Magnetic Field Positioning Technology For Indoor Moving Targets

In daily life,positioning services in outdoor environment are usually provided by satellite positioning systems.Countries and organizations such as the United States,China,Russia and the European Union have independently developed satellite navigation systems.However,there are many high-rise buildings in the community,and there are many cement and steel bars in the indoor environment.It is difficult for satellite signals to pass through the wall to complete the positioning target.Therefore,the positioning software based on satellite navigation is only suitable for providing accurate positioning and navigation services outdoors.As people's demand for location services in indoor environment increases year by year,researchers have carried out research on indoor location technology,and have made many achievements,which are based on bluetooth,ultra-wideband,visible light and so on.Among them,because geomagnetic positioning technology does not rely on other infrastructure,and it is low cost and high positioning accuracy,it provides a direction for many researchers to study indoor positioning technology without infrastructure.Aiming at the geomagnetic positioning technology without infrastructure in static indoor environment,using the characteristic that the magnetic field signal meets the Gaussian distribution,this paper designs the magnetic signal noise reduction processing method based on Kalman filter and the magnetic field fingerprint matching algorithm about the numerical range of magnetic field intensity.The main work of this paper includes:(1)The time-varying and spatial distribution characteristics of static indoor magnetic field signal are collected and analyzed,and the stability of geomagnetic signal as fingerprint is explored,as well as the influence of surrounding environment changes on geomagnetic fingerprint.(2)The magnetic field signal processing method based on wavelet analysis and traditional Karman filtering is designed.Using wavelet analysis to refine the analysis in two domains in time and frequency,the measurement noise and process noise during the detection of the sensor is acquired;the statistical characteristics of the two noise signals are analyzed,and the average value of the noise signal,the variance brought into the equation of Kalman filter;finally use Kalman filter to reduce the original magnetic field signal collected by the sensor.Simulation experiments have shown that this method has better noise reduction effect than wavelet analysis methods.(3)The magnetic field matching algorithm based on multi-constraints is designed.According to the magnetic sensor,the signal is acquired from the Gaussian distribution,the magnetic field fingerprint of the reference point follows the 3-sigma principle,that is?-3??H??+3?,the magnetic field component to be positioned with the reference point corresponds to the magnetic field component,and the difference between the three-axis magnetic field strength is reference points in the error range are recorded as a matching point.Then calculate the distance between all match points and the upper site or the starting point point,and finally determine the matching point coordinates by the distance to retrofitting position between the distance to be positioned.The experimental results prove that the positioning error of the algorithm decreases with the reduction of the error range of magnetic field,and the magnetic field error is set to 3 ?,about 10%points occur the positioning error.