Error Performance Analysis of Kalman Filtering Algorithm for Non-Gaussian System
Lili Ma · Electronics Optics & Control · 2010
Non-Gaussian noise has effect on Kalman filtering accuracy,thus we measured the degree of some non-Gaussian noise distribution sequence by calculating the information distance between nonGaussian and Gaussian distribution.Kalman filtering algorithm was used in simulation experiments for random noise sequence with different distributions,including Gaussian distribution,uniform distribution,Gamma distribution and Rayleigh distribution,which had equal mean value and equal variance.And the error performance of the filtering result was studied.The result of simulation showed that: 1) the non-Gaussian degrees of the three non-Gaussian distributions mentioned above are different with each other;and 2) the error performance of nonGaussian system is basically the same as that of the Gaussian system.