Research On The Key Technology Of Pedestrian Navigation System Based On Kinematics Assistance

Pedestrian navigation,as an important area of navigation and positioning technology development in recent years,has gradually become a hot research direction.Most of the current pedestrian navigation technologies mainly rely on GNSS.However,under the influence of GNSS signals such as occlusion and interference,navigation positioning accuracy is poor.In view of the above insufficiency,stable and reliable position information detection needs to be implemented with other navigation and positioning means.Pedestrian navigation systems based on inertial technology have received extensive attention from domestic and foreign scholars and focused on the advantages of their autonomous navigation and without the need for additional infrastructure.This article focuses on the key technologies of pedestrian navigation systems under GNSS failure scenarios.It covers the identification and correction of micro-inertial devices and system errors,the construction and analysis of the human body’s lower limb kinematics model,the construction of virtual micro-inertial devices based on machine learning,and the fault detection algorithm of micro-inertial devices.The specific research content is as follows:1)Pedestrian navigation and positioning system based on micro-inertial devices foot installation.For the problem of the effectiveness of low-cost inertial devices in pedestrian navigation applications,the accuracy of micro-inertial devices was analyzed.On the basis of analyzing the gait characteristics of pedestrians,a zero-speed detection algorithm with zero-speed multi-conditions is designed,and an adaptive threshold is set for different indoor and outdoor environments.The Kalman filter with velocity error was constructed to estimate the error and fed back to the inertial navigation system,which improved the positioning accuracy.Verified by measured data,this algorithm can improve the reliability of zero-speed detection and suppress the accumulation of positioning error.2)The construction of virtual micro-inertial devices based on human lower limb inertial information model.Aiming at the problem that the micro-inertia devices foot installation method cannot be measured effectively in human body’s high overload motion,a method based on BP neural network to construct the human body’s lower limb kinematics model is proposed.This method uses a BP neural network to approximate the nonlinear kinematics between the acceleration of the lower limbs and the angular velocity,and realizes the construction of virtual micro-inertial devices.The test results show that the method studied in this paper can effectively simulate the output of the inertial devices at the position of the centroid of the foot through micro-inertial devices installed near the hip joint.3)The reconstruction of distributed pedestrian navigation system based on fault detection scheme of micro-inertial devices.Aiming at the problem that the foot installation of the micro-inertia devices can’t effectively achieve accurate navigation in case of fault,a fault detection algorithm based on BP neural network is studied.This method uses BP neural network training model to predict information of virtual micro inertial devices and compares the difference between the predicted value and the measured value as a fault detection strategy.The test results show that the fault detection algorithm can effectively detect fault axial and fault information,and can realize reconstruction of distributed pedestrian navigation system.In order to verify the effectiveness of the above-mentioned key technologies for pedestrian navigation systems,hardware platforms and software modules were designed and implemented to verify the function of position estimation based on key technical assistance.After many tests,the positioning accuracy of the system is less than 2%of the total travel distance under different fault axis and fault information.