Research On Vehicle Dynamics System Aided Low Cost Vehicle Integrated Navigation System

Nowadays,land vehicle integrated navigation and positioning system has been deeply studied,which is mainly aimed at improving the accuracy of autonomous navigation system while reducing its cost.Therefore,the main research contents of this thesis focus on the research of multi-degree of freedom vehicle dynamics model(VDM)and VDM aided inertial navigation information fusion algorithm.A nine-degree of freedom VDM aided integrated navigation system is designed based on the application of low-cost Micro-Electro-Mechanical System Inertial Measurement Unit(MEMS IMU)and combined with common on-board sensors,such as wheel speed sensor(WSS),steering angle sensor(SAS)and magnetometer(MAG).The SINS/VDM/MAG integrated system is built by Adaptive Federated Kalman Filter(AFKF)and the actual vehicle experiments verify that it has high navigation accuracy and fault tolerance performance.The main contents of this paper are as follows:The vehicle inertial navigation positioning algorithm is studied,and the error updating equations are deduced.Based on these theories,a Kalman filter state equation is established.The 9-DOF dynamic vehicle model is established which selects Dugoff tire model and considers the characteristics of vehicle wheel speed sensor and steering wheel angle sensor.To verify the accuracy of the dynamic model,a simulation experiment is designed based on the Carsim platform.Comparing the simulation results,the longitudinal and lateral vehicle speed errors of the 9-DOF vehicle model and the Carsim output are within 0.5m/s;the average errors of the lateral angle and yaw angle are controlled within 0.02° and 4° respectively.By these results,it can be explained that the established dynamic model can estimate the vehicle state parameters accurately.Adaptive Federal Kalman Filter(AFKF)has a good application in multi-sensor data fusion.Based on AFKF,this paper designs a SINS/VDM/MAG integrated navigation system.This system uses SINS as the reference system and designs two local systems which are SINS/VDM and SINS/MAG subsystems respectively.The input of SINS/VDM subsystem are the output data of the IMU,wheel speed sensor(WSS),and steering wheel angle sensor(SAS),and the output information are vehicle yaw rate,lateral speed,and longitudinal speed.The vehicle dynamics model also uses the roll angle and pitch angle which are estimated by SINS as known inputs to enhance its accuracy in estimating vehicle state parameters.Magnetometer(MAG)is a commonly used sensor in vehicle navigation.It is used to establish SINS/MAG subsystem to enhance the accuracy of yaw angle estimation and promote system fault tolerance performance.In order to verify the accuracy of the SINS/VDM/MAG system,two group actual vehicle comparison experiments were designed.One is SINS/VDM/MAG and SINS/WSS/MAG comparison experiment,and the other is SINS/VDM/MAG and SINS/OD comparison experiment.The results of experiment which lasts for 250s show that the MSE of pitch,roll and yaw angle between the reference value and the AFKF filtered output are 0.42°,0.24° and 0.48° respectively.The MSE of the east and north directions is 0.2m/and 0.25 m/s respectively,and the MSE of the east and north directions are reduced to less than 5m.The actual vehicle experiments results verify that the proposed vehicle dynamics-aided low-cost vehicle integrated navigation system has better navigation estimation accuracy,and the fault-tolerant experiment proves that the SINS/VDM/MAG integrated system has better fault-tolerant performance.