Study On Pedestrain Indoor Positioning Based On Inertial Navigation

Location-based services make people's travel more and more convenient,but with the development of urbanization,the increasing number of complex structures makes indoor positioning increasingly important.As an autonomous positioning system,an indoor positioning system based on inertial sensors has the advantages of strong concealment,stable performance,anti-interference,and no restrictions on the use of time and space.Since wearable technology can provide environmental sensing and experimental data based on the integration of various sensors in the wearable device,it is also a research hotspot in the field of pedestrian indoor positioning.Positioning personnel wear wearable devices to achieve autonomous positioning of personnel.This article is aimed at indoor rescue,smart travel and other application scenarios.Based on the cumulative error characteristics of a pure inertial navigation positioning system,a wearable device fixed at the ankle of a pedestrian is designed.The inertial navigation in this paper uses pedestrian dead reckoning to achieve positioning.Combining the characteristics of pedestrian foot movements,the key factors that affect positioning accuracy such as step identification,step length calculation,and heading angle accuracy are studied to achieve autonomous positioning of pedestrians.The specific work of this article is as follows:1.Design the software and hardware parts of the pedestrian indoor positioning system.The hardware part mainly includes the wearable device's motion parameter and environmental parameter acquisition module,and the wireless data transmission modular design.The software part mainly includes the design of the sensor driver program,the design of the data transmission process between the modules,the design of the wireless transmission protocol,and the software design of the three-dimensional building model where the host computer displays the pedestrian positioning.2 Fusion preprocess the inertial sensor data integrated in the two wearable devices,and study the attitude angle solution based on Kalman filter to reduce the error of the inertial sensor.Based on the analysis of the characteristics of pedestrians,a step recognition method based on dual inertial measurement unit was proposed,which combined the attitude and angle transformation of the two legs and the acceleration information for joint gait recognition.3 Analyze the existing step estimation model,and propose a linear regression based step estimation model suitable for wearable devices with slow computing speed and a BP neural network step estimation model suitable for computer terminal with fast computing speed.The model makes full use of the inertial parameters of pedestrians in the process of travel,greatly improving the accuracy of step calculation.The use of barometric pressure sensors to detect the floor where pedestrians are located and the solution of plane coordinates based on pedestrian dead reckoning are studied.4 The experiment verifies the positioning accuracy of the indoor positioning system based on the inertial sensor,and randomly selects multiple experimenters to perform indoor and outdoor walking experiments.The experimental results show that the positioning method described in this paper has high positioning accuracy,and the software and hardware design of this solution is reasonable.The system has stable performance and has broad application prospects.