平方根求容积卡尔曼滤波的组合导航算法

Ship Integrated Navigation Algorithm Based on Square-Root Cubature Kalman Filter

以全球定位系统与航位推算(GPS/DR)相结合的组合导航系统为研究对象。为解决在GPS/DR组合导航研究中传统扩展卡尔曼滤波(EKF)算法的滤波精度和稳定性差问题,引入了平方根求容积卡尔曼滤波算法。作为一类sigma点滤波方法,SCKF有效避免了EKF计算雅可比矩阵,提高了估计精度。仿真结果表明,对于船用GPS/DR组合导航问题,该算法能获得更好的性能指标,更符合实际船用组合导航要求。

An integrated navigation system composed of GPS/DR（Global Positioning System/Dead Reckoning） is studied.To solve the problems of poor filtering precision and stability by conventional EKF（Extended Kalman Filter） in studying GPS/DR integrated navigation system,SCKF（Square-root Cubature Kalman Filter） is introduced.As a kind of sigma points filtering methods,it effectively avoids computing Jacobian matrix and improves estimate precision,comparing with the EKF.Simulation results indicate that better performance of GPS/DR integrated navigation system could be obtained by applying the novel algorithm,which satisfies ship integrated navigation demands in practice better.