A novel constant gain Kalman filter applied to nonlinear system

作者：Liu Bing; Chen Zhen; Liu Xiangdong

来源：ICIC Express Letters, 2014, 8(7): 1949-1954.

• Computational burden
• Computational costs
• Constant coefficients
• Constant gain
• Linear differential inclusions
• Linear modeling
• Linearization errors
• Practical engineering applications

Motivated by improving the computational efficiency of the nonlinear filter for practical engineering application, a novel constant gain Extended Kalman filter is proposed in the paper. An uncertain term is firstly introduced to take into account of the linearization errors of the system model. Then the estimation error system is represented by an polytopic linear model according to polytopic linear differential inclusion theorem. Subsequently, the rectification equations are designed with constant coefficients. Finally, the state estimations are given by updating the state predictions with the rectified quantities. The evident advantage of the proposed filter is that it does not need to evaluate the Jacobian matrixes and update the filter gain online, resulting in less computational cost and much more simplified implementation. The convergent speed of the proposed filter is much faster than that of the existing constant Kalman filter. Its excellent properties are demonstrated by using an example.

• 出版日期2014
• 单位北京理工大学