20 机器人定位中的EKF-SLAM算法，用扩展卡尔曼滤波的方法实现同时定位和地图构建.zip

EKF-SLAM Simulator (version 2.0) ------------------ This simulator demonstrates a simple implementation of standard EKF-SLAM. It permits simple configuration via 'configfile.m' to perform SLAM with various control parameters, noises, etc. Also various switches are available to choose known data-association versus gating, etc. The key file in this simulator is called 'ekfslam_sim.m'. Type 'help ekfslam_sim' for more information of how to use it. In addition to on-line animations, the simulator returns a data-structure of the logged state information for off-line processing. An example use of this data is shown in m-file 'plot_feature_loci.m', which plots the trajectories of the landmark estimates. Tim Bailey and Juan Nieto 2004. Note on Global Variables ------------------------ This version of the simulator uses global variables for all large objects, such as the state covariance matrix. While bad programming practice, it is a necessary evil for MatLab efficiency, as MatLab has no facility to avoid gratuitous memory allocation and copying when passing (and modifying) variables between functions. With this concession, effort has been made to keep the code as clean and modular as possible.