading estimation by using AHDE to correct the heading in an unstable geomagnetic environment. The details of the study are as follows:

1. Due to the problem of magnetic interference in the body itself, the least squares elliptic fitting method is used to correct the magnetometer to obtain reliable magnetic field vector information. The improved quasi-static magnetic field detection (IQSF) algorithm is used to identify the stable magnetic field, and the geomagnetic heading is used to assist in correcting the acceleration data of the mobile phone gyroscope with the help of filtering, and the heading of the pedestrian at this stage is estimated. Experimental analysis shows that this method improves the accuracy of the estimated heading by about 10% compared to the gyroscope acceleration data fusion.


2. It is found that the size of the existence space of non-static magnetic field region is contingent and irregular, and Indoor geomagnetism is difficult to assist in estimating heading. Therefore, with the help of the improved AHDE algorithm, the heading at this stage is corrected to meet the reliable estimation of the heading in this environment by taking the heading at the time of the disappearance of the static magnetic field as the initial heading. Comprehensive experiments show that the two heading correction methods can better compensate for the deficiencies brought by a single heading correction algorithm, The corrected heading estimation accuracy is improved by about 20%, and the restrictions on pedestrian travel routes are greatly reduced.