应答器/SINS信息融合的列车定位方法研究

Research of Train Positioning Method by Balise-and-SINS's Information Fusion

针对列车在高架桥、隧道、密林等地区无法接收卫星信号的问题,采用应答器/SINS信息融合的方法对列车进行位置监测.首先建立了基于固定应答器融合SINS(捷联惯导系统)的数学模型,利用ECEF坐标系进行二者信息融合;其次采用鲁棒滤波器对融合误差进行估计,保证系统的稳定性;最后结合先验数据,采用神经网络预测出捷联惯导系统误差,修正列车定位信息.仿真实验结果表明:与应答器融合后的SINS,一定程度上抑制了定位结果的发散;在400s内不依靠卫星的情况下,也能使定位累积游走误差下降93.6%.

To address the problem that the trains cannot receive the signal from the satellite with-in viaduct,tunnel,dense forest and so on,a method based on the fusion of fixed balise and the Strapdown Inertial Navitation System is used to monitor the trains'position.Firstly,a mathematical model based on the fusion of fixed balise and SINS is established in use of ECEF coordinate system.Secondly,robust filter is used to estimate the fusion errors to ensure the stability of the system.Finally,neural network is used to predict the error caused by the SINS and correct the train positioning data in combination with the prior data.The simulation results show that the SINS integrated with the balise can restrain the divergence of the positioning results to a certain extent;and the cumulative travel error of positioning can also be reduced by 93.6%without relying on the satellite in 400s.