| ABSTRACT:With the continuous improvement of train speed, the environment for train positioning has become more and more complex, which requires the train positioning system should have excellent anti-jamming performance to meet the complex environment. Tightly integrated GNSS/INS positioning system realizes the complementary advantages of GNSS and INS in a high coupled degree, which improves the positioning performance effectively. In this paper, the application of tightly integrated GNSS/INS positioning system in train locating was studied.Firstly, the present status at home and abroad of tightly integrated GNSS/INS was introduced, and both the combination model and related technologies of GNSS/INS were analyzed. Considering the tracking efficiency and processing ability of train locating system, a tightly integrated GNSS/INS strategy for train locating based on vector tracking loop was proposed, and the mathematical model for GNSS/INS integrated system was established.In order to solve the nonlinear problem in tightly integrated GNSS/INS system, the concept of strong tracking filter and CKF were introduced, and a novel nonlinear fusion algorithm (STCKF) with strong robust performance was designed based on both of the strong tracking filter and CKF, while the performance of the algorithm was analyzed to verify the robustness of the algorithm.With the aiming to improve the adaptive change ability to fault data of fusion filter in tightly integrated GNSS/INS system, a fault data detection strategy based on the Takagi-Sugeno Recurrent Fuzzy Neural Network was put forward, and the fault-tolerant strategy was designed, realizing the isolation of fault data.Finally, the simulation system for tightly integrated GNSS/INS train locating was studied, and the locating performance of tightly integrated model and loosely integrated model were compared based on the simulation system, while the related performance of STCKF and the fault-tolerant control strategy were both analyzed.
|
PDF Full Text Request