Research On Trajectory Estimation And Fusion Technology Of MEMS Inertial Positioning System

Satellite-based global positioning system has been widely used in many military and civilian fields for its features of real-time,continuous and accurate location information.However,for special application environments such as emergency rescue,indoor reconnaissance,counter-terrorism,and individual soldier operations,the satellite positioning method is difficult to use due to the blockage of satellite signals by buildings.The inertial principle-based positioning technology does not require external auxiliary information,and the inertial system alone can achieve complete autonomous positioning and navigation.In this thesis,based on the application of pedestrian inertial positioning as the application background,the MEMS gyroscope and accelerometer are used as the dominant orientation and orientation method,combined with the constraint conditions of pedestrian gait characteristics,and the trajectory fusion method is used to correct the speed and position error so as to achieve inertia positioning.According to the characteristics of the pedestrian inertial positioning and the technical difficulties in realizing the positioning,this thesis proposes an overall plan for the bipedal pedestrian-based pedestrian inertial positioning,and introduces and designs the main technology in the overall plan.On the basis of the general plan,a single-track trajectory calculation of pedestrians is studied first.The current attitude,speed,and position information of the carrier are calculated using the original strapdown inertial navigation system,and three conditions are used to comprehensively detect pedestrians' relative zero-velocity moments.The zero-velocity detection algorithm and the zero-velocity update algorithm based on kalman filtering trigger the zero-velocity update when the relative zero-velocity moment is detected.The triaxial velocity at this moment is used as the external observation of the system,and the kalman filter is used to estimate the entire system.Errors that occur over time are finally fed back to the strapdown inertial navigation system calculation module for correction,and corrected velocities and position information are used to complete the trajectory estimation of the pedestrian's single-footed navigation.Because even if the zero-velocity update is used to correct the trajectory,there is still a problem of poor system accuracy.This thesis proposes that a set of MEMS sensors is bound by the left and right feet of a pedestrian,and a dual-pedestrian trajectory fusion algorithm based on inequality constraint kalman filtering is adopted.According to zero-velocity detection method,the zero-velocity state is detected,and the maximum separation distance between the left and right feet is used as a constraint condition,and inequalities are restrained by the kalman filter to be modified to achieve dual-track fusion.Finally,in the experimental verification,the posture,speed,and position information before and after trajectory estimation are compared and analyzed.Three different trajectory routes are also used to perform trajectory fusion experiments.It is experimentally verified that in the pedestrian inertial positioning system,the bipedal trajectory fusion method is used plays a good position correction effect.