Research And Implementation Of RTK/MEMS Fusion Positioning Algorithm

With the arrival of artificial intelligence and internet of everything,high-precision location services have become an indispensable factor for social development.Especially high-tech products,such as unmanned vehicles and UAV,appear in human life and attract people's close attention immediately.Undoubtedly there are huge market prospects and great research value on the high-precision positioning algorithms.The RTK high-precision positioning algorithm is subject to the signal interference of satellite,such as high buildings,shaded trees,which often causes large positioning errors and low reliability.This thesis proposes an RTK/MEMS fusion positioning scheme that uses MEMS advantages of short-term high-precision and strong anti-interference ability to assist RTK algorithm to calculate position solution,and the system software is designed to verify the performance of proposed algorithm effectively.The RTK/MEMS fusion positioning algorithm proposed in this thesis is based on the idea of tight coupling and uses the Kalman filtering to estimate the optimal state vector,which is mainly composed of state predication,measurement update and ambiguity fixing.Firstly,the system state equation is designed based on the MEMS data and the gradient descent algorithm is used to solve the real-time attitude.The INS mechanical orchestration is achieved through the attitude rotation matrix and the prediction information of state vector such as position,velocity,attitude is calculated.Then the double-difference observation vector is established by the pseudorange,carrier phase,and doppler observation data of the rover receiver and base station.Next using the Kalman filtering to update the state vector and obtain more accurate floating solutions and covariance matrices.Finally,searching the ambiguity with LAMBDA algorithm to obtain the optimal fixing solution.The chi-square distribution is used to verify the fixing solution to ensure that the system is not affected by the influence of gross errors and noise.Finally,this thesis designs the RTK/MEMS fusion positioning system software,completes the programming of each algorithm module with C++ language.The vehicle trajectory test results show that the positioning system performance,such as the integer ambiguity fixing rate and positioning accuracy,has been greatly improved.The integer ambiguity fixing rate has been increased by about 25% and the positioning accuracy has increased from the sub-meter level to the centimeter level,which has great engineering practice value.