基于双自适应容积卡尔曼滤波的西部高原铁路列车组合定位方法OA

To address the issues of high maintenance costs,insufficient accuracy,and weak anti-disturbance capability in traditional train integrated positioning systems for western plateau railways,this paper proposes a train integrated positioning method based on Dual Adaptive Square-root Cubature Kalman Filter(DASCKF)by combining pulse velocity sensors,GNSS,electronic maps,and accelerometers.Firstly,on the basis of the standard Cubature Kalman Filter(CKF),the Sage-Husa adaptive noise estimator and fading factor are introduced to dynamically adjust the system noise covariance matrix,enabling the proposed method to adapt to varying noise characteristics in complex environments and thereby improving train positioning accuracy.A dual-filter structure is employed to achieve asynchronous filtering,enhancing the filters' smoothing performance.Secondly,to address the filter divergence problem potentially caused by the Sage-Husa estimator,a square root algorithm is introduced to improve the system's anti-disturbance capability,and the filter divergence criterion is refined to further enhance system stability.Finally,an electronic map is modeled based on actual railway lines,and a simulation environment is built to validate the proposed algorithm.Simulation results demonstrate that,compared with the standard CKF,the proposed algorithm reduces the Root Mean Square Error(RMSE)in the eastward,northward,and mileage directions by 59.2％,56.3％,and 57.9％,respectively.Additionally,compared to non-adaptive filters,the proposed algorithm effectively suppresses filter divergence issues,significantly improves positioning accuracy,exhibits smooth prediction capabilities,and demonstrates a certain level of robustness.