| As an important navigation parameter,attitude has received more and more attention with the rise of technologies such as intelligent driving,drones and unmanned vehicles.Among the attitude determination technologies based on different sensors,GNSS(Global Navigation Satellite System)attitude determination has the advantages of all-weather,high accuracy,and errors that do not accumulate over time,but in GNSS harsh-signal environments,accuracy,continuity,and reliability will rapidly decrease,while Attitude determination based MEMS-IMU(Micro-electro-mechanical Systems-Inertial Measurment Unit,Referred to as MEMS)has the advantages of low cost,small size,and autonomous passive.Information fusion between GNSS and MEMS can achieve complementary advantages and obtain better attitude determination results.According to the number of GNSS antennas,the fusion of GNSS and MEMS can be divided into three different methods of single/dual/multi-antenna GNSS/MEMS integrated attitude determination.Among them,the single-antenna GNSS/MEMS integration is greatly affected by the different motion conditions and the sideslip angle error.and the multi-antenna GNSS/MEMS integration has higher complexity and cost.But,the dual-antenna GNSS/MEMS integration can avoid the disadvantages of the above two methods.At the same time,it combines the advantages of both GNSS and MEMS sensors,so it has more important research significance and application value in the attitude determination field.Based on that,this paper has conducted a detailed study on the dual-antenna GNSS / MEMS integrated attitude determination algorithm and key technologies.The main work is as follows:1)The attitude determination technologies based on dual-antenna GNSS and MEMS is studied and the corresponding error calculation formula is deduced.The advantages and limitations of attitude determination technologies based on different sensors are explained.Accordingly,the importance of dual-antenna GNSS/MEMS integration is theoretically demonstrated.2)Based on the EKF(extended Kalman Filter)and error state filtering method,the mathematical model of the dual-antenna GNSS/MEMS loosely/tightly coupled attitude determination algorithm with misalignment based state model is derived in detail.By comparing the two different algorithm structures,the importance of ambiguity fixing for them and the reason why the tightly coupled structure is more advantageous in the GNSS challenge environment are explained theoretically.3)According to the distribution of MEMS sensor noise in the frequency domain,a Butterworth low passfilter with sliding window is designed.It has a small calculation amount and is easy to implement,which can effectively reduce the influence of sensor noise on the horizontal angle accuracy.4)Using a set of long-term static data,the Allan variance analysis,a random noise modeling and process noise extraction tool,is described and the MEMS error characteristics are futher analyzed in detail.The results show that the random noise of MEMS-IMU is mainly consists of white noise and flicker noise in a short time.5)The dual-antenna GNSS/MEMS loosely/tightly coupled attitude determination algorithm are implemented,and the two algorithms are verified using two dynamic vehicular tests data under different GNSS environments.Under the GNSS open environment,the accuracy of the two algorithms are(0.15,0.12,0.11)degree and(0.14,0.13,0.09)degree and it has little difference between them;Under the GNSS challenge environment,the accuracy of the two algorithms are(0.28,0.35,0.27)degree and(0.21,0.27,0.26)degree.Compared with the loosely coupled algorithm,the heading accuracy of the tightly coupled algorithm is improved by 25%.The above results show that the dual-antenna GNSS/MEMS integrated attitude determination algorithms in this paper can obtain continuous,high-precision and reliable attitude results under different environments.Moreover,the tightly coupled algorithm has more advantages under the GNSS challenge environment. |
PDF Full Text Request