GSNNSimulation

  本文针对SLAM算法中数据关联过程，提出了一种简单易行的改进方法，将欧氏距离与马氏距离结合用于数据关联。算法不必计算地图所有特征与所有量测之间的马氏距离，而是首先利用相对简单的欧氏距离计算缩小了待关联特征的搜寻范围。利用人工合成数据的仿真结果表明，改进后的数据关联方法可以大幅减少系统计算量，提高关联效率，且不会造成错误关联的增加。(This article SLAM algorithm for data association process, a simple method to Euclidean distance combined with the Mahalanobis distance for data association. Algorithms do not have to calculate all the characteristics of the map with all measurements between the Mahalanobis distance, but the first to use a relatively simple calculation of Euclidean distance to be associated characteristics of narrowing the search. The use of synthetic data simulation results show that the improved methods of data association can significantly reduce the system to calculate the volume and improve the efficiency of association, and will not lead to errors associated with an increase.)

all8.mat
errorNN.mat
fNN.mat
mnoise.mat
NIS.m
NISDA.m