Pedestrian Navigation System Based On Gait Feature Constraint

The initial pedestrian navigation system mainly relies on GPS to realize the navigation and positioning of pedestrians.However,the current indoor real-time positioning technology cannot achieve autonomous positioning and simple positioning,which brings certain difficulties to indoor positioning in harsh environments.In this paper,a system that can locate autonomously is designed to achieve rapid positioning in three-dimensional space and realize real-time tracking and positioning of firefighters.Firstly,the gait detection research based on inertial sensor is carried out for indoor positioning,the human body motion state model is constructed,the zero-speed interval is established by the three-condition method,and the median filter is used to suppress the noise to a certain extent.Then,three kinds of intelligent combined correction techniques are used to filter and correct the integrated accumulated error of inertial device navigation.Taking the velocity difference,angular velocity difference and azimuth difference as the measurement values,the error equations of the IMU are deduced,and the Kalman filtering algorithm is used to achieve information fusion.On this basis,a 21-dimensional Kalman filter model is established.Finally,the simulation experiment proves that the error of the method in indoor positioning is 2.25%.Then,according to the gait characteristic law,the six motion modes of walking,running,stationary,going up and downstairs,rapid descent and U-turn are detected,identified and constrained,so that the system can quickly and accurately identify the current motion mode.Then,the indoor and outdoor seamless navigation is designed,and the GNSS position is used to assist the correction.The stability of the system before and after the change of the GNSS signal is tested,and it is verified that the filtering and stable navigation can be performed in the presence of the GNSS signal,and then the ability of the inertial device used in the experiment to maintain stability and maintain the stability of the system is verified.Then combined with the intelligent combination correction technology in the previous chapter,the adaptive adjustment of the parameters of the position-assisted correction fusion speed-assisted correction is proposed,which mainly uses the change of the GNSS signal strength to complete the collection and filtering of the position information.Finally,the effects of the first-order RungeKutta method and the complementary filtering method on the quaternion update algorithm are compared.Finally,a complex motion path is designed to simulate the real fire fighting mode.Based on the pattern recognition,it is proposed to adaptively adjust the position filtering and correction parameters by relying on pattern recognition.The positioning error of a single filtering parameter without pattern recognition is 5.14%,and the positioning error after pattern recognition.Reduced to 3.67%,combined with complementary filtering to update the quaternion,the heading angle error is corrected,and the positioning error can be reduced to1.75%.