A Magnetic Interference Detection-Based Fusion Heading Estimation Method for Pedestrian Dead Reckoning Positioning

Abstract

In indoor positioning, smartphone-based pedestrian dead reckoning (PDR) is a popular method. It does not require any predeployed infrastructure and can provide real-time, continuous position estimation with a known initial position. However, magnetometer sensor data are susceptible to magnetic interference from metal facilities, electrical equipment, and so on, thus reducing the accuracy of heading estimation. To solve this problem, we propose a fusion estimation method for pedestrian heading based on magnetic interference detection classification. First, a support vector machine (SVM) method is used to detect magnetic interference, and an improved relief-F feature selection method is proposed based on the information coefficient and Markov blanket condition. Second, according to SVM-based magnetic interference detection results, different methods for scenarios without and with magnetic interference are proposed to improve the accuracy of pedestrian heading estimation. The experimental results show that the proposed SVM-based magnetic interference detection accuracy reaches 98%, the average error of heading estimation is less than 4°, and the average positioning error (PE) of PDR integrating the proposed heading estimation method is 0.67 and 2.75 m when the walking distance is 84 and 252 m, respectively. The above results prove the effectiveness of the proposed method.