Research On SINS/GNSS/NHC/VKM Vehicle-mounted Integrated Navigation Syste

The existing vehicle navigation systems usually use the integrated navigation system of Strapdown Inertial Navigation System(SINS)and Global Navigation Satellite System(GNSS).The combination of SINS and GNSS can fully exploit the advantages of the two systems and complement each other,providing long time,high accuracy and high frequency navigation information.However,it is susceptible to the influence of the environments where satellite signals are obscured.In order to optimize the SINS/GNSS vehicle integrated navigation system and solve the problem of reduced navigation accuracy caused by satellite signal reception obstruction,this paper conducts research work in three aspects: vehicle navigation information fusion technology,inertial device error compensation methods,and building a compensatory integrated navigation system for GNSS failure conditions.The main research contents are as follows:1)Design an Unscented Kalman filter algorithm incorporating innovation adaptive estimation strategies.Aiming at the problem that the Unscented Kalman Filter(UKF)algorithm is not fully applicable to vehicle navigation systems where the statistical characteristics of noise are unclear due to the varying driving scenarios and motion states,an exponential fading memory weighted average recursive algorithm for measuring the noise matrix is established based on the innovations in the Kalman filter.This algorithm is incorporated into UKF to obtain the Adaptive Unscented Kalman Filter(AUKF)algorithm,which enables the online adjustment of the measurement noise matrix during the filtering process.Through comparison experiments of filtering algorithms,AUKF improves the position accuracy by 22.11% and the heading angle accuracy by 8.37%compared with UKF,and has better robustness in complex driving sections navigation.2)An error modeling method for inertial devices based on noise statistical feature restoration is proposed.In order to mitigate the impact of Inertial Measurement Unit(IMU)device errors on navigation system accuracy,inertial device errors are modeled and compensated for.The traditional power spectral density modelling method is weak in characterising the low-frequency noise of the device.To analyze the relationship between inertial device error and performance indicators,it is recommended to restore device high-frequency noise error through power spectral density modeling firstly,and then restore the low-frequency noise error based on the performance calculation method corresponding to the statistical characteristics of low-frequency noise.Comparative experiments with existing power spectral density modeling methods show that the proposed modeling method has stronger ability to characterize device errors and better accuracy in device error compensation.3)Construct a SINS/GNSS/NHC/VKM vehicle integrated navigation framework that integrates Non-holonomic Constrain(NHC)and Vehicle Kinematics Model(VKM).The GNSS failure judgment algorithm is designed based on the number of satellites received by the satellite receiver and the GNSS Position Dilution of Precision(PDOP).When GNSS is valid,SING/GNSS loose integrated navigation is used.When GNSS is invalid,SINS/NHC/VKM integrated navigation is adopted.VKM calculates the twodimensional position and heading angle of the vehicle in real time based on the IMU accelerometer output and the steering wheel angle sensor information.And the NHC provides pseudo measured values of vehicle lateral and skyward velocity under specific motion states.According to the IMU gyro output and vehicle speed sensor information,the vehicle motion state is judged in real time.NHC and VKM provide different observations for Kalman filter estimation and correction of navigation errors in different motion states.Combining the designed AUKF algorithm and inertial device error modeling method,the real vehicle experiment of SINS/GNSS/NHC/VKM integrated navigation system shows that SINS/NHC/VKM integrated navigation can effectively replace SINS/GNSS integrated navigation in the case of GNSS failure,with a position accuracy of 0.4352 m.