文件列表:

ReBEL-0.2.6\Bugfixes (3476, 2004-02-05)
ReBEL-0.2.6\ChangeLog (5380, 2004-02-05)
ReBEL-0.2.6\core (0, 2004-02-05)
ReBEL-0.2.6\core\addangle.m (1045, 2002-09-22)
ReBEL-0.2.6\core\addrelpath.m (1284, 2002-06-06)
ReBEL-0.2.6\core\cdkf.m (14744, 2002-09-22)
ReBEL-0.2.6\core\checkdups.m (919, 2002-06-06)
ReBEL-0.2.6\core\checkstructfields.m (1355, 2002-06-06)
ReBEL-0.2.6\core\consistent.m (4541, 2002-06-06)
ReBEL-0.2.6\core\Contents.m (3639, 2002-07-31)
ReBEL-0.2.6\core\convgausns.m (4343, 2002-07-15)
ReBEL-0.2.6\core\cvecrep.m (853, 2002-07-30)
ReBEL-0.2.6\core\datamat.m (1171, 2002-06-07)
ReBEL-0.2.6\core\ekf.m (8335, 2002-09-22)
ReBEL-0.2.6\core\fixinfds.m (1954, 2002-06-07)
ReBEL-0.2.6\core\gauseval.m (2850, 2002-07-31)
ReBEL-0.2.6\core\gaussample.m (1624, 2002-07-31)
ReBEL-0.2.6\core\geninfds.m (85597, 2004-02-05)
ReBEL-0.2.6\core\gennoiseds.m (28628, 2002-07-15)
ReBEL-0.2.6\core\gensysnoiseds.m (30341, 2004-02-02)
ReBEL-0.2.6\core\gmmfit.m (7179, 2003-02-25)
ReBEL-0.2.6\core\gmminitialize.m (4649, 2003-02-25)
ReBEL-0.2.6\core\gmmprobability.m (5105, 2003-02-25)
ReBEL-0.2.6\core\gmmsample.m (3000, 2003-02-25)
ReBEL-0.2.6\core\gmsppf.m (10653, 2003-02-25)
ReBEL-0.2.6\core\gmsppf2.m (3505, 2003-02-26)
ReBEL-0.2.6\core\gspf.m (9737, 2002-07-25)
ReBEL-0.2.6\core\gssm.m (28805, 2002-09-03)
ReBEL-0.2.6\core\kf.m (8790, 2002-09-22)
ReBEL-0.2.6\core\mlpff.m (2674, 2002-06-06)
ReBEL-0.2.6\core\mlpindexgen.m (2787, 2002-06-06)
ReBEL-0.2.6\core\mlpjacobian.m (7592, 2002-07-25)
ReBEL-0.2.6\core\mlppack.m (2490, 2002-06-06)
ReBEL-0.2.6\core\mlpunpack.m (3499, 2002-06-06)
ReBEL-0.2.6\core\mlpweightinit.m (3065, 2002-06-06)
ReBEL-0.2.6\core\pf.m (7638, 2002-09-22)
ReBEL-0.2.6\core\remrelpath.m (1281, 2002-06-06)
ReBEL-0.2.6\core\residualresample.m (2268, 2002-06-12)
ReBEL-0.2.6\core\rvecrep.m (851, 2002-07-30)
ReBEL-0.2.6\core\sppf.m (10244, 2002-09-05)
... ...

-- ReBEL : Recursive Bayesian Estimation Library -- A Matlab toolkit for Recursive Bayesian Estimation Copyright 2002, Rudolph van der Merwe 1) WHAT IS ReBEL ? ReBEL is a Matlab toolkit of functions and scripts, designed to facilitate sequential Bayesian inference (estimation) in general state space models. This software consolidates research on new methods for recursive Bayesian estimation and Kalman filtering by Rudolph van der Merwe and Eric A. Wan. The code is developed and maintained by Rudolph van der Merwe at the OGI School of Science & Engineering at OHSU (Oregon Health & Science University). ReBEL is a 'work in progress' and currently contains most of the following functional units which can be used for state-, parameter- and joint-estimation: * Kalman filter * Extended Kalman filter * Sigma-Point Kalman filters (SPKF) o Unscented Kalman filter (UKF) o Central difference Kalman filter (CDKF) o Square-root SPKFs (SRUKF, SRCDKF) o Gaussian mixture SPKFs (not implemented yet) * Particle filters o Generic particle filter o Sigma-Point particle filter o Gaussian sum particle filter o Gaussian mixture sigma-point particle filter (beta) o Auxiliary sigma-point particle filter (alpha) The code is designed to be as general, modular and extensible as possible, while at the same time trying to be as computationally efficient as possible. It has been tested with Matlab 6.1 (R12.1) and Matlab 6.5 (R13_pre-release). 2) INSTALLATION Unpack the ReBEL distribution archive using either 'tar -xzvf ReBEL-X.Y.Z.tar.gz' under Unix or unzip the ReBEL-X.Y.Z.zip archive under Windows. This will create the 'ReBEL-X.Y.Z' directory tree in the current directory. 3) UPDATE MATLABPATH : Add 'core' and 'netlab' directories Make sure the 'ReBEL-X.Y.Z/core/' subdirectory is in the Matlab path by either setting the MATLABPATH environment variable accordingly or using the 'path' command from the Matlab command line. (X = major release number , Y = minor release number , Z = buxfix version number) In order to use the bundeled Netlab neural network toolkit, also add the 'ReBEL-X.X/netlab/' subdirectory to your MATLABPATH. This is needed to use any of the GMM based particle filter and hybrid algorithms. 4) DOCUMENTATION Since ReBEL is an constantly 'evolving' research tool used in the Machine Learning and Signal Processing Group ( cslu.ece.ogi.edu/mlsp/ ), the documentation usually lags behind a bit. For the most up to date documentation, refer to the online documentation section of the ReBEL project website (see below). A small 'Quickstart Guide' is also provided. This together with the (hopefully) well documented code examples in the 'examples' subdirectory will serve as the initial minimal set of documentation for the alpha release of ReBEL. The unofficial accompanying "text" to ReBEL are Chapters 5 and 7 of "Kalman Filtering and Neural Networks" edited by Simon Haykin and published by Wiley. These cover dual extended Kalman filter methods and the Unscented Kalman Filter in great detail. Other algorithms included in ReBEL are also covered in the rest of the book. See the ReBEL website for more detail. 5) WEBSITE For latest information and online documentation, always visit the project website at http://cslu.ece.ogi.edu/mlsp/rebel/ 6) MAILING LIST AND USER DISCUSSION BOARD There is a discussion group for users and developers of ReBEL hosted at YahooGroups. Here is the URL: http://groups.yahoo.com/group/rebel_toolkit/ ----------------------------------------------------------------------- This document was last modified on : September 22nd, 2002.

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：