A New Extension of the Kalman Filter to Nonlinear Systems-英文文献.pdf

A New Extension of the Kalman Filter to Nonlinear Systems Simon J Julier Jerey K Uhlmann sijurobotsoxacuk uhlmannrobotsoxacuk The Rob otics Research Group Department of Engineering Science The University of Oxford Oxford OX PJ UK Phone Fax ABSTRACT The Kalman lterKF is one of the most widely used metho ds for tracking and estimation due to its simplicity optimality tractability and robustness However the application of the KF to nonlinear systems can b e dicult The most common approach is to use the Extended Kalman Filter EKF which simply linearises all nonlinear mo dels so that the traditional linear Kalman lter can b e applied Although the EKF in its many forms is a widely used ltering strategy over thirty years of exp erience with it has led to a general consensus within the tracking and control community that it is dicult to implement dicult to tune and only reliable for systems which are almost linear on the time scale of the up date intervals In this pap er a new linear estimator is develop ed and demonstrated Using the principle that a set of discretely sampled p oints can b e used to parameterise mean and covariance the estimator yields p erformance equivalent to the KF for linear systems yet generalises elegantly to nonlinear systems without the linearisation steps required by the EKF We show analytically that the exp ected p erformance of the new approach is sup erior to that of the EKF and in fact is directly comparable to that of the second order Gauss lter The metho d is not restricted to assuming that the distributions of noise sources are Gaussian We argue that the ease o