基于鲁棒CKF的多传感器全信息融合算法

Multi-sensor all information fusion algorithm based on robust CKF

为了克服不良测量的影响,改善多传感器的融合性能,提出一种基于鲁棒容积卡尔曼滤波(CKF)的多传感器全信息融合算法。基于新息协方差匹配原理构建鲁棒CKF,定义数据质量检测函数,根据测量数据质量选择鲁棒CKF或标准CKF作为子系统最优滤波算法。基于多传感器融合信息,建立子系统软故障检测算法;定义子系统故障系数,通过系统重构实现故障子系统的隔离。利用多传感器系统所能提供的最多信息,建立全信息融合算法。将所建算法应用于船舶动力定位测量系统的仿真实验中,与CKF、局部估计加权融合算法进行比较。仿真结果表明,鲁棒CKF及软故障检测函数提高了子系统的滤波鲁棒性,全信息融合算法进一步改善了系统的融合性能。仿真结果验证了所建算法的有效性。

To overcome the influence of abnormal measurements and improve the performance of multi- sensor fusion, based on robust cubature Kalman filter （CKF） a multi-sensor all information fusion algo- rithm is proposed. By the principle of innovation covariance matching, a robust CKF was built. A data quality detection function was defined. According to the measurements, the robust CKF or normal CKF was selected as the subsystem optimal filtering algorithm. With the estimation information of multi-sensor fusion, a subsystem soft fault detection function was presented. The subsystem fault factor was intro- duced, and the faulty subsystem is isolated by the system reconfiguration. A multi-sensor all information fusion method was proposed based on the all information of the system. These proposed methods were ap- plied to the vessel dynamic positioning system simulation. They were compared with normal CKF and lo- cal estimation weighted fusion algorithm. The simulation results show that the proposed robust CKF and the soft fault detection function improve the robustness and accuracy of subsystem filtering, and the all in- formation fusion algorithm has better performance. The simulation example verifies the effectiveness of the proposed algorithms.