列车自适应交互多模型H∞滤波定位算法

Adaptive Interactive Multi-model H∞Filter Positioning Algorithm for Train

针对在列车运行的复杂工况下噪声特性难以准确获知和状态模型与实际运行不匹配的问题,提出了列车自适应交互多模型H∞滤波算法。首先,依据列车运动状态自适应调整系统状态噪声特性和模型。其次,通过自适应补偿误差协方差增强H∞滤波的鲁棒性。最后,以哈尔滨到大庆的部分实测数据对算法的有效性进行了仿真验证。结果表明:所提算法具有自适应调整噪声特性及模型的能力,在复杂运动适应性和鲁棒性方面具有优势,曲线轨道段优势明显,对于列车定位具有一定的理论指导意义。

Aiming at the problem that the noise characteristics of train operation are difficult to be accurately known and the state model does not match the actual operation,an adaptive interactive multi-model H∞filtering algorithm about train is proposed.First,the system state noise characteristics and model are adaptively adjusted according to the train motion state.Secondly,the robustness of H∞filtering is enhanced by adaptively compensating for error covariance.Finally,the effectiveness of the algorithm is verified by some measured data from Harbin to Daqing.The results show that the proposed algorithm has the ability to adaptively adjust the noise characteristics and model,and has advantages in the complex motion adaptability and robustness,especially in the curved orbital part.The algorithm has certain theoretical guiding significance for train positioning.