A new method for the nonlinear transformation of means and covariances in filters and estimators

  title={A new method for the nonlinear transformation of means and covariances in filters and estimators},
  author={Simon J. Julier and Jeffrey K. Uhlmann and Hugh F. Durrant-Whyte},
  journal={IEEE Trans. Autom. Control.},
This paper describes a new approach for generalizing the Kalman filter to nonlinear systems. A set of samples are used to parametrize the mean and covariance of a (not necessarily Gaussian) probability distribution. The method yields a filter that is more accurate than an extended Kalman filter (EKF) and easier to implement than an EKF or a Gauss second-order filter. Its effectiveness is demonstrated using an example. 

Figures from this paper

Kalman Filtering for a Generalized Class of Nonlinear Systems and a New Gaussian Quadrature Technique

The class of nonlinear systems treated in this technical note consists of the discrete time non linear systems that are formed by the interconnection of linear systems through static nonlinearities with few inputs, and a new quadrature scheme suitable for nonlinear Kalman filtering is introduced.

A derivative-free implementation of the extended Kalman filter

Computationally Efficient Kalman Filtering for a Class of Nonlinear Systems

The proposed recursive state estimation method for the class of discrete time nonlinear systems whose nonlinearity consists of one or more static nonlinear one-variable functions yields good results with small computational cost.

Comparison of standard and modified recursive state estimation techniques for nonlinear systems

A new set of σ-points for the Unscented Kalman Filter is proposed as well as a way to substitute a nonlinear output with a linear one and the results show that the modifications proposed in some cases lead to considerable reduction in estimation error.

A New computation reduction based nonlinear Kalman filter

This article introduces a new algorithm for nonlinear state estimation based on deterministic sigma point and EKF linearized framework for priori mean and covariance respectively. This method reduces

An Iterative Ensemble Kalman Filter

An essential stopping criteria is introduced for the proposed iterative extension to the ensemble Kalman filter to improve the estimates in the cases where the relationship between the model and observations is not linear.

A method on parameter estimation of nonlinear systems

It is shown that the minimizer based on Newton-method becomes to iterated extended Kalman filter (IEKF) and a method on the gain adjustment of the MAP estimator is given so as to improve the performance of it.

Development and numerical investigation of new non-linear Kalman filter variants

This study deals with recursive state estimation for non-linear systems. A new set of σ-points for the unscented Kalman filter is proposed as well as a way to substitute a non-linear output with a

Performance analysis of deterministic sampling filters

This paper deals with a type of nonlinear filters. The deterministic sampling filters (DSFs), including the unscented Kalman filter (UKF) and the cubature Kalman filter (CKF), which use a set of



A new approach for filtering nonlinear systems

A new recursive linear estimator for filtering systems with nonlinear process and observation models which can be transformed directly by the system equations to give predictions of the transformed mean and covariance is described.

Novel approach to nonlinear/non-Gaussian Bayesian state estimation

An algorithm, the bootstrap filter, is proposed for implementing recursive Bayesian filters, represented as a set of random samples, which are updated and propagated by the algorithm.

A General Method for Approximating Nonlinear Transformations of Probability Distributions

A new approach for generalised nonlinear ltering is described, which is more accurate, more stable, and far easier to implement than an extended Kalman lter.

New extension of the Kalman filter to nonlinear systems

It is argued that the ease of implementation and more accurate estimation features of the new filter recommend its use over the EKF in virtually all applications.

The independance of forward and backward estimation errors in the two-filter form of the fixed interval Kalman smoother

In this paper it is shown that the Fraser-Potter two-filter form of the fixed interval Kalman smoother can be derived without assuming that the forward and backward estimation errors are independent.

New nonlinear iterated filter with applications to target tracking

This paper identifies inadequacies inherent to the EKF and IKF, discusses their detrimental effect on performance, and proposes a solution which uses the Julier et al. time update and a new iterated procedure for computing the measurement update.

Implicit Jacobian for linearised state estimation in nonlinear systems

This article describes a method of implementation linearised approximations to nonlinear systems that does not require the direct derivation of Jacobian matrices. The approach is simpler to implement

A finite-difference approach to linearization in nonlinear estimation algorithms

  • T. Schei
  • Mathematics
    Proceedings of 1995 American Control Conference - ACC'95
  • 1995
Linearizations of nonlinear functions that are based on Jacobian matrices can often not be applied in practical applications of nonlinear estimation techniques. An alternative linearization method is


The problem of estimating from noisy measurement data the state of a dynamical system described by non-linear difference equations is considered and a Bayesian approach is suggested in which the density function for the state conditioned upon the available measurement data is computed recursively.

Suboptimal state estimation for continuous-time nonlinear systems from discrete noisy measurements

This paper presents the derivation of the dynamical equations of a second-order filter which estimates the states of a non-linear plant on the basis of discrete noisy measurements. The filter