Path: Common/Estimation
% Unscented Kalman Filter transform. This can be used for the measurements or the states. The weight matrix is used by the matrix form of the unscented transform. The functions fun are passed as pointers and are of the form MyFunction( x, d ) where x is an n by 2*n matrix of states or measurements. d is a data structure. If the unscented transform is used for states, fun propagates the states. For measurements it computes the measurements. The weighting factors are computed by UKFWeight. -------------------------------------------------------------------------- Form: [mU, sU, cU] = UKFUT( m, p, c, wM, w, fun, dFun ) -------------------------------------------------------------------------- ------ Inputs ------ m (n,1) Means p (n,n) Covariance matrix c (1,1) Scaling constant wM (1,2*n) Weighting array w (2*n+1,2*n+1) Weighting matrix fun (1,:) Function pointer for processing x dFun (1,1) Data structure to pass to fun ------- Outputs ------- mU (2*n+1,2*n+1) State matrix sU (2*n+1,2*n+1) Correlation matrix cU (2*n+1,2*n+1) Cross-correlation matrix --------------------------------------------------------------------------
Back to the Common Module page