本文针对SLAM数据关联中使用最为广泛的最近邻方法作了改进，利用特征估计位置与载体预测位置之间的欧氏距离计算代替了全部特征与每个量测之间的马氏距离计算，避免了大量的矩阵乘法计算。该算法简单易行，降低了算法的计算复杂度，有利于SLAM算法的实时执行，且关联效果与全局最近邻法相同-In this paper, SLAM data association in the most widely used methods of improving the nearest neighbor, using the characteristics of the estimated location and the location of vector prediction Euclidean distance between the calculation instead of all the characteristics of each measurement between the Mahalanobis distance calculation, to avoid a large number of matrix multiplication calculation. The algorithm is simple, the algorithm to reduce computational complexity and is conducive to the implementation of real-time SLAM algorithm, and the effects associated with the same overall nearest neighbor method