Motion Pattern Recognition for Indoor Pedestrian Altitude Estimation Based on Inertial Sensor
Qikai Guo,Ming Xia,Dayu Yan,Jiale Wang,Chuang Shi,Qu Wang,Tuan Li
DOI: https://doi.org/10.1109/jsen.2024.3355163
IF: 4.3
2024-03-15
IEEE Sensors Journal
Abstract:In an intelligent information society, indoor pedestrian navigation systems based on inertial sensors are becoming increasingly important due to the advantages of autonomy and continuity in positioning. However, these systems cannot estimate pedestrian altitude because of the divergence of inertial altitude channels and the complexity of pedestrian movement patterns. To address this challenge, this article proposes an indoor pedestrian altitude estimation method utilizing the inertial measurement unit (IMU) sensor installed on the pedestrian's foot. Specifically, the method calculates the height difference of steps with the inertial navigation system (INS), extended Kalman filter (EKF), and zero velocity update (ZUPT)—named the IEZ framework. Then, a convolutional neural networks-support vector machine (CNN-SVM) model is developed to accurately identify complex pedestrian motion patterns, including horizontal walking, running, walking stairs, standing, and taking elevators and escalators. Next, the height difference is adaptively corrected according to the motion modes. For horizontal walking and standing, the height difference based on the IEZ framework is revised and remains unchanged; for walking or running stairs, the pedestrian altitude is determined by the stair step model. In the case of taking elevators or escalators, the vertical displacement is obtained by integrating the barometric height measured as EKF observations with the double integral value of the vertical acceleration. Experimental results show that the CNN-SVM model achieves a classification accuracy of 98.8%, and a positioning error for pedestrian altitude estimation of less than 0.6 m with an indoor walking distance of approximately 500 m.
engineering, electrical & electronic,instruments & instrumentation,physics, applied