基于求积分卡尔曼滤波的交互式多模型算法

Interacting Multiple Model Algorithm Based on Quadrature Kalman Filtering

针对非线性系统中的多模型估计问题,将求积分卡尔曼滤波算法应用到交互式多模型算法过程中,提出一种基于求积分卡尔曼滤波的交互式多模型算法。该算法不需要求取非线性方程的雅可比矩阵,且能够获得比基于不敏卡尔曼滤波的交互式多模型方法更高的滤波精度。仿真结果证明了该算法的有效性。

Aiming at the problem of multiple model estimation in nonlinear system,the Quadrature Kalman Filtering（QKF） is employed to the process of the Interacting Multiple Model（IMM） algorithm and an interacting multiple model algorithm based on QKF is proposed.The Jacobian matrix of nonlinear equations is unnecessary and a little better error performance can be obtained than that of the interacting multiple model algorithm based on unscented Kalman filtering.Simulation results show the effectiveness of the new algorithm.