【摘要】 非线性系统的状态估计问题可以通过容积卡尔曼滤波解决,本文进一步针对非线性系统,提出了基于容积卡尔曼滤波和非线性动态数据融合算法的约束容积卡尔曼滤波方法,用以处理带边界约束的非线性系统的滤波。该滤波器使所有容积点均约束在边界之内,同时依据可相信程度确定相应的容积点权重,使滤波过程充分考虑状态约束的影响,提高滤波精度。但该算法的不足之处是运算量相对较大,随着计算机运算能力的提高,这一问题可以克服。仿真结果表明:相对于容积卡尔曼滤波算法,本算法敛速度更快、收敛精度更高,鲁棒性更好。更多还原

【Abstract】 Cubature Kalman filter(CKF)was put forward to deal with nonlinear dynamic system.A means,constainted cubature kalman filter(CCKF),for nonlinear system with bounded constraints is proposed which combines the vantages of CKF and of nonlinear dynamic data reconciliation(NDDR).This technique violents the sigma points out of the box to the boundary and assigns new respective weights dependent on the probability distribution function.Besides,the state is estimated by NDDR.As a result,the constaints could be satisfied throughout the whole framework.The method shows better accuracy,better rubustness and better convergence compared to CKF.Dispite the high price of the computational cost,the matter would be overcomed with the high development of modern computers.The simulation appears that the algorithm,compared with the CKF,owns better accuracy,better rubustness and better convergence更多还原